# Chapter 3 The Mixed Economy

The devil may indeed dwell in the details, but we first need to find an angel or two in the abstractions that govern […].

— Edward J. McCaffery (2002 K117)

Welfare states are governed by the economic abstractions arising from the co-existence of market and command.

When you organize production and distribution of goods and services by market and by command, as welfare states do, the two contradictory systems “complexly interact” (Perrow 1999) and easily produce “unintended consequences” (Merton 1936) as for example, when ().

To this day, the mixed economy of Postwar Western Europe, in its idealized form, is the closest thing to an angel to ever emerge from this uneasy coexistence between market and plan. To understand first-order desiderata of welfare state design, we need to understand the conceptual compromise of the mixed economy. Let me first reiterate the logic of its two constituent systems.

#### 3.0.0.1 Market vs. Planned Economy.

$sec:market-vs-command$

We can organize economic production and distribution in at least one of two ways,

by centralized, coercive command or

by decentralized, voluntary exchange.50

1. In an ideal-typical (Weber 1920) command economy, whoever wields an effective monopoly on the use of force also directs the economy. A worker constructs, say, a railroad (production) and is fed by a farmer (distribution), both in fear of — however indirect — bodily harm from the monopolist of violence.51

2. In an ideal-typical market economy,52 violence is threatened only to maintain property rights and enforce contracts. People freely exchange goods and services at equilibrium prices that balance the costs to the producer and the utility to the consumer (, p. ). A worker constructs a railroad (labor) in return for an enforceable promise to consume (property) a given amount (wage), which he then redeems in a similar exchange with a farmer for food.

#### 3.0.0.2 Capitalist Welfare State is a Pleonasm.

$sec:interface$ Welfare states combine elements of command and market in the service of equity.53 Specifically, welfare states coercively adjust the distributional outcomes of markets.5455

Welfare states insure their citizens against certain individual risks (disability, sickness, unemployment), fight “poverty” by instituting (unconditional or means-tested) minimal living standards and, sometimes, reduce inequality by compressing the income and wealth distribution of the citizenry.

To reach each of these goals, welfare states have to intervene in voluntary exchanges between buyers and sellers. For better and/or for worse, welfare states change market equilibria. No matter the legal structure, welfare state institutions never exist outside the market: even “nationalized”56 health care needs to buy doctors (at what salary?) and drugs (at what price?) on free markets. No matter the labeling, welfare state institutions never exist independently of the market: even social “insurance”.57 alters labor market outcomes (who should, and can bear the burden?).

Welfare state institutions can interact with markets in more or less attractive ways: they can have a smaller or larger dwls (, p. ), and they can have well or ill-defined incidences (, p. ), but they always interact.

Welfare state institutions can expand (UK socialized health care) or contract (Germany social health insurance) the scope of its command, but they will always interface with the market at some frontier between the two systems, as illustrated in (p. ). This interface is ill-defined, as two incompatible logics collide, twice:

1. The state demand curve (dashed in green) for doctors or hospitals breaks down, as the marginal utility of an additional doctors to citizens (the ultimate consumers) cannot be known.58 The state therefore has to determine the citizenry demand, usually based on a cba or a related procedure — all of which are really nothing but fancy names for sophisticated economic planning.

2. The state cannot command the required supply (of doctors or hospitals), but must instead buy the supply from state revenues on free markets.59 Again, the logic of the market breaks down: the state as the only buyer (of doctors or hospitals) creates a monopsony, causing distributive effects (to the disadvantage of doctors or hospitals) and welfare losses (a dwl). Alternatively — maybe more plausibly — the producers of medical care may capture the planning body and extract rents, distributing away from the public and also causing a welfare loss. Either way, pricing poses unavoidable problems: somewhere down the line, producers require a price, but without atomistic demand, no pareto-optimizing price can be found. This is not to say that if health care — or some other welfare service — were completely privatized, such pareto-optimal equilibrium would be reached. In fact, government procurement of health and other services can be understood as a response to for these goods (p. ). Arguably, government procurement replaces one (horrendous) market failure with another, (milder) form of failure.

I aim here to distill for welfare, what Edward McCaffery (2002) urges us to do about tax: “to find an angel or two in the abstractions that govern $...$(2002 K117).

I look for these angels in an ideal closed, mixed economy. The account I provide in the (p. ) – ) does not resemble any real existing economy, where abstractions are often shrouded in historical idiosyncrasies, and angels rarely found amidst imperfect policies. But this is a question of the (p. ), and to know what is materially doable and normatively desirable we need (p. ), not a posteriori reality.

Even without the details, the abstractions alone need considerable space to be explained. I urge readers to take the time, even if much will seem familiar and some things appear remote to welfare, let alone taxation or democracy. They are not: from (p. ) to network effects (p. ), (p. ). Missing any one of these abstractions, we cannot know what welfare and taxation can, and should do.

Four disclaimers apply to my tentative answer on this very big question:$sec:disclaimers$

1. $itm:not-original$ Not Original. The perspective I take here is hardly original. Many others have, in greater width (Stiglitz 2002) or depth (Sinn 2004), with narrower (Scharpf 1997b) or different foci (Zürn 2000) discussed the first-order shortcomings of “negative” integration (in the eu: Scharpf (1997b)), and economic liberalization everywhere (Stiglitz 2002), and have, in that context, defined the conditions of a mixed economy.

I aim here to review the works of others and to restate some fairly conventional economic concepts in order to build a first-order checklist of welfare state design.

2. $itm:no-test$ No Positive Test. I cannot myself muster the methodological rigor or provide the econometric data, to test the first-order theories of welfare state design, but rely on mainstream literature instead.

The economics of the welfare state are vastly complex, incompletely understood and any policy initiative requires careful (empirical) investigation to balance the often contradictory imperatives of economic policy.

Moreover, as I later find for taxation, reforms of the magnitude implied herein may strain against the current limits of economic science: experimental designs lack in external validity of a vastly complex modern economy and simulations often lack the data, or even computing power to model such inframarginal changes.

3. $itm:no-calibration$ No Calibration. I offer no calibration of the mixed economy and its institutions, and, for the purpose of this chapter, advocate no particular balance between market and state, efficiency and equity or any of the other trade-offs a mixed economy may face. Instead, I highlight the capacities and dysfunctions of markets and potentially able to mitigate these shortcomings (p. ). I consider under (p.  ) and hypothesize how they might explain the (, p. ).

4. $itm:little-macroeconomics$ Little Macroeconomics. I limit this discussion to very basic concepts of the real economy, and ignore many of the more complex models, especially of finance and money. Modern macroeconomics, including such powerful frameworks as the IS/LM model are important (originally Hicks 1937), but would go beyond the already lengthy treatment here.

I also suspect and hope that macroeconomics is best investigated by experts and its policy imperatives safely implemented by technocrats. Monetary policy, for instance, may not raise deep normative questions or offer vexing trade-offs in need of democratic adjudication: its imperatives hinge on contested and imperfect, but merely first-order, positive findings on a mass psychology of price and cost signals (see , p. ). To a lesser extent, finance, too, may be politically epiphenomenal: money and other property rights move in tandem with, and are secondary to material economic exchanges in the ideal mixed economy (see p. ). To the extent that polities can agree on specific and measurable objectives (such as price stability, or risk diversification), macroeconomic policy really can be delegated to independent central banks or other regulatory bodies.

By limiting the discussion to a few rudimentary, but deeply understood concepts of the real economy, I also hope to reconnect regional integration and the welfare state to an econonomic imagination (paraphrasing Mills 1959) of our material affairs as a household — only with a cast of billions.60

Inevitably, much of the detail and complexity that policy makers have to consider, will fall by the wayside.

If the defining characteristic of a welfare state is its uneasy union of market and plan, we must first understand the broader interplay of exchange and command in the mixed economy. (p. ) summarizes exchange (or market) and command (or state) institutions to address five material dimensions of the human condition: (p. ), (p. ), (p. ), (p. ) and (p. ). This is a slightly expanded set, inspired by Musgrave (1959)’s (1959) seminal definition of basic public policy functions: allocation/efficiency, distribution and stabilization (for example, as cited in (Bordo, Markiewicz, and Jonung 2011, 4). It also corresponds to Samuelson (1954)’s authoritative textbook (recently 2005) desiderata of a mixed economy:

1. to let scarce resources be efficiently allocated by competitive markets,

2. to improve the equity of market outcomes through redistribution,

3. to provide public goods by government procurement and

4. to limit inherent market instability by financial regulation and well-directed monetary and fiscal policies

As stands, the game appears stacked against the market, as there is no list of government failures. In fact, failures abound in the command economy: without credible information about individual utility (confer Hayek 1931), and an elegant mechanism for their aggregation (confer Lerner 1944; Lange 1934; Debreu and Arrow 1954) resources are easily wasted and misallocated and even command components in market economies are prone to government failure (Coase 1964). Moreover, a market economy seems to be closely related to liberal democracy,61 and bloated command economies may corrupt politics,62 or even threaten the very constitution of freedom (Hayek 1944; Friedman 1962).

Here again — to “economize on moral disagreement” (Gutmann and Thompson 2004 K226) and to lend credence to my later conclusions — I conservatively place the burden of proof on the state: production and distribution by markets should only be replaced or altered by command when the market can demonstrably not achieve the desired outcomes, that is, when problems materialize (row 4 in ).

I now discuss the five material dimensions of human need in turn.

### 3.0.2 Positional Races

## Taxation and Welfare

Why does taxation matter for human welfare?

As @Schumpeter knew, "in the modern world, taxation *is* the social contract" [@Martin2009a, 1, emphasis in original], even though social scientists have since paid little attention to it [@Tilly2009, K191].
Tax matters especially for current OECD-style welfare regimes, in which markets and states must co-exist [@Stiglitz2011].
In such a mixed economy, government must be able to draft some privately-owned resources to serve its --- ideally democratic --- *command*, without unduly altering the prices of --- ideally competitive --- *exchanges* [@Ardant1975, 165f].
<!-- TODO ref chapter on mixed econ here -->
Much of modern social integration occurs in the balance of these two contrasting logics to produce and allocate resources, "enmesh[ing] us in the web of generalized reciprocity that constitutes modern society" [@Martin2009a, 3].
State plans make the conditions *for*, and set limits *to* *individual* action, as when government builds roads or collects sewage fees.
Conversely, market exchanges produce much of the "fungible resources" used for *collective* choice, as when government pays road builders or procures sewage pipes [@Martin2009a, 4].

Of all conceivable institutions to govern the interface of states and markets, taxation --- not price controls, not expropriation, not debt, not printing money, not tariffs --- is the most equitable, efficient and sustainable [@MusgThet1959; @Stiglitz2011].
<!-- TODO ref tax-matters chapter -->
Welfare states, with their penchant for market interventions for equity, efficiency *and* sustainability, especially, rely on good taxes.

Still, taxation everywhere in the OECD is in crisis.
As alternative sources of economic relief --- monetary expansion and sovereign debt --- are maxed out, structural misalignments persist, and previously forestalling (asset) bubbles have burst into their days of reckoning, public revenues appear to be strictly limited by the longtime coming contradictions of current tax regimes [@Streeck2013].
The popular mixture of (progressive) income, (proportional) consumption and (regressive) payroll taxes appears to offer only harshly unattractive tradeoffs between equity and efficiency [@McCafferyHines2010], as bases have shrunk and schedules flattened [@Ganghof2006].
At the same time --- possibly partly as a result --- inequalities of incomes and wealth have widened [@Butterwegge; @Wagner2007; @Grabka2007].[^no_wealth_distro]
<!-- TODO mention here that inequality findings are contentious -->

[^no_wealth_distro]: Data on the distribution of wealth is conspicuously hard to come by [@Crouch2004, 158] or ordinally summarized in deciles, rendering much of the inequality invisible.

If governments cannot raise the resources necessary to meet democratic demands without incurring prohibitive costs, the social contract is fraying [@Crouch2004].
<!-- %so so shource -->
Whichever way governments now turn, absent better tax, they will violate the post-war capitalist compact of stable, widely shared growth [@Pierson2002; @StreeckMertens2010].

[$sec:market-solutions-production$]{#sec:market-solutions-production label="sec:market-solutions-production"}
Assuming, as I do, (p. ) and conditions for (p. ) markets have at least two attractive properties:
[^20]

1.  when all participants have made all profitable exchanges,
[^21]
markets produce at the quantity and price where the costs to the producers equal the willingness to pay of buyers (see , p. ).
Consumers and producers in any given market enjoy maximum *surplusses*:
all consumers pay prices (at least incrementally) below the utility they receive (by area $A$), all producers receive prices (at least incrementally) above the costs they incur (by area $F$).
In this *competitive equilibrium* (see column 1, row 4 in , p. ), no one can be made better off without making someone else worse off:
it is *Pareto optimal*.
[^22]
[^23]

<!-- ![Market Equilibrium of Supply and Demand[]{label="fig:supply-demand"}](supply-demand){#fig:supply-demand width="100%"} -->

2.  The *price system* (see column 1, row 4 in , p. ) makes markets into supreme "information processors" [@Hayek1931].
[^24]
Because decisions are decentralized, markets can elegantly aggregate dispersed information where a central planner would have to gather them by bureaucratic means (such as a [cba]{acronym-label="cba" acronym-form="singular+short"}).
Because market decisions are always backed by private costs, the market price system can reveal and communicate (some!) private information, where a central planner may face distorted (inflated) information about cost and utility.

##### Conditions for Perfect Competition. {#sec:perfect-competition}

The above properties of market equilibria hold only under the strict assumptions of perfect or atomistic competition.
In one formulation, this entails [@McDowell2006 157f.]:

[$itm:infinite-buyers-sellers$]{#itm:infinite-buyers-sellers label="itm:infinite-buyers-sellers"}
There are so many consumers and producers in the market that an offer or bid by any one of them will have a negligible impact on prices.
Everyone is a price taker.

2.  *Zero barriers to entry and exit*.
[$itm:easy-entry-exit$]{#itm:easy-entry-exit label="itm:easy-entry-exit"}
Firms can start or cease to produce a good at relatively little cost and effort.
All markets are wide open.

3.  *Perfect factor mobility.*
[$itm:perfect-factor-mobility$]{#itm:perfect-factor-mobility label="itm:perfect-factor-mobility"}
In the long run, labor, capital and other inputs to production can move to wherever they earn the highest rents.
It is hire and fire.

4.  *Perfect information*.
[$itm:perfect-information$]{#itm:perfect-information label="itm:perfect-information"}
Consumers know all prices and qualities of goods, producers know all prices and qualities of factor inputs.
People are omniscient and powerful calculators of utility.

5.  *Profit-maximizing firms*.
[$itm:profit-maximizing-firms$]{#itm:profit-maximizing-firms label="itm:profit-maximizing-firms"}
Firms sell at the price and quantity that maximizes their profit.
[^25]
Other, exogenous criteria are not part of firm decision making.
Given the other assumptions of perfect competition, firms sell where marginal cost equals marginal revenue.
[^26]

6.  *Homogeneous products.*
[$itm:homogeneous-products$]{#itm:homogeneous-products label="itm:homogeneous-products"}
Goods produced by one supplier are the same as those produced by another supplier of the same category.
Inputs provided by one factor owner are the same as those provided by another owner of the same factor.
All goods and factor inputs are completely commodified.

To this, one might add, according to @Wikipedia2012:

7.  *Zero transaction costs.*
[$itm:zero-transaction-costs$]{#itm:zero-transaction-costs label="itm:zero-transaction-costs"}
Buyers and sellers can exchange goods, services and make contracts at zero cost.
Search, information, bargaining, policing and enforcement costs are assumed away.
There is no friction.

8.  *Constant returns to scale.*
[$itm:constant-returns-to-scale$]{#itm:constant-returns-to-scale label="itm:constant-returns-to-scale"}
For any additional unit produced, costs rise by the same amount, no matter how much units are produced.
The cost function is *linear*, marginal costs are constant.
There are no (dis)economies of scale or scope.
[^27]

Constant returns to scale are related to, but distinct from assumption [$itm:easy-entry-exit$](#itm:easy-entry-exit){reference-type="ref" reference="itm:easy-entry-exit"} on .
Excessive economies of scale imply difficult entry.
Conversely, difficult entry implies varying marginal costs at low output.

9.  *Property rights*
[$itm:property-rights$]{#itm:property-rights label="itm:property-rights"}
are well established.

Lastly, --- if somewhat redundant, because axiomatically assumed away by neoclassical welfare economics --- I would add:

10. *Utility equals willingness to pay.*
[$itm:same-budgets$]{#itm:same-budgets label="itm:same-budgets"}
The first theorem of welfare economics --- that above specified perfect markets equilibrate at pareto optimality over *initial* distributions --- is often misrepresented to imply that individual *utility* is adequately expressed in willingness to pay, on which free market exchanges operate.
This shortcut works only under the most heroic assumption of all:
that all market participants have the same budget constraint.

If everyone invoking the first theorem were to represent it in full, including the crucial, qualifying "...over *given* allocations", this condition for perfect competition would be unnecessary.
Alas, many do not.

Moreover, equating individual utility with willingness to pay constitutes what might be called the of welfare economics (p. , because it *both* assumes equal *and* unequal budget constraints:
the former to maximize utility, and the latter as incentive.
For a thesis so thoroughly grounded in neoclassical orthodoxy as this, it seems appropriate to feature this contradiction prominently.

I note wherever I relax some of these strict (and rarely plausible) assumptions in the following sections (summarized in , p. ).

It should be clear then that I make the above, conservative or neoclassical assumptions less out of conviction, but for their logical elegance and to "economize on moral disagreement", as [@GutmannThompson-2004-aa K226] have suggested.
I hope these positions will be widely acceptable to readers on the political right, and provisionally tolerable to readers on the left.
If I can show taxation to be wasteful, unfair and unsustainable and democracy to be outmatched, assuming even such liberal orthodoxy, sweeping reform should be all the more obvious.

#### Market Failures.

\phantomsection
[$sec:market-failures$]{#sec:market-failures label="sec:market-failures"}
Markets operate efficiently on private goods:
people can be *excluded* from their *rival* use (see ).
In this case, social costs and utility match private cost and utility.

From some goods, people cannot be effectively excluded and/or are not rivals in their consumption.
[^28]

In these cases, social and private costs and utility diverge and markets may fail.
Goods are overproduced when social cost is higher than private cost (*negative externality*), and goods are underproduced when social benefit is higher than private benefit (*positive externality*).
This problem holds more broadly for *public goods*, *common goods* and *natural monopolies* (see column 1, row 5 in ),
[^29]
summarized in according to @Samuelson-1954-eu's typology of goods [-@Samuelson-1954-eu].

##### Failure: Public Goods

\phantomsection
[$sec:public-good$]{#sec:public-good label="sec:public-good"}
are *underprovided* by markets, because no one can be prevented from using them (non-exclusion), and they do not get used up (non-rivalry).
Potential buyers can always free-ride on other's purchase and are therefore unwilling to pay producers adequately.
Defense is a public good and fireworks are canonical examples of public goods.

##### Failure: Common Goods

\phantomsection
[$sec:common-good$]{#sec:common-good label="sec:common-good"}
are *overused* on markets, because they are rival but again, no one can be prevented from using them [@Hardin-1968-aa].
Potential buyers can free-ride without paying the adequate price for exploiting the rival commons [@Hardin-1968-aa].
[^30]
[^31]
Clean air is a common good, and so may be the enlightened understanding of the electorate [@Caplan2007].

##### Failure: Natural Monopolies

\phantomsection
[$sec:natural-monopoly$]{#sec:natural-monopoly label="sec:natural-monopoly"}
arise where economies of scale abound in the production or distribution of goods or services, such that only a single or very few suppliers can profitably exist.
Natural monopolies are *mispriced at the margin* because after initial fixed costs --- which few single consumers would be willing or able to pay --- marginal cost become negligible.

Excessive economies of scale often occur in businesses dominated by fixed, rather than variable cost.
Sewer systems, electricity grids or search engines can be natural monopolies with prohibitively high entry costs
[^32]
for basic infrastructure (sewers, electricity masts, web indices) and later, negligibly small marginal costs for adding an additional consumer.
Natural monopolies can incur welfare losses in two ways:

1.  If only one market supplier exists, it may charge monopoly prices causing a deadweight loss of underconsumption.

2.  In the extreme, given the high marginal cost for the first buyer, no first buyer may come forth and the otherwise
[^33]
pareto-improving natural monopoly may not be provided at all.

##### Failure: Principal-Agent Problems.

\phantomsection
[$sec:principal-agent-problem$]{#sec:principal-agent-problem label="sec:principal-agent-problem"}
Principal-agent problems are one market failure of the broader class of information asymmetry problems (row 5, column 2 in , p. ), where at least one party to a trade knows less about the service, good or risk being exchanged (pioneered by Nobel laureates @Akerlof-1970-aa, Stiglitz [-@Stiglitz1976] and @Spence1974).

In principal-agent problems, the asymmetrically known quality is the effort exerted by the agent on behalf of the principal.
When agent (say, manager) effort cannot be fully observed by principals (say, owners), and principals have interests (say, long-term returns) diverging from those of agents (say, a pet project), agents may be able to cheat on their contracts.
When agents shirk successfully, they will exert less (or ill-directed) effort than would be pareto-optimal.
In the extreme, the market between principals and agents breaks down completely, as principals anticipate shirking agents and forego the transaction altogether.

Applied game theory and related disciplines offer a host of incentive designs to realign interests of principals and agents, which I need not discuss here comprehensively (but see @Tirole2006).
Solutions include (p. ) --- which creates unemployment --- or deferred compensation and tournaments --- which invites risk-seeking (Holt 1995).
Efficiency wages and tournaments try to alter the probabilistic calculus of would-be shirkers by promising outsized instead of *marginal* rewards and punishment for whichever effort *is* (randomly) observed.
Deferring (part of) the compensation to later may make agents more far-sighted, but will still strictly cap their downside risk, especially when only bonuses are deferred (for example, stock options).
The worst that can happen to an agent in any of these schemes is to loose their job, the tournament promotion or their bonus.
By contrast, the worst that can happen to a principal, is to lose everything.

In addition, if the observations of effort on which such schemes are based are spotty or isolated --- as they often are --- agents can "game the system" and incentives may turn ineffective, or even perverse.

Both (p. ) and tournament compensation also increase economic inequality:
instead of marginal productivity, they reward (p. ) and threaten losers with unemployment.

All these incentive design schemes fall short of the one genuine solution to realign principal and agent interests:
to either sell agents stock or charge them a substantial sign-up fee [@Tirole2006], effectively making agents into principals, too.
Only then can they bear both the full upside and downside risk of the enterprise.
To be able to take on such risk, of course, agents must own substantial assets, which they may not have in an unequal economy.

Principal-agent problems fail markets if, and to the extent that, inequality in assets prevents people from taking equal risks in otherwise welfare-enhancing joint projects.
They are one of the cases, where inequity makes for inefficiency, too.

This is not merely a theoretical conumdrum, but a very real problem for postindustrial and knowledge-based economies (for example, Lisbon Strategy, EU 2020, @Bell-1973-aa).
[^34]
Almost by definition, an economy based on knowledge and innovation will display information asymmetries.
The effort a knowledge worker (say, a programmer) puts in, cannot easily be observed, because that would require the principal to acquire the exact same specialized knowledge (say, a programming language).
Similarly, a would-be innovator (say, an [ict]{acronym-label="ict" acronym-form="singular+short"} entrepreneur) will always know more about her nascent and uncertain innovation than any possible investor, because otherwise, the investor would have done the project herself, already.

In short, for competitive markets to do their magical "stochastic tinkering" [@Taleb2007 211], people need to be equipped and incentivized to act on their local and uncertain ideas, and to specialize.
For a *homo economicus* at least, there can be no *entrepreneurship* without some *ownership*, too.

#### State Responses to Market Failures

\phantomsection
[$sec:state-responses$]{#sec:state-responses label="sec:state-responses"}
States can respond to market failures by fiscal and regulatory interventions.
I discuss them for each type of good (rows 6--8, column 1 in ).

##### Fixing Public Goods.

\phantomsection
[$sec:public-good-response$]{#sec:public-good-response label="sec:public-good-response"}
States can step in to *provide public goods* or subsidize their private provision, both out of the public purse (fiscal policy).
There is no regulatory or monetary response to public goods failure.
[^35]

##### Fixing Common Goods.

\phantomsection
[$sec:common-good-response$]{#sec:common-good-response label="sec:common-good-response"}
States can protect overused commons by a regulatory policy of "fencing in the commons".
[^36]
By doing so, governments follow the [@Coase1960] theorem.
It holds that markets can pareto-optimally resolve externalities if transaction costs are sufficiently low, and if property rights are well-defined.
[^37]
The Coase theorem is often erroneously cited to argue against state intervention.
Maintaining and *issuing new property rights* are, of course, *regulatory* state interventions.
[^38]

Alternatively, states can resolve "Tragedies of the Commons" [@Hardin-1968-aa] fiscally by slapping a *Pigouvian tax* (@Pigou1912, popularized by @Baumol1972) on using the commons.
[^39]
The Pigouvian tax prices in the externality of using a common good.
[^40]

There is no monetary response to overused commons (see , p. ).

<!-- %here are two important notes from prospect theory, and Kahnemann 2012
%I think insurance requires only the weaker (contested by Kahnemann) expected utility hypothesis (bernoulli), who says that risk aversion can be explained merely by diminishing marginal utility (specifically, argues Bernoulli, utility is a logarithmic function of wealth).
%Maybe I can do here with risk aversion and don't need LOSS aversion (based on reference points), which is what Kahnemann is out to explain, and with him prospect theory.
%Consider a weird graph about that in Kahnemann.
% The mixed economy can be understood as those institutional crutches that help system-one humans to behave as system-two humans, because system-two can't be relied upon in the long run, it's too effortful to use.
%So we use institutions to get us there -- because system two is the better.
% Kahnemann raises a big stink with indifference curves:
%he says neoclassical economics is wrong there, because prospect theory shows that people have reference points from which they are loss averse, suggesting that there is no such thing as an indifference curve (I am not sure that is formally right, the indifference curve just wouldn't be linear, as it might be).
% my problem with Kahnemann is that he meanders between positive and normative theory.
%I would think prospect theory is a positive finding, but it need change normative theory.
%Of course, Kahnemann might say, well if utility is experienced according to prospect theory, than maybe THAT is the kind of utility that we should maximize.
%I disagree:
%I think we should let system two reign, and therefore also need not abandon neoclassical economics, we just must make sure that there is always good and enough crutches around. -->

<!-- %\subparagraph{Fixing PCA-Problems}
%\subparagraph{Entrepreneurship Needs Broad-Based Ownership.} \phantomsection \label{sec:Ownership} More deductively, market economies can be thought of as welfare-maximizing because they capture decentralized and/or privately known information, and let diverse solutions compete.
%Nassim Nicholas \cite{Taleb2007} has put this succinctly by praising aggressive trial and error'' (\emph{ibid.}:
%xxi) in free markets that allow people to be lucky'' (\emph{ibid.}:
%xxi)\footnote{Being a quantitative trader by profession, \cite{Taleb2007} actually abandons rational choice when faulting Karl \cite{Marx-1867-aa} and Adam \cite{Smith-1776-lq} for believing that free markets work because of rewards.}.
%This basic impetus for capitalist creativity is expressed in desideratum \ref{des:Entrepreneurship}:

%\begin{desideratum}[Entrepreneurship]
%   A desirable tax will allow entrepreneurs to make their own production decisions according to their independent judgement of private and/or local information.
%   \label{des:Entrepreneurship}
%\end{desideratum}

%But what is required for this stochastic tinkering'' (\citealt{Taleb2007}:
%211) to work?

%Principal-agent theory suggests that maximum effort may not be exercised when effort of agents is imperfectly or non-observable, or other information asymmetries prevail.
%This of course --- unobservable effort and information asymmetries --- are likely features of highly differentiated knowledge economies\footnote{Calls for more startups, patents and research spin-offs, particularly in Germany, may serve as evidence for suboptimal incentive design under the status quo.}, where people produce an idea, not a piece of welded metal (cf.~\citealt{Bell-1973-aa}).
%Game theoretic incentive design suggests that information asymmetry problems can be resolved either by making agents shareholders or by charging them substantial sign-up fees, both of which require substantial assets to begin with \citep{Tirole2006}.

%In short, for competitive markets to work their magic, people need to be equipped and incentivized to act on their local and diverse ideas.
%For \emph{homo economicus}, there can be no \hyperref[des:Entrepreneurship]{entrepreneurship} without some \hyperref[des:BroadOwnership]{ownership}, too.

%This ties in with the simple capitalism desideratum presented earlier, but it also adds a specification:

%   A desirable tax allows for or promotes a broad-based ownership of the means of production.
%\end{desideratum}

%include footnote for all desideratum's of where they are leading.
%reference the same footnote

%\subsection[Welfare Gains]{Welfare Gains:
%How Taxes Can Make the Pie Larger} \label{sec:PurposesOfTaxation}
%The burden of proof is on the state for interventions in the market.
%Witnesses for the defense are summarized in \autoref{tab:ends-mixed-economy} and are entertained in the below.

%Taxes and other state interventions in the economy can in fact enhance market outcomes in several ways:

%this figure is obsolete, it is now \label{tab:ends-mixed-economy}

%\begin{description}
%   \item[Redistribution.] \phantomsection \label{sec:fiscal-redistribution} Governments may respond to excessive \emph{inequality} by taxing people proportionally or progressively to redistribute resources.
%While inequality is further discussed in \autoref{sec:tax-justice} on \hyperref[sec:tax-justice]{equity}, it does bear on efficiency, too, as is argued in \autoref{sec:InequalityIsInefficient}.

%   \item[Risk Pooling.] \phantomsection \label{sec:state-insurance} People can possess \emph{asymmetric information} about things of uncertain quality they exchange on the marketplace.
%Insurance of unemployment, health or disability are principal examples, where sellers of risk (insurants) typically know more about their own risks than buyers of risk (insurers)\footnote{This relaxes perfect competition condition {itm:PerfectInformation} (\hyperref[itm:PerfectInformation]{perfect information}).}.
%Less-informed buyers (insurers) of risks may expect bad risks for \emph{all} insurants, causing premiums to rise and driving low-risk sellers out of the market entirely.
%This mechanism may repeat until no exchanges are made at all, defeating the purpose of insurance.

%   Governments can avoid these \emph{lemons markets} by forcing everyone to take out insurance \citep{Akerlof-1970-aa}.
%When risks are universal --- as is arguably the case for unemployment, health and disability --- the premiums for these insurances are effectively taxes.

%   \item[Public Goods] \phantomsection \label{sec:public-good} can be enjoyed by an arbitrary number of people without exhaustion (non-rivalrous) and no one can be excluded (non-excludable) from its using (\citealt{Samuelson-1954-eu}, summarized in \autoref{tab:Types-Of-Goods}).
%National defense is one example.
%Because people in larger groups can always free-ride on the provision of public goods by others, they are likely to be \emph{underprovided} by self-seeking individuals or markets \citep{Olson-1971-aa}.

%   Governments can improve welfare by providing public goods out of tax revenue.

%the types of good table happens earlier, as    \label{tab:types-of-goods}

%   \item[Common Goods] \phantomsection \label{sec:common-good}  are rival in their consumption but do not allow exclusion:
%everyone can benefit, but they can be exhausted \citep{Samuelson-1954-eu}.
%A \emph{Tragedy of the Commons} occurs when people overuse the common good \citep{Hardin-1968-aa}\footnote{Elinor \cite{Ostrom1990} criticizes the canonically assumed failure of commons in social science and provides an empirically grounded account of their successful, non-coercive governing.}.

%\begin{quote}
%   \emph{What is common to the greatest number has the least care bestowed upon it.}\\*\\*
%   Aristotle, Politics, Book II, Chapter 3 (384 b.c.-322 b.c.)
%\end{quote}

%   Government can avoid the \emph{negative externality} of exhaustion of commons by \emph{pricing in} the costs of its use, an approach also known as Pigouvian taxation (\citealt{Pigou1912}, popularized by \citealt{Baumol1972})\footnote{The alternative, non-tax solution of issuing property rights on the commons (for example through an Emissions Trading Scheme), thereby making it an ordinary private good is of course a government'' solution, too.
%Markets cannot maintain, let alone introduce new property rights.}.%check definitively whether it is Pigovian our Pigouvian.
%note from correction:
%Note that some solutions are costly.

%   \item[Natural Monopolies] \phantomsection \label{sec:natural-monopoly}  arise where economies of scale abound in the production or distribution of goods or services, such that only a single supplier can profitably exist\footnote{This relaxes perfect competition condition \ref{itm:infinite-buyers-sellers} (\hyperref[itm:infinite-buyers-sellers]{price taking}) and {itm:easy-entry-exit} (\hyperref[itm:easy-entry-exit]{easy entry and exit}).}.
%Excessive economies of scale often occur in businesses dominated by fixed, rather than variable costs.
%Sewer systems, electricity grids or search engines can be natural monopolies with prohibitively high entry costs for basic infrastructure (sewers, electricity masts, web indices) and later, negligibly small costs for adding an additional consumer.
% Natural monopolies can incur welfare losses in two ways:
%if only one market supplier exists, it may charge monopoly prices causing a deadweight loss of underconsumption.
%Conversely, if several suppliers exist in one market, each of them may be unable to invest at optimal levels (underprovision).
%%This ain't quite right.
%I'm missing the marginal vs.\ average cost problem.

%   Governments can avoid the deadweight losses of natural monopolies by regulating them (for example, last mile ICT in Germany), franchising or outsourcing them (for example, local railway in Germany), enforcing common carriage (for example, electricity in Germany) or by nationalizing them (for example, public ownership of motorways in Germany).

%   \item[Easy Market Entry.] Problems of prohibitive entry costs are not limited to natural monopolies\footnote{This relaxes perfect competition condition {itm:easy-entry-exit} (\hyperref[itm:easy-entry-exit]{easy entry and exit}).}.
%Competition can also be hampered by market players who enjoy excessive economies of scale by sheer size or past learning curves.
%Aside from regulatory responses (antitrust), governments can react proactively by means of infant industry protection or other industrial policy, the contested (de)merits of which are not under further consideration here\footnote{If infant industry protection \emph{is} considered welfare-enhancing, its medium-term welfare losses to the polity (either in the form of a DWL of subsidizing or protectionism) become a public good to the extent that a successful infant industry generates positive externalities for the  rest of the economy.
%As such, infant industry protection should be partially financed out of general revenue.}.
%\end{description} -->

<!-- %\paragraph{Diminishing Utility is a Fact (Hard) to Observe} Diminishing utility is a plausible intuition grounded in our very nature\footnote{Some findings suggest that we are neurologically hard-wired to display diminishing utility in our feelings \citep{Ng-1997-aa}.}.
%Our quintessential evolutionary features, both our metabolism and propagation display starkly diminishing returns:
%you can only eat so much and rear so many children.
%Survey measures of self-reported happiness also support diminishing utility of wealth and income \citep{Veenhoven-2000-aa, Nickell2008}\footnote{If \hyperref[sec:positional-race]{positional consumption} is, in fact, rampant, survey measures may yet underestimate the diminishing utility of wealth and income.
%When people extract utility from levels of consumption \emph{relative} to others, their valuation of \emph{absolute} wealth and income is probably inflated.}.
%this is a great argument, look for empirical evidence -->

<!-- Some theories for self-reinforcing inequality
%\paragraph{Path Dependence or Cumulative Causation.} The third dynamic of \emph{path-dependent or cumulative causation} refers to situations where small initial state differences in performance lead  to additional opportunities, reinforcing initial inequality.
%These additional, scarce opportunities may be awarded to individuals (or firms, or regions) based on easy but imperfect measures (think SAT scores).
%They may also be awarded based on probabilistic predictions on future performance (think past scholarships), further increasing a self-reinforcing dynamic.
%In the worst, most inequitable (and inefficient case), they are awarded based on meaningless, randomly occurring differences (think mental state on day of testing), haphazard selections (think first come, first serve) or systematic measurement bias (think habitus expectations by assessors).

%Malcolm \cite{Gladwell} illustrates this dynamic in his account of \emph{Outlier} hockey stars in Canada, whose birthdays are significantly more often in the early months of the year.
%\citeauthor{Gladwell} attributes this to very early elite selection in Canadian hockey and a cut-off point between the different leagues on each December 31st, giving children hockey players born early in the year a slight developmental advantages over their peers, amplified by the additional training they receive if initially selected.

%Path-dependent or cumulative causation of inequality are frequently observed in educational systems, particularly in those which track students early (as in Germany).

%These dynamics also apply elsewhere, where initially only slight differences lead to divergent experiences, reinforcing inequality and leading to further opportunities, for example when individuals or firms benefit from learning curves or economies of scale after initial jobs.

%\paragraph{Self-reinforcing Network Effects in Scale-Free Distributions.} The fourth dynamic are self-reinforcing network effects in homopholous networks with a scale-free graph distribution.
%Networks are formalized as \emph{graphs} comprising of \emph{nodes} (individuals, firms) interconnected with (directed, undirected and/or weighted) \emph{edges} (aquaintance, contracts) \citep{Kleinberg-2009-oz}.
%The \emph{degree} of a node is given by the number of edges emerging from it.

%I will first illustrate the graph theory of innovation diffusion \citep{Bass1969} with a simple example, the diffusion of fax machines.
%Assume that individuals (nodes) decide on whether to purchase a fax machine based on the technology's inherent value (the \emph{innovation coefficient}) and the value they realize from the number of their peers (degree) also owning a fax machine (the \emph{imitation coefficient}).
%Consider first what would happen in a grid network, where every individual is connected to each adjacent individual (all nodes have the same degree).
%Assume next that initial adopters are randomly distributed.
%Fax machines would, largely determined by their inherent value, proliferate (or not) relatively homogenously over the entire network.
%It will be unlikely that any given individual (node) has many more peers with fax machines than any other individual, rendering the imitation coefficient relatively inconsequential.
%illustrate this -->

##### Fixing Natural Monopolies.

\phantomsection
[$sec:natural-monopoly-response$]{#sec:natural-monopoly-response label="sec:natural-monopoly-response"}
Governments can avoid the deadweight losses of natural monopolies fiscally, by nationalizing them (for example, public ownership of motorways in Germany) and ideally charging users a fee at *average* cost (for example, trucks and coaches on federal motorways in Germany).
[^41]
Governments can also procure natural monopoly goods and services from private firms and charge users a fee at average cost (for example, local railway in Germany).

Alternatively, governments can allow privately held natural monopolies but tightly regulate them to enforce pricing at average cost and avoid a monopoly [dwl]{acronym-label="dwl" acronym-form="singular+short"} (for example, *last mile* [ict]{acronym-label="ict" acronym-form="singular+short"} or electricity in Germany).

There is no monetary response to natural monopoly problems (see footnote [$fn:monetary-commons$](#fn:monetary-commons){reference-type="ref" reference="fn:monetary-commons"}).

#### Monetary Policy for Price Stability.

\phantomsection
[$sec:price-stability$]{#sec:price-stability label="sec:price-stability"}
Monetary policy contributes best to efficient production --- and almost all other dimensions of material human need --- by staying out of the way of markets with stable prices to allow efficient exchanges in the first place.
[^42]
[^43]
When prices rise (inflation) or fall (deflation) overall, otherwise pareto-optimal exchanges may be hampered.

The welfare losses of inflation include

hoarding (of real assets),

drowning out relative price changes (noise),

increasing transaction costs (such as shoeleather costs
[^44]

[^45]
), as well as

general uncertainty and possibly, unrest.

Cost-wage spirals
[^46]
make inflation self-reinforcing, potentially escalating into hyperinflation.
Inflation is also believed to set off the business cycle [@Friedman1970].

Deflation, while scarcely observed in the western world in the post-Bretton-Woods regime,
[^47]
is similarly damaging.
It also makes

transactions more costly and

heightens uncertainty.

Deflation can also

cause the hoarding of cash and

trap liquidity
[^48]
and self-reinforce into a deflationary spiral.

In the long run, monetary expansion (or contraction) should follow the output of the economy $G$, so that price levels stay stable.
[^49]

Whether inflation and deflation are, as the monetarists would have it, "always and everywhere a monetary phenomenon" [@Friedman1970]
and therefore caused by an over-expansion or contraction of the money supply in the first place, or whether it has it roots in the real economy as the Keynesians would argue, is a very complicated empirical question but --- luckily for the author --- irrelevant to the state job of ensuring an efficient market place.
No matter "who dunnit", monetary policy should aim at price stability, and, perhaps, counteract real price shocks --- if and to the extent that they occur --- with (p. .).

Pooled Risks: Saving the Pie {#sec:risk}
----------------------------

[^50]

#### The Human Condition of Risk.

\phantomsection
[$sec:human-nature-of-risk$]{#sec:human-nature-of-risk label="sec:human-nature-of-risk"} Humans inhabit a volatile environment, marred by low probability, but high impact events (*black swans*, according to @Taleb2007), for example a serious work accident.
Unfortunately, humans also tend to ignore precisely such low probability, but high impact events [@Taleb2007], and overestimate the probabilities of favorable outcomes [@Baron2000 44], especially when they have few cognitive resources available.
When they give it some thought, most people *avoid grave downside risks*:
for example, most will prefer the certain but low cost of car liability insurance over the rare but high cost of paying for a car accident (column 2, rows 1--3 in ).
[^51]

#### Market Solutions to Risk: Insurance.

\phantomsection
[$sec:insurance$]{#sec:insurance label="sec:insurance"}
Markets provide *insurance* as a ready-made institution to address this human need for down-side risk aversion (column 2, row 4).
Insurants can buy protection from their grave downside risk, say, a car crash, by pooling their individual risks.
An insurance deal stipulates that all insurants will regularly chip in a small amount to cover the few unlucky car wreckers, in return for the promise that they, too, will receive a payout if they crash their cars.
[^52]

Aside from car insurance, markets do sometimes, to some extent, provide insurance against the four big life risks commonly associated with the welfare state:

unemployment,

sickness,

accident and

disability.
[^53]

Market insurance of these life risks *does not* involve a redistributive component, though that is easily assumed for unemployment insurance.
The *voluntary* exchange of premiums for coverage, in car and all other insurance, is a pareto-improvement:
all risk-averse insurants are better off, by hedging against downside risks.
Insurance is not obligatory and whoever finds it unnecessary (as may be the case for rich individuals who face limited downside risks) is not affected.

#### Market Failure in Insurance: Asymmetric Information. {#sec:asymmetric-information}

Insurance markets may fail when buyers and sellers possess asymmetric information about the risks to be insured.
[^54]

##### Ex ante,

\phantomsection
[$sec:adverse-selection$]{#sec:adverse-selection label="sec:adverse-selection"}
insurants with privately known high risk may disproportionately take out insurance.
Insurers, anticipating such *adverse selection*, but unable to tell high-risk ("lemons") from low-risk ("cherries") applicants,
[^55]
may expect bad risks for *all* buyers, causing premiums to rise and further driving low-risk insurants out of the market.
This *lemons market* mechanism may repeat until no exchanges are made at all, defeating the purpose of insurance [@Akerlof-1970-aa].

Adverse selection abounds in the insurance of the big life risks.
Ex ante, insurants know more about their likelihood to become unemployed, sick, to be in an accident or become disabled than their insurers.
[^56]

##### Ex post,

\phantomsection
[$sec:moral-hazard$]{#sec:moral-hazard label="sec:moral-hazard"}
sellers of risk (insurants) might take on more risk than they otherwise would have, causing moral hazard and in turn drive up premiums.

Moral hazard, too, may occur in the insurance of big life risks.
Ex post, insurants may be willing to live less healthily, or more recklessly than they would without insurance.

#### State Responses to Asymmetric Information in Insurance.

\phantomsection
[$sec:state-insurance$]{#sec:state-insurance label="sec:state-insurance"}

##### Ex ante,

states can resolve lemons markets by *regulatory* means if they force everyone at risk to take out insurance (@Akerlof-1970-aa, @Barr)
(column 2, row 8).
[^57]

States can also resolve lemons markets by *providing benefits* out of the treasury (column 2, row 9).
As the risks of unemployment, health, accident and disability are near-universal,
[^58]
the contributions for such state-run insurance are effectively taxes.
Some states (such as Germany) outsource insurance to quasi-fiscal organizations.
This "social insurance" may make a difference in administrative, legal or rhetorical terms, but its economics are that of a state-run insurance and its revenues are taxes.
Financing public insurance out of dedicated "social contributions" (usually regressive payroll taxes) instead of general revenue only adds a specific distributive component (a regressive tax on labor) to the overall tax schedule.
[^59]
[^60]

Mandating insurance or, equivalently, providing benefits out of the public purse need not redistribute resources over and above the Kaldor-Hicks improvement from resolving a lemons markets.
Properly understood, state interventions to hedge *individual* risks are meant to improve the efficiency, not equity of outcomes.
Real policy often differs from this (p. ) and layers distributive components on top of the Kaldor-Hicks improvement.
In Germany, for instance, social contributions are proportional, but capped and public health insurance covers an insurant's children at no additional cost.
These additions to schedule and benefits may or may not be desirable, but they properly belong in the realm of (p. ).

##### Ex post,

states and markets have essentially the same, clumsy method to reduce moral hazard:
[^61]
they re-individualize some of the risk by asking for co-payments or provide incentives for prudent behavior.
Exploiting moral hazard is a (p. ), and can therefore be resolved either by (partial) property rights (co-payments) or by Pigouvian taxation (incentives).

Equitable Distribution: Slicing the Pie Fairly {#sec:distribution}
----------------------------------------------

#### The Human Condition of Inequality

\phantomsection
[$sec:human-nature-of-inequality$]{#sec:human-nature-of-inequality label="sec:human-nature-of-inequality"}
Humans, the social animals, can deal with material scarcity in two ways [@Pickett-2009-kx]:

"because members of the same species have the same needs as each other, they have the potential to be each other's worst rival" [-@Pickett-2009-kx 197] but also,

"the potential to be each other's best source of cooperation, learning, love and assistance of every kind" [-@Pickett-2009-kx 198].

Dominance Strategies.

:   The first, @Hobbes-1651-aaian [-@Hobbes-1651-aa]
strategy is one of dominance:
*homo homini lupus est*, man is a wolf to his fellow man.
[^62]
Following *dominance strategies* (column 3, row 1 in ), humans --- much like their primate cousins, the chimps --- maximize their classical evolutionary fitness (@Darwin1859, recently @Dawkins1976) by relying on individual power to secure access to scarce resources (food, shelter, females).

Affiliative Strategies.

:   The second strategy is one of *affiliation* (column 3, row 3) and mutuality:
to be your "brothers keeper" ([kjv]{acronym-label="kjv" acronym-form="singular+short"} Bible, Genesis 4:9).
[^63]
Following affiliative strategies, humans ---as their other primate cousins, the bonobos --- maximize inclusive fitness (@Hamilton1964, popularized in @Wilson1975), or fitness emerging at the group-level [@Wilson2012] by cooperation, trust and reciprocal altruism [-@Pickett-2009-kx 202ff].

*Dominance hierarchies* (column 3, row 2) arise as dominance strategies prevail and (usually male) members of a species fight for higher status to successfully monopolize resources.
All but the highest status individuals are held to suffer from such heightened *relative* inequality (*ibid.*).

Instinct does not determine us to follow dominance strategies (as the chimps, according to *ibid.*) or affiliative strategies (as the bonobos, according to *ibid.*), and so we need culture and institutions to strike that balance.
*Both* markets and states can strike that balance, and reign in on dominance hierarchies.

#### Market Equity

\phantomsection
[$sec:market-equity$]{#sec:market-equity label="sec:market-equity"}
On competitive markets, people enter into voluntary exchanges that, at least, make everyone better off (if not necessarily by the same amount).
[^64]
Interactions under dominance hierarchies are *no such* pareto improvements;
the (physically) stronger will extract from the weaker all she can, possibly short of killing the weaker party if prolonged extraction is in the interest of the stronger party.
[^65]

*Real* existing market economies have sometimes --- but not always --- created sharp inequality.
Still, against the backdrop of dominance hierarchies, *ideal* market economies with , institutionalizing pareto-improving, voluntary exchange must be considered a civilizing achievement.
[^66]

#### Excessive Inequality

\phantomsection
[$sec:inequality-dynamics$]{#sec:inequality-dynamics label="sec:inequality-dynamics"}
Real existing, imperfect, modern markets display some excessively inequitable dynamics that may compromise this ability to at least somewhat compress dominance hierarchies.
In the following, I discuss some of these dynamics and further relax some assumptions of (p. ).

##### Efficiency Wages.

\phantomsection
[$sec:efficiency-wages$]{#sec:efficiency-wages label="sec:efficiency-wages"}
Efficiency wages are, counterintuitively, wages *above* the market-clearing rate.
At efficiency level, wages are so high, that some people cannot find gainful employment and must remain unemployed.
[^67]

Wages may be above market clearing rate because employers want to attract more applicants to choose from, because local traditions demand it, or even to feed malnourished workers.
Two other suggested reasons stand out:

1.  *Reducing turnover.*
Employers may pay above-clearing wages because they want to avoid costly turnover.
When faced with both potential unemployment, attractive, and, especially, seniority-graded wages, workers may not look for a job elsewhere (for example, @Salop1979, on [ldcs]{acronym-label="ldc" acronym-form="plural+short"} @Stiglitz1974a).

2.  *Avoiding shirking.*
Employers may also pay above-clearing wages because they want to deter incompletely contracted and incompletely observed workers from shirking.
Because employers (principals) can observe effort only sporadically and imperfectly, they will catch shirking workers (agents) only some of the time.
Would-be shirkers face some probability of getting caught, a reward for continued shirking, and a punishment for getting caught and may optimize their behavior accordingly.
Efficiency wages can thereby solve this (p. ):
by increasing both wage and unemployment rate, employers widen the spread between the two probabilistic outcomes for would-be shirkers (getting caught, not getting caught).
Risk-averse employees may then work hard to avoid the inflated downside risk of unemployment and loss of a generous wage [@Stiglitz1984].

Both when they increase wages to reduce turnover, and to avoid shirking, employers move the labor market out of equilibrium.
Otherwise pareto-improving employment is lost, and economic welfare foregone.
Still, efficiency wages may persist, because they are efficient --- individually utility optimizing --- for employers, if inefficient for the economy as a whole.
Employers enjoy lower turnover and shirking, at some price of higher wages, but they also push some of the social cost of above-clearing wages to everyone else.
If and to the extent that principal-agent problems are otherwise unavoidable, its socially costly remedy may not be considered a market failure:
there may be no way to make anyone better off (the unemployed) without making someone else worse off (well-paid employed and satisfied employers).

But even if efficiency wages were to destroy no welfare, they certainly redistribute it.
Here again, as in (p. ) or (p. ), people will be rewarded and punished *probabilistically* (not deterministically) and *out of proportion* to the marginal contributions they made --- or could make --- to the market economy.

##### Winner-Take-All.

\phantomsection
[$sec:winner-take-all$]{#sec:winner-take-all label="sec:winner-take-all"}
Five paradigms and stylized dynamics of today's economy point to the possibility of runaway social inequalities that may result in distributions much akin to dominance hierarchies, where whoever is at the top reaps most or all of the benefits [@Frank1996].

1.  \phantomsection
[$itm:non-linear-returns$]{#itm:non-linear-returns label="itm:non-linear-returns"}
*Non-linear returns to scale* in indivisible human capital may disproportionately reward highly-skilled workers, as demand for their skills increases in the knowledge economy and they cannot be replaced by several less-skilled workers.
[^68]
[^69]
Similarly, some highly-skilled work can easily by scaled up (an algorithm can easily be deployed millions of time), but many other occupations cannot be easily scaled as production reaches a physical limit (a hairdresser can only do so many haircuts a day) (originally @Rosen1981, recently @Taleb2007).
Given the [eu]{acronym-label="eu" acronym-form="singular+short"}'s proclaimed goal to become the leading knowledge economy in the world, it is to be expected that non-linear returns to scale in indivisible human capital will further increase [@Commission2007].

*Baumol's cost disease* is a related, but inverse concept.
According to [@Baumol1965], some sectors (such as manufacturing) enjoy faster productivity growth than others (such as nursing), but, competing for the same laborers, both sectors must raise salaries.
In violation of (neo)classical dictum, the wages of, for example, nurses rise, even though they have --- supposedly --- not concomitantly gained in productivity.
Much of these sectoral productivity gains are reflected in similarly increasing non-linear returns to scale in human capital:
as the indivisible innovations (say, laser welding) of engineers are easily scaled up in manufacturing, such innovations are absent in nursing and would resist scaling.
If labor is, as [@Baumol1965] assume, in fact, (p. ) workers will be free to choose jobs with above-linear returns to scale until pay equilibrates at the same level across scalable and non-scalable work.
In this scenario, diverging productivities are priced into *sectoral costs*, not *worker pay*.
As a flip-side to diverging pay from above-linear returns, unscalable sectors will either disappear or become *relatively more* expensive to *consumers* (as seems to be the case with nursing).
is, of course, an implausible assumption.
If workers cannot, in fact, freely choose to work in scalable (computer engineer) or unscalable occupations (hair stylist), those in unscalable occupations will bear the brunt of diverging productivities in lower relative pay.

The truth, as often, will lie somewhere in between, and greatly depend on circumstance.
Some of the divergence between scalable and unscalable occupations will fall on workers, some on consumers and much will be split.
[^70]

2.  [$sec:cumulative-causation$]{#sec:cumulative-causation label="sec:cumulative-causation"}
*Path-dependent rewards and cumulative causation* may also let winners take all or most.
If small initial state differences in performance lead to additional opportunities, initial inequality will be reinforced (@Jackson1968 [@Merton1988] recently popularized by @Gladwell).
[^71]
This pattern of path-dependent or cumulative causation is often observed in educational systems, particularly in those which track students early (as in Germany) or where social permeability is low (as in much of Europe) [@OECD2006].

3.  [$sec:network-effects$]{#sec:network-effects label="sec:network-effects"} *Self-reinforcing network effects*
occur where economic activity occurs along homopholous networks with a scale-free graph distribution (for an introduction to graph theory, see @Kleinberg-2009-oz).
[^72]

As actors (nodes) opt to interact with similar people (homopholy, for example, @Mcpherson2001), and opportunity (innovation) spreads (cascades) only among tightly nit groups (clusters) [@Bass1969] resulting utility (degree) distributions will be decidedly non-normal (scale-free, or fractal [@Mandelbrot2004]).

Whenever features of economic consequence
[^73]
permeate through these networks of tightly clustered, self-similar nodes, opportunities and rewards will be a function of that same power-law distribution.

Inequality, by the very structure of modern society, will be excessive and self-reinforcing [@Cozzi2009; @Keller2005; @Andriani2007].

A similar dynamic, applied to economic activity in space, is implied in the agglomeration and scale effects of (p. ).

##### Different Budget Constraints.

\phantomsection
[$sec:different-budget-constraints$]{#sec:different-budget-constraints label="sec:different-budget-constraints"}
The magic of the (p. ) works over *given* allocations.
In the real world, wealth and income disparities cause market participants to have different budget constraints.
A higher budget constraint will inflate their willingness to pay and a lower budget constraint will deflate their willingness to pay, both at constant levels of utility.

As distributions in the real world are not a blank, egalitarian slate, the demand and supply curves are distorted by differential budget constraints.
Voluntary exchange no longer necessarily equilibrates at the pareto-optimum of *utility*, but at the pareto-optimum of *willingness*, and thereby *ability to pay*, a very imperfect and distorted proxy.

If anything, this glossed-over difference between absolute *utility* and budget-dependent *willingness to pay* is the original, logical sin of neoclassical welfare economics.

On the one hand, different budget constraints distort prices without any informational gain for @Hayek1931's superior information processing system:
budget-distorted willingnesses to pay are *misinformation*.
The old computer science adage applies here, too:
*garbage in*, *garbage out* (GIGO).
A market that equilibrates at multi-million yachts for people with outsized budgets and at malnutrition for others with very small budgets may, be formally pareto-optimal, but it may --- among other things --- not be utility-efficient in any meaningful way.
[^74]

On the other hand, such distortions of individual utility are precisely the sticks and carrots that incentivize homo economicus in market economies.
Market economies reward individuals by letting them amplify their utility signals with a larger budget constraint and they punish individuals by forcing them to subdue their signals with a smaller budget constraint.
These informationally useless spillovers from one exchange (Steve Job's inventions) to other, unrelated exchanges (Steve Job's consumption of yachts) are not merely a side-effect of market economies, they are their motivational essence.

As original sins go, they can never be redeemed in full --- at least not in this world.
So it is with the dirty little secret of neoclassical welfare economics:
to equalize all budgets at all times would be to abandon a market economy to the fullest.

Still, neoclassical economists and everyone else who relies on the first theorem must at least repent this sin by always confessing to it, and by highlighting its normative and policy implications.
Else may not await purgatory, but lies ideology.
We must not seal off, but open up an intellectual edifice to its criticism, we must not assume away but render transparent its inherent contradictions.

##### Diminishing Marginal Utility.

\phantomsection
[$sec:diminishing-marginal-utility$]{#sec:diminishing-marginal-utility label="sec:diminishing-marginal-utility"}
With each additional unit of goods and that people gain, the added utility may fall (theoretically by @Lerner1944 [23], recent empirical support from @Ng-1997-aa [@Veenhoven-2000-aa; @Nickell2008]).
Highly inequitable market outcomes will still be formally pareto optimal,
[^75]
but may leave great Kaldor-Hicks improvements unrealized as the poor stand to gain greater marginal utility than the rich would have to give up at their higher levels of consumption.
In short utilitarian slogan, given diminishing marginal utility, perfect markets do *not* yield "the greatest good for the greatest many" [@Mill1863].

##### Positional Externality.

\phantomsection
[$sec:positional-race$]{#sec:positional-race label="sec:positional-race"}
Thorstein [@Veblen1899] has suggested that people consume excessively not merely to fulfill some manifest function inherent to the good purchased, but that they buy expensive things to "heighten or reaffirm social status" [@Merton-1968-aa 123].
Veblen goods are bought not *in spite of*, but *because* of their price, which is ideally publicly known and serves to communicate wealth and status to others.

People consume conspicuously based on a *relational* rationale:
what matters is what *other* people think about the cost (or sophistication) of a Veblen good.
If people consume conspicuously --- at least partly --- to display status, the rationale is also *relative*:
what matters is how much you spend in relation to what other people spend.

Conspicuous consumption can then --- at least partly --- be understood as *positional* consumption:
people can elevate their relative status if, and to the extent that they spend *more* than their peers.
For this positional motivation for consumption, only *relative* prices (and qualities) matter, not absolute cost or utility.

Positional consumption can be modeled as a [pd]{acronym-label="pd" acronym-form="singular+short"}, as in .
[^76]

+:------------+:-+:-------+:--------+:-+
|             |  |        |         |  |
+-------------+--+--------+---------+--+
+-------------+--+--------+---------+--+
| \cline{3-4} |  |        |         |  |
+-------------+--+--------+---------+--+
|             |  |        |         |  |
+-------------+--+--------+---------+--+
| \cline{3-4} |  |        |         |  |
+-------------+--+--------+---------+--+
|             |  |        |         |  |
+-------------+--+--------+---------+--+
| \cline{3-4} |  |        |         |  |
+-------------+--+--------+---------+--+

: Positional Consumption as a Prisoner's Dilemma[]{label="tab:pd-positional"}

\scriptsize{The Joneses and the Does are peers, positional consumers and in the market for a new car.
Payoffs are the net of cost of car ($VW=0$, $BMW=-5$) and status gain from driving the \emph{relatively} more expensive car ($+10$ for superior, $-10$ for inferior, else $0$).
Larger payoffs are better.\\
Prices of the BMWs are inflated so as to include the societal costs of wasteful consumption, here, as in an ideal world, accruing only to the buyer.}
In a [pd]{acronym-label="pd" acronym-form="singular+short"} of positional consumption, *more* excess will always be a strictly dominant strategy, and the unique Nash equilibrium.
The social welfare optimum of mutual moderation will not be reached.
People would derive same utility at collectively lower levels of consumption, but cannot do so for fear and anticipation of other players cheating by unilaterally consuming more.
[^77]

Consumers of Veblen goods may get stuck in a [pd]{acronym-label="pd" acronym-form="singular+short"}, racing to keep up with others [@Frank1987].
Positional consumption wastes resources because it imparts no additional utility and exerts a negative externality on others by devaluing their purchases (a process known as *expenditure cascades* *ibid.*).

##### Monopsony Employers.

\phantomsection
[$sec:monopsony-employers$]{#sec:monopsony-employers label="sec:monopsony-employers"}
When few, big firms face many, unorganized workers, labor markets may turn monopsonistic and cause both a welfare and a distributive effect.
[^78]

Welfare

:   is lost as some workers, facing lower wages stay home, who might otherwise have been gainfully employed at a higher, but still profitable wage.
Under this monopsony [dwl]{acronym-label="dwl" acronym-form="singular+short"}, some otherwise pareto-improving exchanges are not undertaken.

Redistribution

:   .
Further, compared to a competitive labor market, economic surplus is redistributed from workers to employers.

This power imbalance between labor and capital is arguably one of the deepest contradictions that capitalism created.

The free market, universalizing the commodity form, treats capital and labor as equal factors of production.
Capital and labor, are fundamentally *not* equal factors of production and have vastly different bargaining power for at least three reasons:

1.  Historically --- and to date --- capital tends to be *concentrated*:
few have capital.
By contrast, labor was historically, and is to date, greatly dispersed:
many are workers.
Larger groups with widely dispersed interests are harder to organize [@Olson-1971-aa].
Industrial production requires the cooperation and coordination of many, often thousands of workers who face collective action problems as well as colossal transaction costs.
In the absence of institutions fostering collective action, employers may be able to permanently suppress worker cooperation by attracting individually rational defection.
Collusion here, as always in atomistic markets --- for better or for worse --- is a [pd]{acronym-label="pd" acronym-form="singular+short"}.

2.  Capital is *easily cooperated and coordinated* into production, if pooling of resources over several capitalists is necessary at all.
Almost by definition,
[^79]
any market economy will provide the institutions to pool capital (such as a public company) and match small owners with large projects and large projects with small owners (for example *convenience denomination* by financial intermediaries).
By contrast, labor is (again increasingly) unorganized.

3.  Capital can afford to lie idle and forgo a rent.
Owners can instead convert it into the goods and services of a modern economy.
Labor cannot afford to be unemployed, as consumption possibilities would then be reduced to the outputs of autonomous subsistence production, which are far below those produced under separation of labor.
Not trained in, and often prevented from subsistence production, the worker faces the very real possibility of starvation if she does not enter into an agreement with capital.
In economic terms, labor supply will always be *relatively less price elastic*, compared to capital supply.

As a result, the free market, treating labor and capital equally as commodity, instituted a systemic power imbalance.

#### Redistributive Policy.

\phantomsection
[$sec:redistributive-policy$]{#sec:redistributive-policy label="sec:redistributive-policy"}
Government has a number of tools to redistribute market outcomes.
These tools differ in effectiveness and efficiency:
some alleviate only some kinds of market inequality (for example, , p. ), and some are very costly for any given increment of resource redistributed (for example, p. ).

##### Price Controls.

\phantomsection
[$sec:price-controls$]{#sec:price-controls label="sec:price-controls"}
Government can respond to inequities by *regulating* price ceilings (such as rent control, in ) or price floors (such as minimum wages, in ).
As intended, binding price controls shift welfare:
consumers gain in surplus from price ceilings, producers gain in surplus from price floors (column 3, row 6 in ).

But they gravely lack in efficiency and effectiveness.

![The [dwl]{acronym-label="dwl" acronym-form="singular+short"} and distributive effects of a price ceiling with unit-elastic demand and supply (for example, of housing)](price-ceiling){width="100%"}

\scriptsize{Compare with \nameref{fig:supply-demand} in \autoref{fig:supply-demand} (p.~\pageref{fig:supply-demand}).}
[$fig:price-ceiling$]{#fig:price-ceiling label="fig:price-ceiling"}

![The [dwl]{acronym-label="dwl" acronym-form="singular+short"} and distributive effects of a price floor with unit-elastic demand and supply (for example, of labor)](price-floor){width="100%"}

\scriptsize{Compare with \nameref{fig:supply-demand} in \autoref{fig:supply-demand} (p.~\pageref{fig:supply-demand}).}
[$fig:price-floor$]{#fig:price-floor label="fig:price-floor"}

Efficiency.

:   In addition to the desired, redistributive zero-sum component, price controls also destroy welfare:
much like taxes (), they cause a [dwl]{acronym-label="dwl" acronym-form="singular+short"}.
Price controls are a negative-sum allocation.
Ceilings push prices (say, rents) below equilibrium levels, decrease supply (of flats) and increase demand (for flats), causing excess demand.
Some potential tenants cannot find good housing, even though they would be willing to pay a higher rent or some potential landlords will not built or improve housing, even though willing tenants could be found.
Conversely, floors push prices (say, wages) above equilibrium levels and increase supply (of labor) and decreases demand (for labor), causing excess supply (unemployment).
Some workers cannot find gainful employment, even though they would be willing to work for lower pay or some potential employer will not hire an available worker, because her cost (wage) outweighs her utility (productivity).
Under binding price ceilings or floors, some otherwise pareto-improving exchanges will not be undertaken, causing decrepit real estate
[^80]
or unemployment, respectively.

Such [dwls]{acronym-label="dwl" acronym-form="plural+short"} are the opposite of gains from trade, they invariably reduce to total consumer (producer) surplus of people able to buy (sell) goods at prices lower (higher) than their reservation price.

The size of the [dwl]{acronym-label="dwl" acronym-form="singular+short"} of price controls depends on the relative price elasticities of supply and demand (just like the , p. ).
The more price inelastic demand, the smaller is the [dwl]{acronym-label="dwl" acronym-form="singular+short"} of price floors:
even at pushed-up prices, few buyers will be able to exit the market and almost all sellers will find buyers.
Conversely, the more price inelastic supply, the smaller is the [dwl]{acronym-label="dwl" acronym-form="singular+short"} or price ceilings:
even at pushed-down prices, few sellers will be able to forego sales and almost all buyers will find sellers.

In the real world, and depends on many factors (p. ).

In housing markets, supply (landlords) may be inelastic in the short run (for existing housing), but very elastic in the long run (for improved, or new housing).
[dwls]{acronym-label="dwl" acronym-form="plural+short"} (or decrepit housing) are probably sizeable.
In labor markets, too, demand (employers) will likely not be price inelastic across the board.
Often, labor can be substituted by capital.
Here, too, [dwls]{acronym-label="dwl" acronym-form="plural+short"} (or unemployment) will be sizeable.
In any case, it will be very hard for states to gauge elasticities and calibrate price controls accordingly to specific industries, time periods and regions.
[^81]

Effectiveness.

:   Price controls are also limited in their effectiveness.
They may, in principle --- if at colossal cost and complexity --- redress some of the inequities of markets (p. ), (p. ) and (p. ).
Short of a near planned economy, they do little to dampen s (p. ):
consumers can simply buy different (more extravagant) categories of goods, or more of the same to fulfill positional impulses.
They can also hardly reign in on (p. ).
By definition, their demand for labor is highly price elastic, given plenty of alternatives.
Worker supply of labor, absent another employer, is almost perfectly price inelastic.

Governments should, however, respond to monopsony employers in labor markets either by anti-trust action or by leveling the playing field.
Regulating a right to strike and the right to form trade unions can put monopoly employees opposite monopsony employers in collective bargaining.
[^82]
Employment protection and labor contract regulation can also alleviate the power imbalance in labor markets.

##### Affirmative Action.

\phantomsection
[$sec:affirmative-action$]{#sec:affirmative-action label="sec:affirmative-action"}
To a limited extent, governments can also respond to winner-take-all markets with affirmative action or equal opportunity legislation.

##### Redistribution by Fiscal Policy

\phantomsection
[$sec:fiscal-redistribution$]{#sec:fiscal-redistribution label="sec:fiscal-redistribution"}
The principal redistributive tool of government is a progressive tax, by which government collects more from the rich to pay for public expenditures, and/or finances handouts to the poor.
[^83]

##### Redistribution by Monetary Policy

\phantomsection
[$sec:distributive-effects-of-inflation$]{#sec:distributive-effects-of-inflation label="sec:distributive-effects-of-inflation"}
Monetary policy cannot effectively mitigate inequitable market outcomes.
Inflation and deflation redistribute wealth, but not between the rich and the poor.
Instead, inflation redistributes wealth between credits denominated in real terms (for example, house ownership), debts denominated in nominal terms (for example, a fixed-rate mortgage) and credits denominated in nominal terms (for example, fixed-rate pension) and debts denominated in debts denominated in real terms (for example, a futures contract) as summarized in .

\centering
![The distributive effects of inflation, with some examples[]{label="fig:distributive-effects-of-inflation"}](distributive-effects-of-inflation){#fig:distributive-effects-of-inflation width="100%"}

Such arbitrary redistribution does not correspond to of markets (p.) and cannot be defended on normative grounds.
[^84]

Additionally, redistributing by monetary policy causes severe (p. .
Today, as financial products have diversified, redistributing based on denomination and between creditors and debtors will have increasingly arbitrary results.
[^85]

Monetary policy on distributive goals, as on (p.), shines by staying neutral.
It must strive to maintain price stability to avoid any arbitrary and undue distributive effects that would otherwise interfere with the distributive dynamics of state and market.

Consistency Over Time: Saving Tomorrow's Pie {#sec:time}
--------------------------------------------

#### The Human Condition in Time

Human beings are myopic planners [@KahnemanTversky1979].
They discount the future hyberbolically;
the more remote a future event is, the more humans discount its rewards (@Ainslie1975, @Thaler1981).
Human beings also succumb to herding:
they do as others around them do.
Both these frailties can lead to time inconsistency, where we simultaneously hold incompatible preferences for our present and future selves, for example when we would like to retire somewhere warm and mild, but also want a new car every three years.

To be time consistent, we have to do unto future generations (or selves) as we would have them do unto us (the Golden Rule of reciprocity).
Applied to the material world, we have to save --- at least --- at that level, which allows the highest constant level of current and future consumption (@Phelps1966a, @Solow1956).
If we are more generous to our children, or consider technological innovation to be endogenous, we might have to save even more.

Time consistency bears both on equity and efficiency.
An *equitable* policy ensures same (discounted)
[^86]
utility for people living today vs. people living in the future.
An *efficient* policy ensures maximum (discounted) growth over all periods (along the long-term growth path of the economy).

#### Interest

\phantomsection
[$sec:interest$]{#sec:interest label="sec:interest"}
In the exchange economy, capital markets align preferences of present and future generations (and selves) by compensating savers with an interest.
Ideally, risk-free interest rates,
[^87]
and other (hedged) returns on capital (for example, stock market indices) equilibrate at that level, where future marginal utility equals present marginal cost of saving.
At this intertemporal optimum,
[^88]
present generations (and selves) *do* unto future generations (and selves) as vice versa.

Under the assumptions of neoclassical orthodoxy, efficient capital markets should guarantee *some* degree of time consistency.

##### Garbage In, Garbage Out.

But because markets are ultimately *aimless* and operate on (assumed!) *exogenous* preferences, the *formal* ability of capital markets for intertemporal optimization is strictly limited.
Interest rates --- as all other prices --- are elegantly aggregated (time) preferences of people.
If these preferences are inconsistent, or amoral, so, too, will be the resulting market "optimum".
As in statistical analyses, no matter the algorithm, the quality of the output depends strictly on the quality of the inputs.

People may be relatively consistent --- or at least our best bet --- to gauge some components of observed discount rates, including the likelihood of exogenous growth (innovation), or continued human existence (compare footnote [$fn:3components$](#fn:3components){reference-type="ref" reference="fn:3components"} (p. ).
For these components, people may input reasonably good 'data', and markets will output reasonably good prices.

By contrast, the 'data quality' of *pure discount component rate of the future* --- discounting future utility simply because it lies in the future --- is deeply questionable for reasons (compare footnote [$fn:3components$](#fn:3components){reference-type="ref" reference="fn:3components"} (p. ):

1.  *Inconsistency.*
People are, as cognitive psychology has shown, *time inconsistent* even in their own lives, and the short- to medium term (@Ainslie1975, @Thaler1981).
If they are similarly inconsistent when deciding how much saving to supply at any given interest rate, *overall* interest rates will be distorted.
Myopic individuals will *cause* the very inflated interest rates meant to keep their short-sightedness in check, and overall saving will be too low.
[^89]

2.  *Morality.*
As [@Samuelson2005 54] has recently reminded us, *term uncertainty*, especially pure discounting of the future is a *normative concern*.
Just how much we value the welfare of future generations and selves is a *moral* judgement.
There is nothing in our nature or our environment that could positively tell us what a *true* rate would be, there are only normative judgments of what a *right* rate should be.
Once people have agreed on such a rate, markets can do the @Hayek1931ian number crunching.
But markets, because they will accept *any* rate as input, cannot solve this moral problem for us.
Instead, only a democratic polity --- and its government --- can decide the pure discount rate.

People often say they want their children to live a *better* life.
If similar sentiments are deeply-held and widespread, people might democratically decide to save *more* than intertemporal *equity* demands.
If indeed, we shall give to our children *better*, than we ourselves received, we are at least looking for a *Kaldor-Hicks* (not Pareto!) improvement without and above the equilibrium capital return.

Even ignoring such substantive concerns, real world capital markets are also plagued by *formal* intertemporal failures, causing their own costly and inequitable time inconsistency.
The capitalist algorithm not only receives bad inputs, but it is also ridden with bugs, that fail both in the short and the long term.

#### Short-Term Intertemporal Failures {#sec:short-term-inconsistency}

Economic activity tends to fluctuate in the short-term.

are periodic fluctuations, typically relatively mild in amplitude and slope.
They are endogenously caused by lumpy, lagged decision-making of market participants, for example in inventory cycles.
[^90]

According to the dissenting, minority view, business cycles are *true* shocks that may or may not be endogenously amplified, but are exogenous in origin (@Kydland1982).
[^91]

##### Bubbles and panics

can lead to abrupt and dramatic fluctuations in economic activity, with great slope and amplitude.
Bubbles and panics envelope market actors when they make their decisions based on the anticipated decisions of others, in a beauty-contest type game [@Keynes1936].
Similar rationales also emerge when making privately informed decisions creates a positive externality for other participants to free-ride on your inadvertently disclosed information [@Banerjee-1992-aa].
[^92]

##### Fluctuations are Costly.

Both types of fluctuations in economic activity are inefficient because they temporarily leave factors of production idle (overused) and increase periodic transaction costs.
Fluctuations leave factors of production (for example, an assembly line) idle during the downturn or building up excess capacities during the upturn (for example, a real estate overhang).
The costs of short-term fluctuations are particularly virulent in labor markets, where hiring is expensive and unemployment can degrade worker morale.
Aside from these material costs, short-term and uncertain employment (and economic prospects in general) cause stress and hardship.

Such idleness or mania diverts the economy away from its (exogenously) long-term growth path and thereby slows down the economies progress along the growth path.

By contrast, "true" exogenous shocks (such as an earthquake) or genuinely new information about future utility of present savings (for example, because of a new technology) is not a deviation from, but a *shift* of the long-term growth path.
It is, for better or for worse, unavoidable.

#### Short-term: State responses

States can respond by regulatory, fiscal and monetary means to reduce the frequency, depth and duration of economic fluctuations.
In the following, I discuss the state responses to a downward deviation from the long-term growth path of the economy.
States should likewise, if with opposite sign, respond to upward deviations from long-term growth.

##### Regulatory

States can reign in on short-term downturns simply by outlawing respective actions of market participants.
[epl]{acronym-label="epl" acronym-form="singular+short"}, for example, can include lay-off protection, making it harder for employers to fire workers in a downturn.
Other regulatory interventions include the bans on short-selling of recent 2007ff fame.

Regulatory responses can be effective, but blunt instruments as they also affect economic activity outside of endogenous downturns.
Generous lay-off protection, for instance, can cause a [dwl]{acronym-label="dwl" acronym-form="singular+short"} of unemployment during upturns, keeping the economy under its long-term growth path.
Regulatory interventions can also fail to distinguish between endogenous and exogenous fluctuations in the economy overall and some of its markets.
[^93]
As collateral damage, generous lay-off protection may prevent workers from moving to new, more productive firms and sectors (for example services rather than mining).
Similarly, bans on short-selling may prevent or defer necessary market readjustment in case of real, exogenous price shocks (for example, a war in an oil-exporting country).

##### Fiscal Stimulus.

\phantomsection
[$sec:fiscal-stimulus$]{#sec:fiscal-stimulus label="sec:fiscal-stimulus"}
States can also respond to short-term downturns by propping up aggregate demand, following @Keynes1936ian demand management.
[^94]

By definition, to push up aggregate demand, government must *dissave*.
Fiscal stimulus must be paid out of newly issued debt, or prior savings --- but it cannot be paid out of increased *current* period tax revenues, for such hikes would cut private demand, as public demand expands.
[^95]
Still, even *deficit spending* will *crowd out* some private demand, as it soaks up savings in capital markets.
Government bonds must be sold to someone, and that someone will not be able to spend or invest that same money elsewhere.

Luckily for government (and all of us), it can target its spending wisely to those areas where it will have a maximum *multiplier effect*, that is, where demand begets as much more demand as possible.
For example, when government procures a new railroad line, these construction workers will have more resources to spend on entertainment, or invest in a family home.

Of course, government demand is not the only demand that multiplies;
private demand also multiplies.
However, because government (ideally) aims *not* to maximize profit, but overall economic recovery, it can seek out that spending that maximizes multiplication.
Private demand, by contrast, will seek maximum profit at low risk, which, in economic downturns, is often found in a "flight to safety", hoarding of cash or equivalents.

A good stabilizer should go entirely into immediate consumption or investment, and possibly incentivize further dissaving.
[^96]

While easier to understand for the often maligned deficit spending type of stimulus, governments must trade off the multiplication effect versus the crowding out effect during *all* fiscal expansions.
Even when (or rather, hypothetically, *if*) governments were to pay for stimulus "out of pocket", these now dispersed savings were collected at an earlier time, ever since they crowded out that same amount in private demand.

Maximizing multiplied demand, for every increment of demand crowded out is a challenge that government must always face in fiscal expansion.
Where to best direct resources is an empirical question for econometricians, statistical physicists and other experts that I happily need not, and cannot engage here.

*Automatic stabilizers*
can support demand without legislative intervention.
They include unemployment insurance
[^97]
and short-time working benefits, both of which smooth out (consumer) demand by substituting the market income of laid-off, or shorted workers.
[^98]

*Discretionary stabilizers*
are often called for when bubbles, panics or amplified shocks cause output fluctuations too large and fast to be captured by automatic stabilizers.
They come in the form of tax breaks, stimulus subsidies or public investment.
They, too, rely on dissaving, and the incurred debt must be repaid in good times.

*Side effects.*
As all potent drugs, fiscal stimulus must be prescribed with great care.
It has at least two dangerous side effects:

1.  *Structural painkiller.*
All stabilizers can disguise real needs for adjustment as cyclical problems, further diverting an economy from its equilibrium and long-term growth path.

2.  *Playing favors.*
Especially discretionary stabilizers are also prone to clientelism, that easily creeps in when legislatures pass often very targeted stimulus packages.
When specific decisions are made, rather than general rules passed [@Weber-1918-aa], as discretionary stabilizers require, special interests are more likely to gain the upper hand.
[^99]

##### Monetary Stimulus.

[$sec:monetary-stimulus$]{#sec:monetary-stimulus label="sec:monetary-stimulus"}
Government can also prop up the economy in the short run by monetary stimulus.
[^100].
Government injects more cash into the economy, to make up for liquidity frozen up during the downturn, thereby (p. )[^101]

The monetary contraction in the course of an endogenous downturn can best be described as a general loss of confidence in the profitability of future projects (such as investments).
As market participants become squeamish about economic prospects in general, they hoard cash (or equivalents) or flee to safety (such as gold, or --- before 2009 --- government bonds).
[^102]
Less funding is available even for the most profitable and unrisky projects and the lack of liquidity can render (balance sheet) solvent firms (cash flow) insolvent.
The money supply contracts.
When these defaults further feed the market panic and cause more debts to sour, a self-reinforcing debt-deflation crises may ensue.

A monetary stimulus bolsters market confidence by essentially placing an optimistic, massively government-backed bet.
When central banks resort to quantitative or qualitative easing, they declare whichever assets they buy or collateral they accept as worthwhile (profitable) investments by fiat.
When, if and to the extent that markets buy into this stunt of optimism, they will be increasingly willing to lend the injected (and their own) money again, too.

Monetary stimulus has some drawbacks, too.
Monitoring the money supply in the economy is very difficult and imprecise, and related policy imperatives are incompletely understood.
To avoid inflation, central banks must contract their money supply again as soon as the monetary stimulus has worked, and market liquidity has recovered.
The timing and calibration of interventions to maintain a constant money supply are crucial, but difficult to get right.

Monetary stimulus, too, can disguise and defer unavoidable realignment when the cause of the downturn is exogenous, such as a true price shock (as for example when revised Greek fiscal numbers became available in 2008).
When the long-term growth path is shifted (downward), the money supply *must* contract in absolute terms as the underlying economy, too, will contract, or at least grow slower than previously expected.
Monetary stimulus does not allow economies to live beyond their real, exogenous means for long;
it just defers the pain to a later contraction or risks inflation, which in turn (p. .

#### Long-Term Inconsistency {#sec:long-term-inconsistency}

##### Market problems

The market institution of equilibrium interest rates may fail to guide humans to temporal consistency in the long run for three reasons.

1.  (Exogenous) growth theory discussed so far assumes technology to be constant, or exogenous:
as if the rate at which new technology is discovered cannot be altered by humans.
Accordingly, once a maximum steady-state output is achieved, further *capital deepening* or other policy will be inconsequential for growth.
Exogenous technological innovation appears unreasonable in a knowledge-based economy (as recently endorsed in @Communities2009).
When we *can* endogenously improve our creativity, we might have to save more to achieve intertemporal Kaldor-Hicks improvements.

Additionally, markets may fail to adequately save for and invest in fundamental [rnd]{acronym-label="rnd" acronym-form="singular+short"}, because innovation is often a (p. ).
Market innovation may also be hampered by under asymmetric information (p. ).

2.  Ancillary conditions frequently observed in post-industrial economies imply *real* (public or private) and require offsetting, greater *nominal* savings (p. ).
They include aging,
[^103]
structural unemployment,
[^104]
or, equivalently, (p. ), underprovided public goods (basic research?) and overused common goods (@Stern-2006-aa:
climate change!).
Markets can, by definition, not provide these goods or solve these problems, and to the extent that government (or lovers) have neglected to do so in the past, these failures constitutes real dissavings.

3.  Market return rates on capital set a saving rate by equilibrating *self-interest*.
Savings rates, however, are a public good, which cannot be provided out of self-interested exchange.
One person's saving bestows a positive externality on other people in two ways:

1.  When technological change is held to be *exogenous*, other people will benefit from capital deepening and approximating of steady-state growth.

2.  When technological change is allowed to be endogenous, these benefits include greater economy-wide productivity and innovation and will be even larger.

Markets, when left to their own devices, will save less than optimal because not all of the societal benefit can be recouped in interest payments.

##### Government Saving.

\phantomsection
[$sec:government-saves$]{#sec:government-saves label="sec:government-saves"}
If the polity has democratically prescribed a savings rate and/or the market fails to save at positively optimal levels, government can prop up saving by regulatory and fiscal means.

In *regulatory* policy, government can mandate citizens to save into a pension or other saving scheme.
[^105]
Mandatory pensions and similar financial products solve time inconsistency both at the individual level (present and future self) and the societal level (present and future generations).

The difference between fiscal and regulatory policy I draw in (p.  is a bit murky and overdrawn, when it comes to saving.
Here is my attempt to draw it anyway:

Regulatory

:   I treat policy as regulatory, when it ties made contributions to *entitled benefits* later in life.
Much like when it is fencing the commons, government guarantees legally enforceable, quasi-property in both private and public pension schemes.

Such a regulatory pension scheme, tying benefits to contributions, organises intertemporal production (saving) and distribution (dissaving) in a *single* institution --- much like the market economy does.
[^106]

Fiscal

:   I treat a policy as fiscal, when contributions do *not* entitle payers to specific benefits.
By definition, this is what taxes do.

In fiscal policy, government can encourage private saving with clever subsidies or save resources itself by raising more taxes than spending in the current period.
Government can invest budget surpluses in publicly run projects such as education or infrastructure, or privately run enterprises in [swfs]{acronym-label="swf" acronym-form="plural+short"} (here again, the public-private is irrelevant to the saving component (p. ).

Such a fiscal intertemporal regime keeps intertemporal production (saving) and distribution (dissaving) apart.
Each step in the intertemporal deal requires a separate political decision.
For example, government will have to decide when to auction off [swfs]{acronym-label="swf" acronym-form="plural+short"} and what to use the revenue for.

Convergence Over Space: Growing the Pie Smoothly {#sec:space}
------------------------------------------------

#### The Human Condition of Space

Humans have evolved in small, intimately known groups of hunters and gatherers of no more than a few hundred individuals, covering a relatively small area of Earth in their lifetime (popularized, reviewed in @Diamond1997).
In evolutionary terms, mass society and fast travel are recent developments.
And so we may be ever disposed to parochially define our context in terms of locale (and ethny)
[^107]
(for example, @Van-den-Berghe-1981-aa).

When we let our parochial sentiments make policy, economically costly autarky results.
We are rich and prosperous, because at the national, regional and global level we are integrated by organic solidarity of *functional* differentiation, not the mechanical kind of locale and ethny [@Durkheim-1893-aa].
Under modernity, autarky is always regression.

The cosmopolitan mindset transcends our parochial heritage, and defines our group and place in the most encompassing, radically modern way:
species homo sapiens on planet Earth.
This is the jurisdiction of human rights, the scope of science and the marketplace of economic liberalisation.

However, modernity and prosperity have not arrived nor progressed simultaneously everywhere, but the have swollen very differently (p. ).
Our world, today, is a very unequal place.

#### Market Solutions

\phantomsection
[$sec:trade$]{#sec:trade label="sec:trade"}
Markets offer the institution of trade to mediate between our parochial tendencies and the cosmopolitan demands of our modern world.
Trade kindles our self-interest, binds us to strangers far away, increases our common welfare and, sometimes, makes us more equal, too.

In the following I (very) briefly describe four mainstream trade theories and highlight their welfare and distributive effects, but will have to ignore much real world complexity and empirical refinement (for another review with empirical evidence, see @Beckfield2009).

I here apply the paradigms of trade theory to commerce *within* an idealized, closed economy, even though trade is traditionally understood as *inter-state* exchanges between two or more open economies.
However, some of the same basic dynamics also govern welfare and distribution *inside* one closed economy.
Of course, those institutions that differ from one open economy to another --- including monetary policy, trade barriers and factor mobility --- do not apply to trade within the closed economy:
there is only one money, no tariffs or quotas, and great (if not perfect) capital and labor mobility.
But the overall dynamics of divergence and convergence are always the same.
In fact, thinking, as we do, *differently* about commerce *within* a country, and trade *between* countries is an exercise in cognitive compartmentalization (Gabbard 2010), and it often borders on mercantilism or reeks off nationalist sentiment.

\phantomsection
[$sec:trade-theories$]{#sec:trade-theories label="sec:trade-theories"}
Trade is always and everywhere the voluntary exchange between people who differ in location, ability or something else of economic value.
Trade is, in short, the mode of a market economy.
We know of at least four different theories to explain growth, convergence and divergence under trade:

:   [$itm:absolute-advantage$]{#itm:absolute-advantage label="itm:absolute-advantage"}
Trade on absolute advantage occurs between at least two parties that can produce different goods at the cheapest cost [@Smith-1776-lq].
[^108]
The parties will both specialize in whichever good they can produce more efficiently, and trade the surplus production (over their domestic demand) for goods other parties have an absolute advantage in producing.
All parties benefit from greater overall productivity.
Aggregate output increases:
everyone gets richer.

Crucially, when a party has no absolute advantage in producing any good, it will not trade at all.

The benefits are divided according to productivity.
In @Smith-1776-lqian trade, everybody gains, if not equally.
It exploits exogenous productivity differences, but does not otherwise lead to convergence in productivities or prosperity.

:   [$itm:comparative-advantage$]{#itm:comparative-advantage label="itm:comparative-advantage"}
Trade on comparative advantage occurs between parties that produce different goods at *relatively* different productivities [@Ricardo1817].
[^109]
A party will specialize in that good, for which they have to give up the least other production (opportunity costs).
The party then trades the surplus production (over their domestic demand) for goods other parties have a comparative advantage in producing.

A party will trade, even if it has no absolute advantage in producing any good.

The benefits of trade are divided according to the terms of trade.
In Ricardian trade, too, everybody gains, but not equally.
It exploits exogenous productivity differences, but does not otherwise lead to convergence in productivities or prosperity.

Factor Price Equalization

:   [$itm:FPE$]{#itm:FPE label="itm:FPE"}
According to the [fpe]{acronym-label="fpe" acronym-form="singular+short"} theorem, parties trade with parties that differ in their factor endowments (such as labor and capital) [@Stolper1941].
[^110]
A party will specialize in producing goods intensive in their more abundant factor and trade the surplus production (over their domestic demand) for goods other parties have specialized in.

Overall output increases, as factors are put to the most productive use.

Within the party, the relative factor returns change as trade commences.
The relatively more abundant factor (in rich countries, capital) is in higher demand and reaps a higher return.
The relatively less abundant factor (in poor countries, low-skilled labor) is in higher demand and reaps a higher return.
Over the long run, according to [@Stolper1941] specialization continues until all factors are equally abundant in all parties, and command equal returns (prices).

Between the parties, according to Hekscher-Ohlin trade, everybody gains, but not equally.
It equalizes factor prices, but does not further convergence of endowments or prosperity.
Immediately, specializing according to factor endowments can reinforce exogenous, pre-existing inequality:
as a (human) capital-poor party opens to trade, the return on capital (education)
[^111]
may fall.

Economies of Scale

:   [$itm:NTT$]{#itm:NTT label="itm:NTT"}
According to *[ntt]{acronym-label="ntt" acronym-form="singular+short"}*, parties trade with one another not to exploit any exogenous, pre-existing difference in productivities or endowments, but because specialization itself pays [@Krugman-1980-aa].
[^112]

A party will specialize in the goods in which it is already specialized, or at the least, in which no other party is specialized to produce at competitive prices.
Similarly, a group of geographically (or otherwise) clustered parties will specialize in one category of goods (food in north-western France) or one industrial sector (automobile in southern Germany), to benefit from specific resources (specialized labor) and networks (for example, supply chain, trade fair).

In [ntt]{acronym-label="ntt" acronym-form="singular+short"}, overall output increases, as more production happens at higher scale and dense networks are efficiently shared.
[ntt]{acronym-label="ntt" acronym-form="singular+short"} also implies significant distributive effects:
whoever specializes first and is closely clustered, wins.
This *agglomeration* may increase spatial inequality and counteract convergence.
In extreme cases, economies of scale and network effects may breed monopoly or oligopoly producers, causing additional distributive effects and welfare losses.

##### Balance of payments

In the short and medium term parties can defer all of the distributive effects of trade by running *current account deficits*, including trade deficits.
Current account deficits are offset by *capital account surpluses*, including sales of domestic assets and issued debt.
In the long run current account deficits (and counterparty surpluses) built up untenable imbalances and can trigger *balance of payments crises*.
At the end of the day, current account deficits cannot persist but must be (p. ).
Running current account deficits may defer the distributive pain (at a cost), but cannot avoid it.

In the final analysis, trade deficits, surpluses (and capital account deficits, surpluses respectively) are meaningful only net of trade with *all* other parties.
If party A has a a trade deficit with party B, which has a trade deficit with party C, which has a trade deficit with party A, overall current accounts may be balanced.
Only the net deficit trade distilled into capital account surpluses (debt or foreign ownership) matters for economic imbalances.

This can be explained with reference to personal finance.
All people except for suppliers and employees run a "trade deficit" with their local supermarket;
yet, as long as they have offsetting surpluses with other parties (for example, their clients), they may be financially sound.
Only if they spend more than they earn (or own), will they be in trouble.

\phantomsection
[$sec:adjustment-costs$]{#sec:adjustment-costs label="sec:adjustment-costs"}
All of the above four trade models assume costless adjustment of the economy.
As some sectors wax and others wane in the course of specialization, factors of production (capital, labor) are transferred from one use to another with no cost.
This is an unrealistic assumption.
Specialized capital (machinery) or (trained) labor may not be useful in another industry without some retooling or retraining, if at all.

To the extent that adjustment costs are concentrated in one party or industry, as they are likely to be, adjustment also has "domestic" *distributive* effects.
People who have a stake in the industry favored by trade, such as trained workers and owners win.
Workers and owners in declining industries lose.

##### Factor Mobility

\phantomsection
[$sec:factor-mobility-trade$]{#sec:factor-mobility-trade label="sec:factor-mobility-trade"}
Modeling the spatial dimension of a closed market economy on international trade theories may appear as a bit of a stretch.

The analogy works only to the extent that factors of production stick to place and industry.
By definition, labor and especially capital face no *formal*, spatial boundaries *within* the closed economy.
In a perfectly mobile labor and capital market, none of the above distributive effects would apply:
factories and workers would fluidly and without cost move to wherever they can earn most, counteracting any spatial factor rent differentials.

In the real world, neither capital nor labor is perfectly mobile, and to that extent, trade theory applies.
[^113]

Domestic and international trade are subject to essentially the same economic dynamics.
They differ not dichotomously, but gradually, depending on factor mobility.

If I have stretched the concept of international trade here, it is to show that the closed, mixed economy experiences the same welfare and distributive dynamics, and has, as I explain in the following section, found policy responses to mitigate them.

#### Government Solutions

Government can forge spatial convergence to counteract divergent trade dynamics by fiscal, and to a lesser extent, by regulatory means.

##### Regulatory Policy

According to (neo)classical economic doctrine, to allow everyone to partake in and gain from trade government should "fight factor market rigidities" (a euphemism for letting wages fall and consumer prices rise).

Two welfare state institutions are frequently blamed for wage rigidities:

1.  Formal minimum wages or income-substitution benefits can establish effective price floors in labor markets and cause a [dwl]{acronym-label="dwl" acronym-form="singular+short"} of unemployment, or, in the terms of trade theory, a group or region of people with no absolute advantage.
[^114]

2.  Powerful trade unions and corporatist style collective bargaining can make wages (especially downward) rigid.

To further trade, or, equivalently, clear labor markets, government should avoid price floors and, if found downwardly rigid, tamper or bust trade unions.
[^115]
Both these measures will have adverse (vertical) distributive consequences, which government may counteract with fiscal transfers.
Again, as I argued earlier, to distribute vertically (p. ).
Equity can best be achieved by fiscal measures.

In international trade, government can protect "infant industry" to remedy the distributive dynamics of Hekscher-Ohlin trade and to check agglomeration by setting import tariffs or quotas.
Under the shield of protection, the economies accumulate capital (or high-skilled labor) until it becomes the more abundant factor and join international trade to participate (in arguably more value-adding) capital-intensive production.
Alternatively, protection allows domestic firms or sectors to grow to a scale and density that allows them to reap returns to scale and network effects before they enter international competition.
By definition, there are no tariffs or quotas *within* the closed economy, so government cannot *protect* infant industries.

##### Fiscal Policy

Government can nurse infant industries, by pursuing industrial or structural policy.
[^116]
In industrial policy, government forges capital deepening, builds forward-looking infrastructure and favors "picked winners" in sectors and firms.
Similarly, structural policy supports underperforming regions or sectors by granting subsidies or tax breaks.

The toolset of industrial and structural policy is varied and complex, often centered on fiscal means, but also including a banking regime, ownership structure, corporate governance and institutions in education and training (for an impressive survey of different configurations of these interlinked institutions, see @HallSoskice-2001-aa).

Government intervention in the economy is also rightfully controversial.
Instead of nurturing infant, but promising industry, it can protect sclerotic, inefficient sectors doomed for "creative destruction" [@SchumpeterSwedberg-1942-aa].
In the worst case, industrial and structural policy succumbs to clientelism and turns into crony capitalism.
Famously, *picking winners*, in industries and regions, is very hard and government may be particularly bad at it.
Both the targeting and timing of government support for regions, industries and firms are tricky.

Subsidies must be *targeted* at industries that later can successfully compete in an open market.
Government often has limited knowledge or ability to make these calls, and clientelist politics easily creeps in.

Subsidies must also be *timed* to reliably recede, to put sufficient pressure on industry to become competitive and to avoid windfalls.
Phase out a subsidy too soon, and the industry dies.
End it too late, or never, and you invite rent-seekers.

As subsidies, tax breaks or funding for local government, central government can compensate for adjustment costs, give side-payments to losers from trade and generally smooth structural change.
Progressive transfers can also dampen self-reinforcing agglomeration and counteract suppressed factor returns.

##### Monetary Policy

By definition, a closed economy only has one currency, and thereby, one monetary policy.
Consequently, monetary policy cannot affect the spatial dimension of economic activity *within* that closed economy.

## The Means of a Mixed Economy

> *"The revenue of the state *is* the state."*\
> --- Edmund @Burke1790 [111, emphasis added]

To pursue the (p. ) of (p. ), (p. ), (p. ), (p. ) and (p. ), the mixed economy relies on an intact set of (p. ), (p. ) and (p. ) means.

Effectively commanding regulation, taxation and fiat money is not trivial.
To be effective, these institutions of the mixed economy must be designed to *anticipate* and *minimize* adverse interactions with the independent market economy.
As is the defining feature of the mixed economy, (p. ), both in its policy *ends* and *means*.

The Means of Regulatory Policy {#sec:regulatory}
------------------------------

Effective regulatory policy is (p. ) but (p. ), and always (p. ) with the reach of economic activity.

Capability.

:   [$itm:capability$]{#itm:capability label="itm:capability"}
To regulate economic activity, government must command an effective monopoly on the use of force and possess the administrative capability to pass, monitor and enforce regulation.

Regulation will often be exceedingly complex and require frequent modification and government must be equipped to keep up with the creativity and dynamism of private business.
[^117]

Restraint.

:   [$itm:restraint$]{#itm:restraint label="itm:restraint"}
On the other hand, the regulatory power of government must also be strictly limited by the rule of law, particularly the right to property, and the norms of good governance.
Markets must be protected from arbitrary or unpredictable regulation of the economy that would disrupt their smooth functioning.

Restrained government maximizes planning reliability for market participants, protects their confidence and avoids retroactive effects.
[^118]

Congruence.

:   [$itm:congruence$]{#itm:congruence label="itm:congruence"}
Lastly, the scope of regulation must match that of the economic activity in question.
[^119]

When regulation covers an area smaller than the respective market, *regulatory arbitrage* ensues.
For example, when health and safety in manufacturing are regulated at the county level, factories will relocate to wherever the rules are laxest.
Because manufactured goods can easily be moved, consumers in the strictly regulated county will buy (cheaper) goods produced under lax conditions.
Regulation at the county level will be ineffective.
When local jurisdiction have an incentive to attract manufacturing such as employment or tax revenue, standards may equilibrate at a the (low) Nash equilibrium in a [pd]{acronym-label="pd" acronym-form="singular+short"}-type game.
When jurisdiction and the mobility of factors and goods does not match, a *race to the bottom* may ensue.
[^120]
Conversely, when the scope of regulation far exceeds that of markets, government may overreach and violate norms of *subsidiarity*.
For example, when health and safety of hairdressers are regulated at the national level, hypothetical regional or local differences in customer preferences would be negated for no compelling reason.
People will usually not travel to get a haircut, and haircuts cannot be transported.
Regulatory arbitrage is hence unlikely to occur.
Local government is free to set training standards at whichever level its electorate sees fit, without fear of hairdressers relocating or customers fleeing to other, laxer jurisdictions.
[^121]
As @Bordo2011 put it, *this* is the "*raison d'être*" [-@Bordo2011 4] for fiscal federalism, as defined by [@Oates1972].

In the closed economy, government can always regulate at the largest possible scope of markets, the national level.
By definition, no relocation or trade beyond the closed economy are possible.
Central government can devolve jurisdiction to local government according to market scope, as it sees fit.

The Means Fiscal Policy {#sec:fiscal}
-----------------------

To redistribute market outcomes, fund public goods and natural monopolies or to change market incentives, government relies on the treasury.
These fiscal activities in (p. ) differ by and , summarized in (p. ):

Base.

:   \phantomsection
[$itm:base$]{#itm:base label="itm:base"}
Fiscal revenue generation differs in its *nominal* base, or *what is taxed*.
[^122]

Schedule.

:   \phantomsection
[$itm:schedule$]{#itm:schedule label="itm:schedule"}
Obligatory transfers differ in their desired schedule.
They can apply to everyone at a *flat* rate, *proportional*, or *progressive*, charging the fortunate above a linear tariff on their ability to pay (or , p. ).

Differing in base and schedule, fiscal activities fall in three broad categories:

\phantomsection
[$sec:levies$]{#sec:levies label="sec:levies"}
fund (p. ), including suboptimal saving.
[^123]

In english, they are also known as Pigouvian *taxes*, which is imprecise, because they are no proper taxes.
In german, they are known as literally "steering taxes" (*Lenkungssteuern*), which is apt, because they are meant to change market activity.

Pigouvian levies are *specific* in base:
only whoever uses common good pays.

Pigouvian levies --- ideally --- are scheduled at *marginal cost*:
everyone pays for the increment by which she has exhausted the common good.
Equivalently, everyone pays according to the marginal social cost they inflict on everyone else.

<!-- %  \subparagraph{\hyperref[sec:common-good]{Common Goods} and \hyperref[sec:natural-monopoly]{Natural Monopolies}} should both be financed out of specific levies.
%For common goods, these take the form of Pigouvian levies, pricing in the costs of using the non-excludable, but rival commons.
%For natural monopolies, these take the form of fees, ideally at average (not marginal) costs of production collected by the state or a publicy-owned firm.

%As the rate of the levy is determined only by the costs of producing the respective good or service, a schedule does not apply. -->

\phantomsection
[$sec:fees$]{#sec:fees label="sec:fees"}
fund (p. ).

Fees are *specific* in base:
only whoever benefits from a natural monopoly pays.

Fees are scheduled at *average* cost:
everyone pays the same share of the overall cost, no matter how much she added in incremental cost (see footnote [$fn:why-ac-fees$](#fn:why-ac-fees){reference-type="ref" reference="fn:why-ac-fees"}, p. ).

\phantomsection
[$sec:taxes$]{#sec:taxes label="sec:taxes"}
fund all remaining government outlays, including those for .

Taxes, no matter their use, are *general* in base:
everyone pays unconditionally, and without individualized return.
[^124]

Taxes can be scheduled to achieve *any* desired redistribution, even though most would understand redistribution to imply at least a proportional or progressive schedule.

<!-- %  \subparagraph{\hyperref[sec:fiscal-redistribution]{Redistributive} taxes} should, by definition, be general in incidence and proportional or progressive on ability to pay, depending on the chosen equity norm.
%A \emph{specific} redistributive tax is a contradiction in terms:
%redistributing only between some entities in turn constitutes a redistribution between the included and the exempted entities.
%Moreover, specific inclusion or exemption of entities based on something other than the accepted equity norm (for example, ability to pay) conflicts with liberal-democratic norms of non-discrimination. -->

Taxes meant to raise general revenue to finance all remaining outlays (see , p. ) imply only a flat, or per-capita schedule.
Funding, for example,

,
[^125]

,
[^126]

\hyperref[sec:fiscal-stimulus]{stimulus}
[^127]
or

\hyperref[sec:government-saves]{government saving}
imply flat schedules:
these programs are all about maximizing welfare, have universal coverage and justify no redistribution.
However,
[$sec:fiscal-redistribution-and-revenue-are-one$]{#sec:fiscal-redistribution-and-revenue-are-one label="sec:fiscal-redistribution-and-revenue-are-one"}
because redistributive taxes for equity and general-revenue taxes for welfare share the *same base*, they can also be rolled up in one combined schedule.
In addition to cutting redundant administration, taxing only according to one schedule makes more sense.
When two taxes have overlapping objectives, such as a general base, their parallel implementation may conflict.
This is also the case for general revenue and redistributive taxes, simply because a flat financing of, for example, public goods, already embodies *one* --- possibly contested --- equity norm.

#### A Norm of Ordoliberal Hygiene.

\phantomsection
[$sec:ordoliberal-hygiene$]{#sec:ordoliberal-hygiene label="sec:ordoliberal-hygiene"}
These three categories of obligatory payments to the treasury --- Pigouvian levies, fees and taxes --- are different public policy beasts entirely, and must not be confused.

<!-- % \cite{Dwyer2009} 340:
%Indeed, Leicht and Fitzgerald (2006) argue that the middle class has been ‘lent what it should have been paid’ and that a middle-class lifestyle was maintained during a time of stagnant income growth for many at the cost of financial security. -->

<!-- %\paragraph{Redistributive Taxes and General Revenue are One} The remaining fiscal instruments with the goals to redistribute and raise general revenue are treated as one.
%There is no meaningful distinction between taxes that are meant to redistribute (or its progressive element) and those (or that component), which is meant to finance public/common goods or collectivize risk.
%
%A separate treatment of the merely redistributive component of taxation and its revenue generating element is unnecessary because the population concerned, is, in theses cases, everyone (by definition), and hence same.
%
%We can then implement the degree of redistribution that is required over the entire range of general revenue generating taxes.

%\paragraph{A Norm of Ordoliberal Hygiene.} This typology of fiscally relevant state interventions into the market carries implications for the efficient design of a tax.

%General and specific fiscal interventions follow, in fact, two \emph{opposite} logics.

%A Pigouvian tax or fee is meant to fall on those very market interactions they are raised to discourage or finance.
%Pigouvian taxes \emph{should} change relative costs in the market vis-a-vis the status quo to reduce otherwise overconsumption of commons.
%An efficient local sewage system fee \emph{will} fall only on actual private or business use of the system, and not on other activity, say, a sewage-free server farm.

%General taxes to redistribute, pool risks or finance public goods, by contrast, should, to \hyperref[sec:DWL]{minimize their deadweight loss} (desideratum \ref{des:minimal-DWL}) on otherwise Pareto-optimal markets leave relative prices unchanged.

%From this follows a norm of ordoliberal hygiene to keep these two types of fiscally relevant state interventions into the market apart.
%Because I am not concerned with Pigouvian levies and fees here, I exclude them from the following in desideratum \ref{des:ordoliberal-hygiene}. -->

[$des:ordoliberal-hygiene$]{#des:ordoliberal-hygiene label="des:ordoliberal-hygiene"}
A desirable tax should not include any Pigouvian or fee-like components.

Neither Pigouvian levies nor fees should be designed to redistribute or raise *any* general revenue.
Conversely a redistributive or general revenue tax should not alter behavior.

General and specific fiscal interventions follow, in fact, two *opposing* logics:
A Pigouvian levy or fee is meant to fall on those very market interactions they are raised to discourage or finance.
Pigouvian levies *should* change relative costs in the market vis-a-vis the status quo to reduce otherwise overconsumption of commons.
General revenue or redistributive taxes, by contrast, should leave relative costs and levels of market activity unaffected, otherwise, a [dwl]{acronym-label="dwl" acronym-form="singular+short"} ensues.
An efficient local sewage system fee *will* fall only on actual private or business use of the system, and not on other activity, say, a sewage-free server farm.
[^128]

Of course, Pigouvian levies and fees sometimes cause undesired distributive effects.
For example, taxes on fuel consumption often hit lower and middle class people hardest.
Still, government should not alter a Pigouvian schedule, let alone issue offsetting subsidies, such as the German commuter tax relief ("Pendlerpauschale").
Instead, government should use general-base redistribution for equity.

#### Redistribution and Revenue-Generation are One.

\phantomsection
[$sec:fiscal-redistributionAndRevenueAreOne$]{#sec:fiscal-redistributionAndRevenueAreOne label="sec:fiscal-redistributionAndRevenueAreOne"}
On the other hand, this typology of fiscally relevant state interventions also suggests that redistributive and revenue-generating taxes for and provision of can be consolidated into one.

The desired general incidence in both these classes of taxes implies that the universe of taxable entities is the same.

Redistributive and general revenue taxes *should* be rolled into one for two reasons.
economicus

:   The collection of any tax will be administratively costly.
Implementing fewer taxes will be cheaper.
The flat or proportional schedule of revenue generation can be easily included in a redistributive (proportional or progressive) schedule as a base rate.

General Revenue and Redistribution Interact.

:   When two taxes have overlapping objectives, such as a general base, their parallel implementation may conflict.
This is also the case for general revenue and redistributive taxes, simply because a flat or proportional financing of pooled risks and public goods already embodies *one* --- possibly contested --- equity norm.

A flat or proportional financing of general revenue, could, for instance, conflict with efficiency desideratum des:low-price-floor in .
Conversely, *levels* of risk pooling and public good provision are shown to bear equity consequences in .

A desirable tax exclusively finances general revenue and redistributes through a single schedule.
[$des:redistribution-and-revenue-are-one$]{#des:redistribution-and-revenue-are-one label="des:redistribution-and-revenue-are-one"}

#### Congruence.

In addition to a well-functioning tax administration, tax collection also requires congruence between the economic activity being taxed and jurisdiction of taxation.
In @Oates1972' seminal formulation:

> *"The result of tax competition may well be a tendency toward less than efficient levels of output of local services.
> In an attempt to keep taxes lows to attract business investment, local officials may hold spending below those levels for which marginal benefits equal marginal cost, particularly for those programs that do not offer direct benefits to local business."*\
> --- @Oates1972 [143]
> [^129]

For example, when Pigouvian levies on pollution in aluminum production are set at the local level, but aluminum is widely traded, production will relocate to wherever taxes are lowest.
When factors and/or goods are sufficiently mobile, tax levels will race to the bottom.
This cooperation problem of taxation at the local level also applies to redistributive taxation where the rich will relocate to avoid progressive taxation.
Even fees for natural monopolies will be affected:
when sewage fees are set at the local level, other municipalities can attract more establishments by pricing their sewers at (initially much lower) marginal cost, rather than otherwise optimal average cost.
As more people and firms relocate to the cheaper municipality, it eventually becomes unable to provide the natural monopoly at marginal cost.

In the closed economy, tax competition is usually not a problem, as most taxes and vulnerable fees are set at the national level.
[^130]

#### (Many) More Tax Desiderata.

For general revenue and redistributive taxes, the list of desiderata goes on (see @Held2010a).
I focus here on only two criteria that are important for the welfare state and regional integration:
and (p. ).

##### Well-Determined Incidence on Natural Persons.

\phantomsection
[$sec:well-determined-incidence$]{#sec:well-determined-incidence label="sec:well-determined-incidence"}
In @Vickrey1947's seminal clarification, "genuinely progressive taxation is necessarily *personal* taxation" [-@Vickrey1947 1, emphasis added].
Albeit often ignored, this is merely a definitional clarification:
distributive norms only concern the relative utility of *natural persons*.
Corporations, for example, are no moral subjects, only people are.

Factually, corporations are also never the ultimate recipients of utility.
The welfare of all non-natural, private juristic persons (say, a public company) ultimately accrues to natural persons as their owners (say, shareholders), workers or customers.

Taxing a corporation --- as [cits]{acronym-label="cit" acronym-form="plural+short"} try to --- is as nonsensical as it is impossible.
It is nonsensical because we do not, and *cannot* know what it would *mean* to tax, say, the income of Deutsche Bank AG, because Deutsche Bank AG is an institutionalized fiction, but not a moral subject of fairness.
It is also impossible because taxing the income of Deutsche Bank AG *will* instead reduce the welfare of its owners, workers or customers (who *are* moral subjects).
Crucially, natural personhood is a condition *only* for general revenue and redistributive taxation, because they share the same *general* base (all people).
By contrast, pigouvian levies and fees *can* and *should* be levied on any merely formal market participants, including corporations, that use the respective common good or natural monopoly (a *specific* base).
While a [cit]{acronym-label="cit" acronym-form="singular+short"} on the income of Deutsche Bank AG makes no sense, the corporation should still be billed for its sewage or carbon footprint.

Not only will a nominal tax on the income of a corporation factually fall on its owners, workers and customers, but, to make matters worse, we cannot even know which of these stakeholders pays.
In the language of economics, the *effective* incidence of a tax can be difficult to ascertain, and it often matters little who nominally pays the tax as shown in (p. ).

\centering
![The Incidence of a Tax on Suppliers with Same Elasticities for Producers and Consumers[]{label="fig:same-incidence"}](same-incidence){#fig:same-incidence width="100%"}

Taxation is always an (p. ).
The incidence of a tax is borne by all parties to these market exchanges in proportion to their relative price elasticities of demand and supply, respectively, as shown in (p. ).
Whoever can cheaply exit or substitute the taxed market exchange dodges most of the tax:
because these price elastic suppliers *would* quickly cut their quantity supplied at lower prices, they can extract relatively high prices from buyers.
Buyers, in this example, cannot cheaply exit or substitute the taxed market exchange:
they *would* hardly be able to cut their quantity demanded in response to higher prices.
At the post-tax equilibrium, buyers foot most of the tax.

\centering
![The Incidence of a Tax on Suppliers with Relatively Less Elastic Demand[]{label="fig:different-incidence"}](different-incidence){#fig:different-incidence width="100%"}

In the real world, relative price elasticities of demand and supply in many markets are unknown and would be difficult to ascertain.
[^131]
Elasticity of demand and supply depends on many factors, including the availability of substitutes (for example,, margarine for butter) as well as the mobility and specificity of factors (for example, immobile, specific typewriter factory vs mobile, unspecific car rental).

Therefore, to fairly and effectively redistribute between natural persons, government must tax market interactions where the relative price elasticities of demand and supply can be easily known.
This is often the case for very broad categories of market exchanges, such as labor ([payroll]{acronym-label="payroll" acronym-form="singular+short"}) or consumption ([vat]{acronym-label="vat" acronym-form="singular+short"}):
[^132]
people cannot substitute working *somewhere* or employing *someone*, they cannot avoid to buying or selling *something*.
[^133]
The price elasticities for demand and supply in these broadly defined markets are small and similar.

In the closed economy, government can effectively tax broad categories of market exchanges, such as consumption.
[^134]
By definition, economic activity does not extent beyond the borders of the closed economy, and there can therefore be no substitution to, for example, spending at home.

##### Minimal Market Distortions, Welfare Losses.

\phantomsection
[$sec:minimal-DWL$]{#sec:minimal-DWL label="sec:minimal-DWL"}
When government taxes a market exchange, buyers and sellers may react in ways that reduce overall welfare.

When the post-tax price of the transaction is higher than the buyers utility or the sellers cost, they exit the market.
Their otherwise pareto-improving exchange does not occur, as shown in (p. ).
Government also gains no revenue from these non-occurring transactions.
Everybody loses.
The sum total of these losses is known as the [dwl]{acronym-label="dwl" acronym-form="singular+short"} of taxation (areas C, E in figure [)](fig:DWL).
An efficient tax minimizes [dwl]{acronym-label="dwl" acronym-form="singular+short"} for maximum revenue (areas B, D).

\centering
![The [dwl]{acronym-label="dwl" acronym-form="singular+short"} of a Tax with Unit-Elastic Supply and Demand[]{label="fig:DWL"}](dwl.pdf){#fig:DWL width="100%"}

This ratio of tax revenue to [dwl]{acronym-label="dwl" acronym-form="singular+short"} --- much like the (p. ) --- depends on the price elasticities of supply and demand.
The more inelastic supply and demand, the smaller the [dwl]{acronym-label="dwl" acronym-form="singular+short"} for any given tax rate, as shown in (p. ).
If buyers and sellers cannot easily cease their welfare-improving exchanges, they will continue to trade when taxed.
A tax on such a market will not distort or depress market activities.

\centering
![The Deadweight-Loss of a Tax with Inelastic Supply and Demand[]{label="fig:smaller-DWL"}](smaller-DWL){#fig:smaller-DWL width="100%"}

The price elasticities of demand and supply are, again, a hard, empirical question.
As discussed above, supply and demand in broad categories of market exchanges, such as all domestic consumption ([vat]{acronym-label="vat" acronym-form="singular+short"}) or all domestic labor ([payroll]{acronym-label="payroll" acronym-form="singular+short"}) are relatively inelastic, and the [dwl]{acronym-label="dwl" acronym-form="singular+short"} of such taxes is likely to be small in a closed economy.

Still, [dwls]{acronym-label="dwl" acronym-form="plural+short"} of taxation may abound, even in closed economies.
In mixed economies, the market for low-wage labor is particularly vulnerable to large [dwl]{acronym-label="dwl" acronym-form="singular+short"}s.
Welfare state governments to people who earn too little to maintain a socially acceptable standard of living (p. ).

In this scenario, taxation of low-wage labor has two related, negative effects.
[^135]

1.  Taxation raises the effective cost of living for low-wage earners:
for each euro they earn on the market, they can afford a lower, real standard of living.

2.  If people can apply for handouts, their supply of labor may become perfectly price elastic, when market incomes are close to welfare incomes.
If they could earn less on the market, than by collecting welfare, low-wage workers may be forced to exit the labor market:
a large [dwl]{acronym-label="dwl" acronym-form="singular+short"} ensues.
[^136]
Any tax on low-wage labor will only increase the [dwl]{acronym-label="dwl" acronym-form="singular+short"}:
for every euro of wages taxed, more people will be pushed outside the (official) labor market.
[^137]

In the closed economy, government can minimize the tax burden on low-wage labor by relying more on progressive taxes, including a [pit]{acronym-label="pit" acronym-form="singular+short"}, instead of proportional taxes ([vat]{acronym-label="vat" acronym-form="singular+short"}, [payroll]{acronym-label="payroll" acronym-form="singular+short"}).

The Means of Monetary Policy {#sec:monetary}
----------------------------

Government provides the economy with legal tender to serve as

a medium of exchange,

a unit of account and

a store of value.

Only government can do this, because supplying money is a (p. ).
To avoid the (p. ) and (p. ) of de- or inflation,
governments should be devoted to price stability.

They set that quantity of money where it equilibrates with variable money *demand* at stable prices.
The demand for money, in turn, is determined by the output of the economy.
[^138]

#### Tools.

Government extends or contracts the supply of money to meet a quantity, interest rate (US) or inflation rate (Bundesbank) target.
Fiat money is supplied in one of two ways:
[^139]

Government creates *base money* by

buying government bonds ([omo]{acronym-label="omo" acronym-form="singular+short"}) or

other financial assets ([qe]{acronym-label="qe" acronym-form="singular+short"}) from private holders who receive legal tender out of thin air
[^140]
in return.
[^141]
Central banks can also

lend money to financial institutions at a set rate (discount window rate).

Government stimulates the multiplication of *broad money* in the fractional reserve banking system by lowering the reserve requirements for private banks.

\centering
![The Policy Trilemma of Central Banks[]{label="fig:triangle-cb"}](triangle-cb){#fig:triangle-cb width="1\linewidth"}

#### Role.

The impact of monetary policy on the real economy is, of course, incompletely understood and highly controversial.
I do not need to recapitulate these controversies here (for a recent review, see @Wapshott2011).
I note two hopefully uncontroversial points:

1.  *Monetary Neutrality.*
[$itm:monetary-neutrality$]{#itm:monetary-neutrality label="itm:monetary-neutrality"}
Following the classical dichotomy, *nominal* economic variables, such as the price level, do not matter in the long run.
The prosperity of an economy is determined by its level of technology, human capital and physical capital --- not by the amount of intrinsically worthless currency in circulation.

On this ultimate end of monetary policy, Monetarists and Keynesians can probably agree:
good money stays out of way of the real economy to let actual output track the long-term growth path of the economy.

2.  *An Empirical Question.*
[$itm:empirical-macroeconomics$]{#itm:empirical-macroeconomics label="itm:empirical-macroeconomics"}

Keynesians

:   believe that market participants are swayed by changes in nominal variables.
As a result, the prices for goods and services and factor prices (especially wages) do not equilibrate promptly or perfectly.
Self-reinforcing (debt-deflation) crises of excess supply and depressed demand may result [@Fisher1933].
The state can smooth out resulting aggregate fluctuations in output by altering the money supply.

Monetarists

:   believe that market participants are fairly rational and will consider expectations of future inflation (and taxes) in their current decisions.
As a result, prices for goods and services and factor prices, especially wages, equilibrate quickly and keep the economy on its long-term growth trajectory.
State intervention in the business cycle or the money supply are unnecessary and ineffective.
[^142]
[^143]

Determining our irrational *animal spirits* [@Keynes1936] is an empirical question best left to cognitive psychology, behavioral economics and related fields (recently @Akerlof2010, see , p. )
"Keynes vs. Hayek" [@Wapshott2011] need not be a normative struggle, but a pragmatic balance based on empirical evidence.
Whichever policy stabilizes economic output along the long-term growth trajectory is best.

If ever there was a policy that should be judged on consequentialist terms, it is monetary policy.

Because monetary dynamics *do not* and *should not* matter to our 'household with a cast of billions' I can refer the issue to empirical clarification and need not entertain it further here.
Suffice it to remind us that government should set that quantity of money where it equilibrates with variable money demand at stable prices.
The rest is empirical details.

## Trade-Offs of the Mixed Economy

Given its conflicting (p. ) the mixed economy can arrange its (p. ) to make at least two basic trade-offs.

1.  *Command vs. Exchange.*
It can organize more or less production and distribution by command instead of exchange.

2.  *Consumption vs. Saving.*
It can defer more or less of current consumption to the future.

We can look at these trade-offs by (p. ) and by (p. ) of the entire economy.

#### By Expenditure.

\phantomsection
[$sec:by-expenditure$]{#sec:by-expenditure label="sec:by-expenditure"}
A mixed economy can devote its resources (expenses) to public or private, investment and consumption goods.
The four expenditure components and revenue flows are plotted on a two-dimensional coordinate space of the mixed economy in (p. ).
I provide (very roughly) equivalent macroeconomic variables as in (p. ).
[^144]

\centering
![Coordinate Space of the Mixed Economy[]{label="fig:coordinate-space"}](coordinate-space){#fig:coordinate-space width="1\linewidth"}

\centering
##### Save/Consume More or Less.

Government increases the private savings rate by taxing consumption (instead of income), encouraging investment (factory) through industrial policy or other fiscal stimulus and increases public saving by putting net tax revenue into durable public goods (basic research), natural monopolies (a bridge) or .

Steering in the opposite direction, government increases private consumption by cash transfers, fiscal stimulus and public consumption by handing out in-kind benefits (school milk), or providing short-term public goods (fireworks) and natural monopolies.
The saving-consumption trade-off roughly expressed in the gross savings rate in (p. ).
[^145]

##### More or Less Government/Market

Government sets the exchange-command mix of the economy by taxing more or less of economic output, and commissioning public investment and consumption from the revenue.
The exchange-command mix is (roughly) expressed in the public expenditure quota in (p. ).

##### How to Strike the Best Balance.

Welfare economics and the of a mixed economy (p. ) suggests that there are one (or more) *optima* in the balance between saving and consumption, between market and command.
These ends also bind each of the government interventions, such as providing a public good, to *specific* justifications,
[^146]
such as a market failure in providing the public good.
Normatively, government *should* not take an arbitrary position on ().

However, I should point out that *positively*, government in a closed economy *could* employ its to choose any of the possible trade-offs between consumption and saving, market and command economy.

##### Tax is Key.

Of these of the mixed economy, tax is the dominant tool.
To redirect resources between the four quadrants, government relies primarily on (p. ) rather than on (p. ) which has limited applications and (p. ) which is neutral in the long run.

\small
-------- -------------------------------------------------------- ------------------------------------------------------------------- -------------------------------------------------------------- --------------------------------------------------------------------- --

(r)3-4                                                                                         *Private*                                                         *Public*
*[I]{acronym-label="I" acronym-form="singular+short"}*    Factory, [rnd]{acronym-label="rnd" acronym-form="singular+short"}                      Bridge, basic research                       *$\sum=$[gfcf]{acronym-label="gfcf" acronym-form="singular+short"}*
*[C]{acronym-label="C" acronym-form="singular+short"}*                            TV set, vacation                                               School milk, fireworks                         *$\sum=$[fce]{acronym-label="fce" acronym-form="singular+short"}*
*$\sum=$($\gls{I}+\gls{C}$)*                      *$\sum=$[G]{acronym-label="G" acronym-form="singular+short"}
[^147]*
(r)3-4
-------- -------------------------------------------------------- ------------------------------------------------------------------- -------------------------------------------------------------- --------------------------------------------------------------------- --

: GDP Components by Expenditure in the Closed Economy[]{label="tab:GDP-Comp-Exp"}

#### By Changes in Net Worth

\phantomsection
[$sec:delta-net-worth$]{#sec:delta-net-worth label="sec:delta-net-worth"}
The same trade-offs of the mixed economy can also be understood as different changes in net worth.

As economies, firms and households save and consume, their net worth changes.
This truism is expressed in the Haig-Simons income identity:

$$\label{eq:haig-simons} \text{Income}=\text{Consumption}+\Delta\text{Wealth}$$

Or, trivially transformed:
$$\label{eq:haig-simons-trade-off} \text{Savings}-\text{Dissavings}=\text{Income}-\text{Consumption}$$

applies to the savings trade-off of households, firms and government, including some examples.

\centering
![Individual and Collective Haig-Simons Identity of Income[]{label="fig:haig-simons-individual-collective"}](haig-simons-individual-collective){#fig:haig-simons-individual-collective width="1\linewidth"}

The logic of the Haig-Simons identity, and is easy:
you can only have your cake *or* eat it.
Any change in consumption must be matched by a change in income or net worth --- and vice versa.

For instance, a household can afford to build a house either by taking out a mortgage, earning more, or consuming less.

The identity can be balanced *across* households, firms and government, too.
These transfers occur through different financial products and fiscal institutions.
For instance, a household can also afford to build a house if it pays fewer taxes and government accepts less revenue.
Government looses the amount in income that households earn.
[^148]

A comprehensive Haig-Simons identity also includes depreciation (for example, neglected roads) and depleted natural resources (for example, fossil carbohydrates) as real dissavings.
For instance, an economy can consume more than it earns for some time by burning oil.
[^149]
Conversely, a comprehensive Haig-Simons identity also includes real savings such as a newly developed technology or public infrastructure, even when these are not (yet) market priced.
[^150]
For instance, an economy can consume less today at equal income and channel the surplus into basic research, [rnd]{acronym-label="rnd" acronym-form="singular+short"} or send more people to (costly) college.

The Haig-Simons identity of income is a truism similar to the law of conservation of matter.
Surprisingly, it is often ignored or misconstrued, even in (p. ).

I offer two notes to further clarify:

1.  *Saving $=$ Investment.*
In the long run, when industry has adapted and monetary effects have neutralized, all savings are invested.
Saving does *not* depress aggregate demand and choke the economy.

True Keynesian shortfalls in aggregate demand result from people hording *cash* or equivalents, and not because of an increase in investment.
In deflationary spirals, anticipating lower prices, people cut *both* consumption *and* investment:
the contracted money supply freezes up *all* economic activity.

This popular conflation of monetary dynamics with savings rates may be one of the most formidable obstacles to enlightened, democratic choice of economic policies and tax in particular.
[^151]

Instead, saving *changes*, but need not depress, aggregate demand;
more capital goods and fewer consumer goods are in demand.
If given enough time, the economy can transform from "S-classes to school buildings" without write-downs on unamortized capital investment.
Well-regulated, competitive financial intermediaries will always channel saving into investment.

An investment is a valuable transformation or improved understanding of our physical world.
It requires labor.
Be it a [swf]{acronym-label="swf" acronym-form="singular+short"} or a savings account, a blast furnace or a green tech patent --- in the final analysis, saving always means to build more things that last longer and/or that we will consume later, instead of things that last a short while and that we consume now.

Of course, any *one* investment will *ultimately* be consumed or depreciate away.
And we *can* save too much, when capital goods depreciate faster and have such decreased marginal returns that they outstrip our current utility from the same resources (This follows from [@Solow1956] theory of growth, p. ).

But while that means we should not save endless amounts at any point in time, it does not mean that at some point in time, we should save no more.
There is no economic reason why we could not roll over (limited) savings to our children in perpetuity.

2.  *Dissaving $\neq$ Debt $=$ Deposits $\neq$ Saving.*
[$itm:credits-debits-wash$]{#itm:credits-debits-wash label="itm:credits-debits-wash"}
Dissavings are not the same as debt.
Dissaving is a decrease in net worth of households, firms and economies:
we diminish some durable thing in its value.
Conversely, saving is an increase in net worth:
we add value to some durable thing.

In contrast, going into debt does not affect net worth of households, firms or government:
we temporarily gain access to an *already existing* valuable thing (construction man-hours), potentially transform it into something else (a house) and return the valuable thing later (with interest).
Conversely, putting in a deposit (or other credit) also does not affect net worth:

\small
*Households*                             *Government*
-------------------------------------- ------------------------------------------- --------------
$Income - Spending<0$                    $Revenue - Spending<0$               *$\sum$*
private debt                              public debt                      $=$
(for example, credit card, mortgage)        (for example, government bonds)         *All Debt*
$20pt$                          $Revenue - Spending>0$              *$\sum=$*
$Income-Spending>0$
private credit                            public credit                     $=$
(for example, deposits, bonds)      (for example, reserves, sovereign wealth)   *All Credit*
$\sum=0$

: Debt and Credit in the Closed Economy[]{label="tab:Debt-Credit"}

Trivially, public and private debt will always equal public and private credits in the closed economy as summarized in .
By definition, every debtor needs a creditor.

Debt and credit define the short-term control of and long-term claims to valuable things and may have distributive consequences, but they do not affect net worth.
Net saving, by definition, does.
[^152]

Smoke and Mirrors of the Mixed Economy

> *"If it's too good to be true, it's too good to be true."*\
> --- The author's landlady, trained nurse, single mother of three and foreclosed homeowner in Irvine, CA (2007)

A mixed economy can seemingly overcome its physical limitations and the (p. ) between different ends using a set of smoke and mirrors.

\centering
![Circular Flow of Income in the Economy, with Macroeconomic Imbalances](circular-flow-with-imbalances.pdf){width="100%"}

\scriptsize{Compare \autoref{fig:circular-flow-with-imbalances}.}
[$fig:circular-flow-with-imbalances$]{#fig:circular-flow-with-imbalances label="fig:circular-flow-with-imbalances"}

I briefly explain how three such practices allow us to live beyond our means in the short term:

1.  *Credit Bubbles.*
[$itm:credit-bubbles$]{#itm:credit-bubbles label="itm:credit-bubbles"}
Within the closed economy, (p. ).
Still, excessive debt and credit can serve to hide or defer economic trouble ahead.

In efficient financial markets, credit is extended to households, firms and governments at an interest rate that fully reflects the risk of default.
To assess credit risk, creditors make (and update) predictions about the future solvency (and liquidity) of debtors.

In the real world, these guesses are sometimes overly optimistic given the available information and credit is extended at too low an interest rate to too many people and organizations.
[^153]
As long as these risks are not reappraised or do not materialize, an economy can seemingly live beyond its material means.
Eventually, of course, credit bubbles will burst and debtors will partly (have to) renege on their promises to repay interest and principal.
Now, the economy as a whole has to pay for the fat years.

Credit bubbles are thus always an intertemporal redistribution of wealth from the future to the presence.
[^154]
When they burst, credit bubbles also jumble ownership rights as defaults spread [@Stiglitz2010].
[^155]
Depending on this resulting, quintessential political-economic struggle between debtors and creditors, the brunt of the bursting bubble is allocated between different households, firms and government [@Coggan2011].
[^156]
Similar to inflation and asset bubbles, bursting credit bubbles also redistribute somewhat arbitrarily based on financial product and timing:
people who own equity or leave overheated credit markets early enough, win.
The suckers and laggards loose.

2.  *Asset Bubbles.*
[$itm:asset-bubbles$]{#itm:asset-bubbles label="itm:asset-bubbles"}
Concomitant to credit bubbles, bubbles can arise in certain classes of overvalued assets, often including real estate (2007ff), stock (2000f) but also art, oldtimers or essentially useless gold.
Assets are overvalued, when their capital gains are more than what returns can reasonably be expected, given all available information.

As long as prices do not revert to intrinsic value, people and the economy as a whole can live off the virtual capital gains and beyond its material means.

Asset bubbles, too, redistribute wealth from the future to the presence.
In addition, as pyramid schemes, they redistribute between early investors and later, "greater fools" (see @Stiglitz2010 for a good discussion of the 2007ff crises).

3.  *Inflationary Pressure.*
[$itm:inflationary-pressure$]{#itm:inflationary-pressure label="itm:inflationary-pressure"}
Excessive monetary expansion, aside from fueling credit and asset bubbles can also cause demand-pull inflation.
The onset of inflation and its costs, however, need not be instantaneous.
As upcoming wage-price spirals and increasing inflationary expectations silently add to built-in inflation [@Gordon1988], an economy may enjoy temporarily heightened output.
[^157]
As the price level eventually creeps up,
[^158]
the economy pays the costs of inflation through depressed growth.

Grandfathered inflation also redistributes from the future to the presence, and arbitrarily redistributes between cash-denominated and other ownership claims, between debtors and creditors.
[^159]

#### Too Good to be True.

In summary, an economy cannot produce and consume above its long-run growth path of outward shifting aggregate supply.

Trivially, the wealth of an economy is determined by its natural resources, technology, human and physical capital --- all *tangible things*.
When these capacities are fully utilized (as they are *not* in a cyclical downturn or debt-deflation crisis), no financial or monetary charlatanism can take us beyond them in the short term.
Any short-term gain that a bubble or excessive monetary expansion will bring only defers the day of this reckoning.

## Real Dissavings of the Mixed Economy {#sec:real-dissavings}

> *"Give me chastity and continence, but not yet."*\
> --- @St.AugusteofHippo397 Confessions (VIII, 7)

Our economies substantially dissave in ways that are not reflected in conventional macroeconomic data.
These *real* dissavings, or "off-budget fiscal activities" [@Bonker2006 49] include population aging,
[^160]
depleted natural resources (land, oil, water), exhausted common goods (global warming)
[^161]
or failed public goods (immunization?), to name just a few.

These processes all unambiguously degrade something of economic value, and should be recorded as consumption in our aggregate Haig-Simons accounts (as I have suggested in , p. ).

[^1]: [$fn:tilly$]{#fn:tilly label="fn:tilly"} This positive description does not imply normatively, as liberal entitlement theory would have it, "that a person is entitled to those goods acquired in uncoerced exchanges with others" [@Nozick1974; @Friedman1962 149].
Uncoerced *exchange* does not mean absence of coersion.
At the very least, markets rely on a large-scale coercive power (aka. the state) for property rights and states and markets [@Tilly-1985-aa].
Naturalizing whatever allocative results the free market produces as entitled, inalienable, private property is as ahistorical as it is uncritical.

[^2]: Because I am interested in economic abstractions, a materialist theory of the state (such as @Tilly-1985-aa) suits me better than, for example, contract theories.
Again, this does not imply positive or normative claims.

[^3]: I prefer the terms "market" and "command" economy, because the conventional *capital*ism-socialism dichotomy is misleading:
real existing Soviet (especially Stalinist) socialism also accumulated capital (for example, electrification, railroads), just in different hands.
Market exchange and planned command describe the different modes of production and distribution more precisely.

[^4]: Mixed economies also alter market outcomes to improve *efficiency* in case of market failures.

[^5]: Power resources theory [for example @Korpi2003], Marxist interpretations [for example @Offe1972] or recently, some advocates of a [big]{acronym-label="big" acronym-form="singular+short"} might disagree:
for them, the purpose of the welfare state is to change power relations.
These are reservations of the (p. ) and are discussed in (p. .

[^6]: Conversely, a socialist welfare state is an oxymoron.
Just as welfare states are defined by their coexistence with a market economy, there is no such thing as a socialist welfare state, no matter the (however corrupted) distributional goals of nominal socialisms.
In a predominantly planned, or even command economy, the state directs much of both production and distribution and therefore does not require the (welfare state) institutions governing the mixed economy, including taxation, insurance and pension funds, business cycle smoothing and dedicated social service provision.
Even where these institutions nominally existed, they never faced the (otherwise defining) condition of independent market prices.
For example, where "prices" are *set* and "profits" owned by the state, taxes become a meaningless category:
they are really just changes in the set prices [for example, @Bonker2006 23].

[^7]: A misnomer.
Even the [nhs]{acronym-label="nhs" acronym-form="singular+short"} is not fully socialized medicine, it is just a tax-financed near state monopsony on health services.

[^8]: In fact, social insurance contributions are a regressive

[^9]: Individual citizens are incentivized to mis(over)-represent their marginal utility from additional doctors, absent individually accruing costs.
This is equivalent to a [cpr]{acronym-label="cpr" acronym-form="singular+short"} problem or the cooperation problem of a [pd]{acronym-label="pd" acronym-form="singular+short"}.

[^10]: ...or socialize parts of the labor market, which would further shift the frontier, but not alleviate the problem.

[^11]: ...as the etymology of economics would suggest:
the science of managing the *oikos*, greek for household.

[^12]: [$fn:also-in-mpp$]{#fn:also-in-mpp label="fn:also-in-mpp"}
This section is based, in part, on earlier, unpublished work which I submitted to the Hertie School of Governance as my Master of Public Policy thesis in [@Held2010a].
I have since revised and expanded it.

[^13]: [$fn:human-condition$]{#fn:human-condition label="fn:human-condition"}
I choose the term "human condition", because it captures what might be (positively) inescapable in human existence, without specifying either a natural or cultural explanation.

[^14]: Some of the frailties follow the heuristics and biases of human cognition identified by the prospect theory research program [@KahnemanTversky1979; @Kahneman2011]

Prospect theory is appropriate to consider for any first-order theory on the material:
it provides a positively descriptive model of human cognition and decision making as opposed to the normatively optimal (but rarely observed) expected utility hypothesis, on which neoclassical economics rest (first formulated by @Bernoulli1738, later specified by @VonNeumannMorgenstern1944).

In part, the mixed economy and the welfare state (for example, mandatory pensions) are institutions to get prospect theory humans, easily mislead by their fast but erroneous "system one" to follow the slow, but accurate "system two" [@Kahneman2011]).
Supposedly, when we give it some thought, we find that decisions based on expected utility [@Bernoulli1738] maximize our happiness, but ever the "cognitive misers" [@FiskeTaylor-1991-aa], exhausted and overwhelmed, we easily fall back to *system one* shortcuts and fail to rationally maximize utility.
Institutions, characteristically enabling and restricting human behavior --- including those of the mixed economy --- can help prospect theory humans to approximate decisions based on expected utility.

Prospect theory, alas, is still an emerging field (together with yet dejunct evolutionary psychology and cognitive neuroscience), and any link to the mixed economy must in the meantime remain tenuous.

[^15]: Indeed, as Dr. Mildner of the Hertie School of Governance never tired of telling me:
"There is state failure, too"

[^16]: As suggested by the late [Kim-Jong Il's choice of grey jumpsuits](http://kimjongillookingatthings.tumblr.com/).

[^17]: As in modernization theory, for example, [@InglehartWelzel-2005-aa].

[^18]: Specifically, resource-based "rentier states" may hinder liberal democracy [@Beblawi1990].
Generally, distributive decisions in planned economies appear easily as zero-sum games (as opposed to the per-definition non-zero-sum, pareto improvements in competitive markets), which may corrupt politics.

[^19]: This broad definition of efficiency is more demanding than pareto efficiency.

[^20]: Here, as always, it is important to keep up (p. ).
The first theorem of welfare economics invoked here is an exercise in positive logic.
It does not imply:

1.  that *actual* markets display these properties --- that would be a first order *empirical*, not logical finding.
The assumptions of are, in fact, quite heroic and may rarely be observed in the real world.

2.  that market allocations are beyond normative reproach.
The first theorem of welfare economics, crucially, operates only on *ex ante* distributions, which may or may not be distributively just.

[^21]: (Neo)classical economics takes an oddly static view of the economy.
In reality, equilibria are more often in flux, a matter of becoming, not being.

[^22]: [$fn:1st-theorem$]{#fn:1st-theorem label="fn:1st-theorem"}
Formally, the first theorem of welfare economics states that over a *given* distribution, the competitive equilibrium will be a pareto optimum (demonstrated first graphically by [@Lerner1944], mathematically by [@Lange1934], [@Debreu1954] and others).

[^23]: An allocation is pareto-improved if at least someone receives more, with others receiving (at least) the same.
When all such improvements are exhausted, an allocation is pareto optimal.
Crucially, pareto improvements and optimality are always in reference *only* to some ex-ante allocation.
Also, pareto improvements do not imply that everyone would be better off by the same amount.

[^24]: Nassim Nicholas @Taleb2007 has recently put this succinctly by praising "aggressive trial and error" [-@Taleb2007 xxi] in free markets that "allow people to be lucky".
Being a quantitative trader by profession, @Taleb2007 abandons rational choice when faulting Karl @Marx-1867-aa *and* Adam @Smith-1776-lq for believing that free markets work because of *rewards*.

[^25]: @Friedman1970a's dictum applies:
"The Social Responsibility of Business is to Increase its Profits" [-@Friedman1970a], both as commandment and promised salvation.

[^26]: This assumption is politically consequential.
If firms and their owners, in fact, merely maximized their profits, their property would *not* bestow power.
By definition, to be a profit-maximizer in a competitive market is to *only* do things that pareto-improve everyone over a given distribution.
For property to bestow power in the classic definition [@Geoff2002 8ff], owners must sacrifice some of their profit to make other people do something they would *not* otherwise (pareto-optimally) have done.
Under microeconomic dictum, owners exert power by *transferring* or *relinquishing* some of their surplus production.

Clearly, to assume profit maximization and thereby to effectively define away a power aspect of property is implausible, and it rests on some heroic assumptions of human rationality and utility-maximization.
More importantly, profit maximization implies --- as all such static microeconomics --- that future profits, and thereby, future demand *can* be known, at least probabilistically.
From a more dynamic or evolutionary perspective, capitalism is *path dependent*, and those rich enough to bet on future profits now will make the very paths on which that future depends.

Property *does* bestow power, and the kind of axiomatical contortions to define that away probably constitute the kind of hegemonic tendencies of microeconomic thinking that so (rightly) infuriates its critics.
However, any emancipatory or equity argument on the distribution of property should explicate its relaxation of profit maximization, and provide some preliminary account of the motivations of rich people, instead.

I take up this argument when I discuss the $fairness of taxation$ (p. ), especially of a .

[^27]: Constant returns to scale contrast with the very idea of functional differentiation, economic modernization and division of labor discussed (p. ) and (p. ).

[^28]: I discuss only a popular subset of market failures in this paper.
I ignore many of the problems recently highlighted by the 2007ff financial and sovereign debt crises.
While common and public good failures explain much (for example herding and information spill-overs in @Banerjee-1992-aa), the 2007ff crises require a dedicated first-order theory of financial capitalism, including an in-depth appreciation of its institutions (for example, derivatives), practices (for example, bank equity) and policies (for example, interest rate).
[@Cassidy2010] provides an equally accessible and comprehensive account of "How $financial$ Markets Fail".

[^29]: Providing a public good has a positive externality, using a common good has a negative externality.
Underproviding a public good and overusing a common good can also be represented as [pds]{acronym-label="pd" acronym-form="plural+short"}, where underprovision or overuse is the Nash Equilibrium.

Whoever first buys a natural monopoly at very high marginal cost, extends the positive externality of very low marginal costs to all subsequent buyers.

[^30]: Elinor [@Ostrom1990] criticizes the canonically assumed failure of commons in social science and provides an empirically grounded account of their successful, non-coercive governing.

[^31]: Public and common goods are often hard to distinguish in public choice.
*Not* overusing a common good is a public good, and providing a public good is a common good.

[^32]: This relaxes the perfect competition assumption [$itm:easy-entry-exit$](#itm:easy-entry-exit){reference-type="ref" reference="itm:easy-entry-exit"} on .

[^33]: Otherwise, if *all* buyers could pay for the initial costs equally.

[^34]: Calls for more startups, patents and research spin-offs, particularly in Germany, may serve as evidence for suboptimal incentive design under the status quo.

[^35]: [$fn:monetary-commons$]{#fn:monetary-commons label="fn:monetary-commons"}
There are no monetary responses to any of these market failures, because (p. ) boosts or retards economic activity *overall* by extending or restricting credit and money creation.
Markets fail in the provision of public goods, common goods and natural monopolies because these goods are ill-priced *relative* to other goods in the economy.
States cannot deliberatively correct relative prices in the economy by monetary policy.

Inflation, if and to the extent that it is caused by monetary overexpansion, may, however appear first in some goods, including long-term assets such as real estate, as appears to have happened in the US monetary expansion up until the 2007ff financial crisis.
These relative price changes on the first impact of inflation cannot be effectively targeted by government and eventually propagate into an overall increase in price levels.

[^36]: The expression stems from Ireland, where the British king passed enclosure legislation to privatize and built fences around previously communal pastures.

[^37]: The Coase theorem further holds that it does not matter to whom the property rights are initially granted, say whether the polluters are granted a right to pollute, or citizens are granted a right to clean air (invariance thesis).
While *welfare* neutral, this original granting of property rights does have distributive effects:
whoever is granted the property right, enjoys a windfall.
Economists often suggest to auction off new property rights at competitive prices, to minimize arbitrary distributive effects.

[^38]: Here's what happens upon privatization:
Enclosed public goods become natural monopolies.
Enclosed common goods become private goods, as per (p. .

[^39]: States can also regulate common goods by outlawing its overuse, as the US does by instituting minimum [cafe]{acronym-label="cafe" acronym-form="singular+short"} standards on automakers.
Regulation by defining maximum acceptable use of commons is inefficient, because it does not incentivize market participants to save the commons wherever it is marginally cheapest to do so.
Standards regulation cause a [dwl]{acronym-label="dwl" acronym-form="singular+short"} much like minimum wages or other price floors and ceilings.

[^40]: The German term for a Pigouvian tax, helpfully, translates to "steering tax" (Lenkungssteuer).

[^41]: [$fn:why-ac-fees$]{#fn:why-ac-fees label="fn:why-ac-fees"}
Recall that prohibitevely high initial marginal cost marred private markets in the first place.

[^42]: Except for (p. ).

[^43]: For two technical reasons, most economists and central banks prefer a positive, but moderate and stable rate of inflation of around 2% rather than perfect price stability at 0% inflation:

1.  A positive inflation rate gives central banks room for maneuvre in fiscal stimulus.
At moderate inflation, central banks can pursue a *negative* real interest rate (as was the case in 2011ff in many markets).

2.  A positive inflation rate can mitigate the hypothesized downward stickiness of nominal labor costs.
Real wages will fall even when nominal wages stay constant.

[^44]: The cost of minimizing cash holdings, including many trips to the bank, during which shoe leather is supposedly worn down.

[^45]: The costs of repricing for business, including the cost of changing restaurant menus.

[^46]: Experiencing higher consumer prices, workers will demand higher wages, which in turn prompts producers to increase sale prices.

[^47]: With the exception of Japan's lost decade.

[^48]: Under deflation, holding (increasingly valuable) cash is relatively more attractive than investing.

[^49]: This, if nothing else, is the reason why backing a paper currency with specie (such as the Gold standard) or pegging it to another currency (fixed exchange rate) are bad ideas:
under such fixed or pegged regimes, the money supply expands and contracts *independent* of output.
Under the gold standard, money supply tracks discovery and extraction of gold --- a process which is likely unrelated to economic output.
This, in addition to the transaction costly lunacy of digging up gold in one place (a mine) only to bury it in another place (a vault).

Similarly, under a pegged (or fixed) exchange rate, money supply follows the monetary command of *another* economy.

[^50]: I concentrate here on *individual* risk.
Governments and markets should also reduce *aggregate* risk and systemic uncertainty [@Knight1921].
They should avoid extreme negative risks, seek extreme positive risks [@Taleb2007] and avoid systems too complex and tightly coupled to be safely operated by humans [@Perrow-1999-aa].
To the extent that such precaution, requires fiscal, regulatory or monetary policy intervention, this end also depends on intact of the mixed economy (p. ).

[^51]: If people were neutral between upside and downside risks, they would be indifferent between a lifetime of car insurance premiums and a (low) probability of accidental personal bankruptcy.

[^52]: With sophisticated financial markets, insurants buy this protection not just from one another, but also from other (rich) people or financial intermediaries willing to stomach the risk for a premium.

[^53]: I do not include old age here, because the risk component (a long life in retirement) of pensions is small compared to the saving component.
Other insurances, notably health insurance, also include a saving component (for high morbidity in old age), but the risk component dominates.

[^54]: This relaxes perfect competition assumption [$itm:perfect-information$](#itm:perfect-information){reference-type="ref" reference="itm:perfect-information"} on .

[^55]: The conventional terminology stems from @Akerlof-1970-aa's original paper, in which he modelled quality uncertainty in the market for used cars.
In American slang, a lemon is a car "that is found to be defective only after it has been bought" (Wikipedia).
A cherry, conversely, is a good used car.

[^56]: Prior the Obama administration's [ppaca]{acronym-label="ppaca" acronym-form="singular+short"}, in the US, *pre-existing conditions* were frequently exempted from private health insurance contracts, defeating the purpose of insurance for many chronically ill or disabled persons.
Private insurers minimize the ex-ante problem of asymmetric information by excluding those risks about which the insurants may already know.

In Germany, insurants experience a similar frustration if they seek to take out (recently privatized) occupational disablement insurance, facing greatly limited coverage (for example, no cardio-vascular conditions), eligibility (for example, no back-related conditions for construction workers) or premiums (for example, higher rates for burnout-prone teachers).

[^57]: To the extent that government also regulates coverage and admittance, as it would have to do to avoid a lemons markets, the mandatory, nominally private insurance becomes a partly quasi-fiscal institution.
When central qualities of the products (insurance) are regulated and buyers forced to buy, premiums are to a large extent pre-determined.
Insurance firms will differ only in the few, administrative components of their business outside the reach of state command.
The part of the premium that covers the mandated services is effectively a tax, and the insurance company but an outsourced service provider to the government.
Ignoring, as I have here done, the (vexingly complicated) supply side of medical care, the difference between mandatory private and public health insurance appears small.

[^58]: Even people with a privately known risk of zero should be included, as low-risk insurants may otherwise find it attractive to misrepresent their privately known risks.
This extreme case of a privately known risk of (close to) zero cannot be justified under an efficiency norm of Pareto optimality, as everyone is made better off by making some people worse off.
Instead, it requires the stronger Kaldor-Hicks efficiency norm [@Kaldor1939; @Hicks1939].
Resolving a lemons market may cost some people less than it will benefit others.

<!-- %  The costs of risk pooling should incur generally to all entities, irrespective of their individually, asymmetrically known risk\footnote{Moral hazard, the ex-post problem of asymmetric information, is ignored \citep{Arrow1971}.
%It is assumed here and in the following that the insured, privately known risks do not depend on the behavior of entities.
%If risky behavior does in fact occur and create moral hazard, disincentivizing of such risky behavior may enhance efficiency.
%\emph{Specific} sources of revenue such as co-payments or variable premiums should then be extracted from some, risk-seeking entities only.
%Such a specific revenue component of moral hazard is, in fact, a Pigouvian taxation of a commons problem.
%It is further discussed below.}.
%Even entities with a privately known risk of zero should be included, as low-risk entities may otherwise find it attractive to misrepresent their privately known risks\footnote{This extreme case of a privately known risk of (close to) zero cannot be justified under an efficiency norm of \hyperref[sec:pareto]{Pareto optimality}, as everyone is made better off by making some entities worse off.
%Instead, it requires the stronger \hyperref[sec:kaldor-hicks]{Kaldor-Hicks efficiency} norm described in \autoref{sec:tax-optimality} \citep{Kaldor1939,Hicks1939}.
%Resolving a lemons market will cost some people less than it will benefit others.}. -->

[^59]: Because (p. ).

[^60]: In perfect markets with sufficient price flexibility it also does not matter whether social contributions are levied from employers or employees.
The incidence of a tax depends only on the relative price elasticity of supply and demand for labor.
In labor markets, employees (supply) are typically less price elastic than employers (demand) who can substitute labor for capital, or move their capital.
Employees end up paying for "social insurance" no matter the nominal burden.
Only the matters (p. ).

[^61]: Again, states and markets are not the only way to organise production and distribution of material goods, emphatically not in the mainstays of moral hazard, such as decisions about medical care.
[@Schwartz2010] argue passionately for a patient-doctor relationship based on trust and "Practical Wisdom" and cogently argue how *any* system based on incentives can, and *will* be perverted.

[^62]: According to [@Hobbes-1651-aa] "life in a state of nature" would be "solitary, poor, nasty, brutish and short".

[^63]: Whether a capacity for altruism really developed out of genetic nepotism, as both the book of genesis and the inclusive fitness hypothesis (@Hamilton1964 [@Wilson1975]) would have it has recently been questioned [@Wilson2012].

[^64]: This, again, is the first theorem of welfare economics (see footnote [$fn:1st-theorem$](#fn:1st-theorem){reference-type="ref" reference="fn:1st-theorem"}).

[^65]: This, of course, is the very exploitation that Marx has criticized (Manchester) capitalism for ([-@MarxEngels-1848-aa; -@Marx-1867-aa]).
By definition, *proletarians* --- those who only have their offspring --- only receive the minimum wage necessary to reproduce themselves, and their labor.

[^66]: *Price discrimination* strategies are one often unacknowledged practice that redistributes utility from the rich to the poor in advanced market economies.
To tap into different willingnesses to pay, firms often try to price same or similar goods to different market segments.
Especially when the goods are same (toothpaste with or without coupon rebate), or similar in production cost (mid-priced sedan and luxury sedan), price discrimination can redistribute resources from (rich) consumers with a high willingness to pay to (poor) consumers with a lower willingness to pay --- an effective, if normatively imperfect way to redistribute.

Compared to a no-trade scenario, price discrimination causes a pareto-improvement as per the (see footnote [$fn:1st-theorem$](#fn:1st-theorem){reference-type="ref" reference="fn:1st-theorem"}).
Compared to a trade scenario *without* price discrimination, however, (poor) buyers with a low willingness to pay are better off compared to (rich) buyers with a high willingness to pay.
With price discrimination, buyers of mid-priced sedans enjoy some of the innovations (airbags) first paid for by luxury car buyers.
Pareto optimality makes everyone better off, but not by the same amount:
a caveat that can have unexpected implications.
If ever there was a way that wealth would "trickle down", this must be it.

In addition to this unexpected distributive effect between consumers, much of price discrimination also constitutes monopolistic competition, and as such redistributes resources from consumers to producers and causes a [dwl]{acronym-label="dwl" acronym-form="singular+short"}.

[^67]: Consequently, some part of real observed structural unemployment may stem from such heightened wages, and therefore, be 'natural' [@Schlicht1978], as opposed to 'voluntary', as the Monetarists would have it.
Keynesians may then conclude that an expanded money supply should continuously push against such natural, but suboptimal unemployment.

[^68]: By contrast, in perfect capital markets one investor with \$10 should be roughly replaceable by two investors with \$5 each, ignoring transaction costs, information asymmetries, risk aversion and other complications.
High-skill labor markets may display invisibilities:
one computer scientist with a PhD in artificial intelligence may not be replaceable by two (or more) computer scientists with an undergraduate degree in the same area.

[^69]: This relaxes assumption [$itm:constant-returns-to-scale$](#itm:constant-returns-to-scale){reference-type="ref" reference="itm:constant-returns-to-scale"} on .

[^70]: Such flexibility does justice to [@Baumol1965] original insight, which started out as an empirical observation on the relative pay of the performing arts, and not as an(other) *iron* law of wages (cf. @Malthus1798).

[^71]: These additional, scarce opportunities may be awarded to individuals (or firms, or regions) based on easy but imperfect measures (for example, standardized test scores).
They may also be awarded based on probabilistic predictions on future performance (think past awards), further increasing a self-reinforcing dynamic.
In the worst, most inequitable (and inefficient case), they are awarded based on meaningless, randomly occurring differences (for example, mental state on day of testing), haphazard selections (for example, first come, first serve) or systematic measurement bias (for example, habitus expectations by assessors).

[^72]: This relaxes assumption [$itm:easy-entry-exit$](#itm:easy-entry-exit){reference-type="ref" reference="itm:easy-entry-exit"} on .

[^73]: Such as job opportunities on professional networks [@Benkler2006], the marks of social distinction received at the dinner table (@Bourdieu-1984-aa, recently @Hartmann2002), academic peer citations [@Jackson1968; @Merton1988], or plain social influence [@Asch].

[^74]: This is precisely the kind of critique from which the first theorem of welfare economics is rightly immune:
it assumes *given distributions* and it --- as pareto optimality in general --- have nothing to say about the virtue or vice of these original allocations.
This is, however, the kind of critique that must be waged against any neoclassical, or otherwise account that swiftly jumps from formal logic to normative judgement to policy implication without taking due note of these limitations.

[^75]: Marginal utility diminishes, but does not turn negative.
Consequently, the rich will still enjoy greater utility, if only slightly more.

[^76]: The canonical version can be found in (p. .

[^77]: When positional consumption is a pure [pd]{acronym-label="pd" acronym-form="singular+short"} cooperation problem, it becomes not just inequitable, but inefficient, too.
When defection (buying BMWs) is widespread, curbing positional consumption is even .
Economist [Robert H. Frank](http://www.robert-h-frank.com) has long ([-@Frank1987]) argued provocatively that positional taxation may benefit the rich and recently testified for a [pct]{acronym-label="pct" acronym-form="singular+short"} in front of the US [House Financial Services Committee](http://financialservices.house.gov/) on May 16, 2007.

[^78]: This violates the assumption [$itm:infinite-buyers-sellers$](#itm:infinite-buyers-sellers){reference-type="ref" reference="itm:infinite-buyers-sellers"} of .

[^79]: Historically, at least, public companies (the Dutch East India Company in 1602) and banks (Venice-based Medici bank in 1397) predated capitalism, and to this day, [ifis]{acronym-label="ifi" acronym-form="plural+short"} urge nascent "capitalisms" to legislate corporate law and build financial intermediaries.

[^80]: Rent control is famously derided as "the next best way to destroy a city" aside from carpet bombing.

[^81]: In addition to the mere complexity of such regulatory intervention, price controls are also prone to rent-seeking clients.
As the gains (and losses) will often be obvious to, and concentrated in some citizens, they may lobby their government to affect changes [@Peltzman1976; @Posner1975; @Krueger1974].
As in oil-rich rentier states [@Beblawi1990], widespread zero-sum games may corrupt politics.

[^82]: Such as in Article 9, Section 3 of the German Basic Law.

[^83]: A cash-flow based, post-paid [pct]{acronym-label="pct" acronym-form="singular+short"} (recently @McCaffery2002 [@McCaffery2005]), a [lvt]{acronym-label="lvt" acronym-form="singular+short"} [@George1879], and, if necessary, a [nit]{acronym-label="nit" acronym-form="singular+short"} to fight structural unemployment or working poverty (first suggested by Milton @Friedman1962), and a [wt]{acronym-label="wt" acronym-form="singular+short"} on net worth to rein in extreme inequity are to respond to these inequities (p. ).

Notably, even such sharply progressive taxation, may be unable to mitigate the inequities (and welfare losses) of monopsony employer labor markets.
Potentially, monopsony employers will counteract progressive schedules by further lowering their wages.
Some regulatory interventions, especially a right to strike and to unionize, may be essential.

[^84]: Rich people are more likely to be creditors, and poor people more likely to be debtors.
Telling, as [@Coggan2011] does, the history of mankind as the struggle between debtors and creditors [paraphrasing @Marx-1867-aa] may be roughly adequate and enlightening, but the simplification carries only so far.
As [@Coggan2011 K6-24-04] himself points out, German hyperinflation wiped out the (often nominally denominated) savings of the middle class, but left relatively unaffected the (often real denominated) wealth of the upper class and the subsisting lower class with neither debt nor savings.

[^85]: There is a case to be made for deliberately sharing the burden of systemic and unsustainably high levels of debt, as may be case with Greek sovereign debt in the 2007ff financial crisis [@Coggan2011].
As @Coggan2011 reminds us, we do not need inflation to achieve that;
a (partial) default will do the trick, colloquially referred to as a "haircut" in the 2007ff financial crisis.

[^86]: [$fn:3components$]{#fn:3components label="fn:3components"}
In his treatise on the economic costs of global warming, @Stern-2006-aa lists three components to discount to future utility:

future people may value an additional unit of consumption less (or more) if they have more (or less) output available overall (elasticity of marginal utility).

expected but *exogenous* growth, such as a chance discovery of a new technology, may increase future output without requiring present people to give up anything.

people will discount future utility simply because it is in the future and uncertain (the pure discount rate of the future) [@Stern-2006-aa 52].

[^87]: Idiosyncratic risk of specific investments are diversified or hedged away.
Risk-free interest rates still include a premium for *systematic* (not systemic) or aggregate risk, such as the continued existence of life on earth.
The remaining interest component, the pure discounting of future utility, represents the hedonic loss of present, myopic individuals.

[^88]: Specifically, saving (below @Solow1956's optimal rate) without interest is a @Kaldor1939-@Hicks1939 improvement:
future utility is greater than present cost.
Interest can be considered an intertemporal sidepayment from the future to the present.
With interest, sub-optimal level saving can be pareto improved:
upon saving more, both present and future generations (and selves) are at least equally well off.

[^89]: Myopia implies a *shift* in the supply curve of saving.
For any given interest rate, people will supply less.

[^90]: *Endogenous* business cycle durations range from 3--5 years for lagged inventory decisions [@Kitchin1923], 7--11 for fixed investment [@Juglar1862], 15--25 for infrastructural investment [@Kuznets1930] and 45-60 for technological revolutions [@Kondratiev1925].

For a recent empirical test of all four business cycle theories, see [@Korotayev2010].

The latter two of these are probably beyond reliable anticipation and certainly beyond the short-term.
They are also contested as empirical artifacts [@Howrey1968] or restricted to heterodox, evolutionary economics [@Modelski2010].

[^91]: [rbct]{acronym-label="rbct" acronym-form="singular+short"} proponents consequently argue against state interventions to counteract these exogenous shocks.

[^92]: For a fully-fledged account, involving the debt, equity and asset market as well as monetary and regulatory correlates of *Maniacs, Panics and Crashes*, see [@KindlebergerAliber-2005-aa].
Herding has also been complemented and expanded into the (heterodox) debt-deflation theory of economic cycles, where credit cycles magnify initial bubbles, panics or shocks [@Fisher1933].

[^93]: Generous lay-off protection has been hypothesized to contribute to "eurosclerosis".

[^94]: According to the opposing, now minority view of [ret]{acronym-label="ret" acronym-form="singular+short"}, fiscal expansion will be ineffective, because market participants will correctly anticipate that current deficit spending (expansion) will be offset by future tax hikes (contraction).
Anticipating such future losses, they will hoard more cash and equivalents today, than they would otherwise, thereby negating any current period effect.

Even though controlled experiments are not readily available at the level of entire economies, deficit spending seems to work, at least a little.
Here, for once, human ignorance (of [ret]{acronym-label="ret" acronym-form="singular+short"}) is not only bliss, but also a blessing for all of us.

[^95]: A democratic polity can public for private demand as it wishes (p. ), but any such balance will not affect aggregate (public *and* private) demand.
Democracies can call this shot, but it will not alleviate slumps.

[^96]: The 2009 German *cash-for-clunkers* scheme --- for all its infamous waste, cannibalizing effects and blatant clientelism --- included such an incentive:

[^97]: To the extent that it serves as an automatic stabilizer and/or pools the near-universal risk of unemployment, "unemployment insurance" is, again, a misnomer.
The demand stabilization bought by unemployment benefits are a public good, idiosyncratically financed out of a specific tax in many countries.

[^98]: (German) automatic stabilizers have recently faired well during the 2007ff financial crisis and received great praise by [ifis]{acronym-label="ifi" acronym-form="plural+short"} (@IMF-2008-ab [20], @WorldBank2008 [19]) and experts [@BofingerFranz-2007-aa 8].

[^99]: The 2009 [vat]{acronym-label="vat" acronym-form="singular+short"} break handed out to the hospitality sector by the new liberal-conservative coalition in Germany is a recent, brazen example.

[^100]: Technically, central banks can lower the (federal funds) interest rate by buying (back) government bonds (*[omo]{acronym-label="omo" acronym-form="singular+short"}*), pump cash in the economy by buying assets () or accept riskier collateral from commercial banks (*qualitative easing* [@Buiter2008]).

[^101]: "Monetary expansion" is thereby a misnomer.

[^102]: This is the (Post-)Keynesian consensus.
Strict monetarists argue that the money supply *always* remains at equilibrium without government intervention.

[^103]: The second demographic transition delivered low, often below-replacement level [tfr]{acronym-label="tfr" acronym-form="singular+short"} and low mortality (@Davis1945, restated by @Caldwell-1976-aa).

US 2009 estimated [tfr]{acronym-label="tfr" acronym-form="singular+short"}: 2.05 [@CIA2009], Germany 2009 estimated [tfr]{acronym-label="tfr" acronym-form="singular+short"}: 1.41 [@CIA2009], EU-25 2002 [tfr]{acronym-label="tfr" acronym-form="singular+short"}: 1.37 [@Demeny-2003-aa 2].
Life expectancy at birth for EU-25 is 69 years for males and 78 years for females [@Demeny-2003-aa 2]

[^104]: When understood endogenously, strata of underqualified workers are also a real (if not nominal) dissaving in human capital.
Saving, then, means to (publicly) educate most everyone to be competitively productive in a skill-based economy that outsources or offshores much of manual labor.

[^105]: It does not matter for the economy-wide savings rate whether citizens save into privately offered financial products or state-guaranteed (often [paygo]{acronym-label="paygo" acronym-form="singular+short"}) schemes.
I discuss the (somewhat epiphenomenal and misconstrued) later (p. ).

[^106]: Real existing (public) pension schemes are a lot murkier and less (p. .
Frequently, pension schedules feature equity or other policy goals such as child-rearing.
In these, more realistic cases, the link between contributions and benefits is distorted.

[^107]: A neologism suggested by @Van-den-Berghe-1981-aa to replace the clumsy "ethnic group".

[^108]: Formally, the trade theory of absolute advantage posits two countries, one factor of production (labor), perfect factor mobility within party, no factor mobility between parties, and constant returns to scale.

[^109]: Formally, the trade theory of comparative advantage posits two countries, one factor of production (labor), perfect factor mobility within party, no factor mobility between parties, and constant returns to scale.

[^110]: Formally, Hekscher-Ohlin trade theory posits two countries, two factors of production, perfect factor mobility within party, no capital mobility between parties, and constant returns to scale.

[^111]: An effect known as *brain drain*.

[^112]: Formally, [ntt]{acronym-label="ntt" acronym-form="singular+short"} relaxes perfect competition assumption [$itm:constant-returns-to-scale$](#itm:constant-returns-to-scale){reference-type="ref" reference="itm:constant-returns-to-scale"} of (p. ).
Specialization pays, because marginal costs of production fall (positive returns to scale).
Clustering pays, because network effects (such as in the transmission of innovation, according to @Bass1969) reward and reinforce tightly-nit networks (technically a group of nodes with some degree of heightened interconnectedness).
For a brilliant introduction to network theory, see [@Kleinberg-2009-oz].

[^113]: Such divergence between *formal* and *real* factor mobility and market flexibility is crucial not only to trade theory, but also economic policy making.
I return to this issue when I discuss the theory of (p. ).

[^114]: German reunification provides a realistic example for this scenario.
Eastern workers were less productive than their western compatriots, but soon expected to be paid equally.
Arguably, labor costs in the East and the West converged too far and too soon causing structural unemployment.

[^115]: This does not contradict a (p. ).
Trade unions --- supply cartels by another name --- are justified if and to the extent that they help to *balance* the playing field between (monopsony) employers and atomized workers.
By definition, unemployment suggests that the field has tilted too far:
there are surplus workers that cannot find a willing employer at the price imposed by the supply cartel.
Unemployment resulting from wage rigidity is the welfare loss equivalent to the reduced output of a monopoly provider.

Recognizing a level playing field in the real world will be very difficult.
Employers will always (and sometimes truthfully) argue that they would hire more workers, if only the wages were lower.
Unions will always (and sometines truthfully) argue that employers could pay more without laying off workers.

[^116]: I use the term more narrowly and do not, as others, include market failures or protection.
are adressed by government to improve (p. , not to improve convergence.
Protection, such as [isi]{acronym-label="isi" acronym-form="singular+short"} is not available within the closed economy.

By industrial policy, I here mean the discretionary and deliberative intervention in the economy by government, maybe best captured in the french term "gouvernement economique".

[^117]: The failure of bank regulation in the run-up to the 2007ff financial crisis has recently shown that government is easily ill-equipped to prevail in this "cat-and-mouse-game".

[^118]: I cannot provide here, but merely refer readers to a thorough discussion of these and related legal norms of the rule of law in market regulation.

[^119]: [@Zurn-2000-aa] speaks of a broader incongruence between the popular inputs and (economically constrained) outputs of EU-level democracy.

[^120]: For a more formal treatment of *systems competition* see [@Sinn2004].

[^121]: This is a hypothetical and admittedly implausible example.
There may be other compelling reasons to legislate health and safety, at the national level, for hairdressers, too, including fairness and simplicity.

[^122]: Nominal base, sadly, differs from the effective (p. ).

[^123]: As I explained in , the economy-wide savings rate is a commons, too.
A pigouvian levy on dissaving such as a [pct]{acronym-label="pct" acronym-form="singular+short"} internalizes the social cost of consuming over the government-set optimal savings rate [@Held2010a].

[^124]: One might think that a redistributive tax is *specific* in base, but that is a misunderstanding.
Redistribution favors some, and disfavors others, but it always affects *all* people somehow.

Indeed, a *specific* redistributive tax would be a contradiction in terms:
redistributing only between some entities in turn constitutes a redistribution between the included and the exempted entities.

Moreover, specific inclusion or exemption of entities based on something other than the accepted equity norm (for example, ability to pay) conflicts with liberal-democratic norms of non-discrimination

[^125]: Even entities with a privately known risk of zero should be included, as low-risk entities may otherwise find it attractive to misrepresent their privately known risks.
This extreme case of a privately known risk of (close to) zero cannot be justified under an efficiency norm of , as everyone is made better off by making some entities worse off.
Instead, it requires the stronger norm (p. )[@Kaldor1939; @Hicks1939].
Resolving a lemons market will cost some people less than it will benefit others.

[^126]: Even those who receive *no* utility from the public good should pay, both because their non-utilizing cannot be observed (non-exclusion) and because any additional utilization does not create costs (non-rivalry).
This case of little or no utility from public good again requires , not [@Kaldor1939; @Hicks1939].

<!-- %The costs for providing public goods should likewise incur generally to all people.
%Even those who receive \emph{no} utility from the public good should pay, both because their non-utilizing cannot be observed (non-exclusion) and because any additional utilization does not create costs (non-rivalry)\footnote{This case of little or no utility from public good again requires \hyperref[sec:kaldor-hicks]{Kaldor-Hicks}, not \hyperref[sec:pareto]{Pareto efficiency} \citep{Kaldor1939,Hicks1939}.}.

%Schedules for pooled risks should be flat, or at maximum proportional where a flat tax would create undue market distortions between entities of different size\footnote{The complication of a possibly desirable proportional schedule applies only when entities \emph{are} in fact of different size.
%This is the case for firms, where the fixed cost of a flat tax would place smaller firms at a disadvantage versus larger firms.
%I resolve this problem of different sizes later when introducing desiderata \ref{des:personal-taxation} of \hyperref[sec:PersonalTaxation]{taxing only natural persons}.
%Natural persons are, in an economic sense, entities of identical size.
%weight, impact?
%}.
%Proportional taxation, in this case, is not based on an equity consideration. -->

[^127]: Fiscal stimulus occurs only government spending *exceeds* tax revenue in bad times.
Resulting debt has to be paid back in good times with increased tax receipts.

[^128]: Confusion, alas, *is* everywhere.

For example, a [ftt]{acronym-label="ftt" acronym-form="singular+short"} was frequently touted by the left to pay back sovereign debt, or to redistribute between rich and poor.
However, a [ftt]{acronym-label="ftt" acronym-form="singular+short"}, as originally conceived by [@Tobin1970], is a *Pigouvian levy*, meant to disincentivize short-term trading.

Accordingly, it should be judged on *those* merits, and on those *only*:
can it curb the commons of "speculation", or, more precisely beauty-contest [@Keynes1936] herding [@Banerjee-1992-aa]?

Whether or not a [ftt]{acronym-label="ftt" acronym-form="singular+short"} is, in fact, able to also raise revenue, let alone *redistribute* according to some norm of fairness is dubious, at best.
Traders may be able to push the additional cost of a [ftt]{acronym-label="ftt" acronym-form="singular+short"} down their customers, including small-time bank customers.

For a Pigouvian levy, this is a good sign:
the price signal of an exhausted commons *should* perculate through the entire economy.
For a *redistributive* tax, however, such dubious incidence is arbitrary:
redistributive taxes must fall only on those intended by legislators.

[^129]: Political economists such as [@Dehejia1999] have recently added more nuance to the spectre of tax competition.
It might, for example, not quite lead to the bottom, as attracted investment also has diminishing returns [-@Dehejia1999 416], and may further favor smaller economies more than larger ones, where the ratio between attracted investment and foregone public spending is higher.

For a succinct overview of tax competition theories, also see [@Wilson1999].

[^130]: The German [lbt]{acronym-label="lbt" acronym-form="singular+short"} is a sad exception.
While it is ostensibly meant to raise general, not specific revenue for municipalities, it is set by local government.
Predictably, many municipalities find themselves forced to lower [lbt]{acronym-label="lbt" acronym-form="singular+short"} rates.

[^131]: It is usually assumed that incidence applies only to *indirect* taxes such as [vat]{acronym-label="vat" acronym-form="singular+short"} and [cit]{acronym-label="cit" acronym-form="singular+short"}.
I find this limited application of incidence implausible.
The concept remains applicable for *direct* taxes, including [pit]{acronym-label="pit" acronym-form="singular+short"} and [wt]{acronym-label="wt" acronym-form="singular+short"}, too.

A [pit]{acronym-label="pit" acronym-form="singular+short"} on labor incomes, for instance, may partially fall on employers when labor markets are very tight:
tax-depressed worker supply may cause employers to pay more.
Even a [wt]{acronym-label="wt" acronym-form="singular+short"} may fall on people other than the owners when demand for their collateral is sufficiently inelastic:
interest rates may rise when private capital becomes less abundant.

[^132]: A proportional [payroll]{acronym-label="payroll" acronym-form="singular+short"} is actually equivalent to a [vat]{acronym-label="vat" acronym-form="singular+short"} in its incidence.
A [payroll]{acronym-label="payroll" acronym-form="singular+short"} taxes (only) labor income before it is spent (prepaid), a [vat]{acronym-label="vat" acronym-form="singular+short"} taxes (only) labor income after it is spent (postpaid).

[^133]: Saving is a third alternative if you subscribe to a *[y2c]{acronym-label="y2c" acronym-form="singular+short"}*:
when you consider returns on capital a genuine *change* in welfare, you can substitute present spending for *more* future spending.

If, instead, you believe in a *[osn]{acronym-label="osn" acronym-form="singular+short"}*, returns on capital are just compensation to risk, uncertainty and delayed pleasure:
welfare is *postponed*, but not changed.
Consequently, under a [osn]{acronym-label="osn" acronym-form="singular+short"}, you can only substitute current consumption for same future consumption, and have no alternative but to spend your money at some point.

For a summary of the savings norm debate see [@Held2010a] or, in the brilliant original, [@McCaffery2005 819].

[^134]: I use consumption and, broadly equivalently, labor as examples here because unlike income taxation ([pit]{acronym-label="pit" acronym-form="singular+short"}), they do not require me to decide on a savings norm ([y2c]{acronym-label="y2c" acronym-form="singular+short"} or [osn]{acronym-label="osn" acronym-form="singular+short"}).
Both norms are unattractive in the extremes and only a [pct]{acronym-label="pct" acronym-form="singular+short"} can resolve the tension (@Held2010a).

Still, it is important to point out that, no matter the savings norm, savers in a closed economy have limited or no alternative to investing at home.
Therefore, the incidence of a [pit]{acronym-label="pit" acronym-form="singular+short"} will also be relatively well-defined in a closed economy.

[^135]: As explained in the above, both a [payroll]{acronym-label="payroll" acronym-form="singular+short"} and a [vat]{acronym-label="vat" acronym-form="singular+short"} burden low-wage labor.

[^136]: In this extreme, unlikely case, the [dwl]{acronym-label="dwl" acronym-form="singular+short"} of a welfare handout is equivalent to the [dwl]{acronym-label="dwl" acronym-form="singular+short"} of a price floor on labor at the same level.

[^137]: People leaving the labor market in this scenario has nothing to do with laziness.
The legislated minimum standard of living is after all considered minimally acceptable.
People who opt for welfare instead of poorly paid work may not have a choice.

[^138]: This very brief account follows the *quantity* theory of money.
Competing theories are discussed in the below.

[^139]: If for no other reason, fiat money would be preferable to intrinsically valuable moneys *because* governments can adjust its supply to output in the long run and at little cost.
By contrast, specie moneys are costly to extract and their supply changes arbitarily on new discovery.

The futility of a gold-based economy is aptly summarized by investor Warren Buffett:

> *"$Gold$ gets dug out of the ground in Africa, or someplace.
> Then we melt it down, dig another hole, bury it again and pay people to stand around guarding it.
> (...)
> Anyone watching from Mars would be scratching their head."*\
> --- Warren Buffett

[^140]: Actually, the balance sheet of the central bank records the acquired bond as an asset, and the currency as a liability.

[^141]: Central banks are customarily barred from buying government bonds directly from government to increase transparency.
By buying government bonds from a middlewoman, central banks can still "print money".

[^142]: They are unnecessary according to .

[^143]: They are ineffective according to the policy ineffectiveness proposition.

[^144]: [gdp]{acronym-label="gdp" acronym-form="singular+short"} and its macroeconomic components, while commonly reported, falls short of a comprehensive account of the economy.

Rather than a measure of prosperity, it indicates market value level of economic *activity*.
Roughly, [gdp]{acronym-label="gdp" acronym-form="singular+short"} is to prosperity, as a corporation's cash flow statement is to its income statement.
[gdp]{acronym-label="gdp" acronym-form="singular+short"} growth and positive cash flows are a necessary, but not a sufficient condition to prosperity.

[gdp]{acronym-label="gdp" acronym-form="singular+short"} does not measure changes in net worth as an income statement would:
for example, an earthquake (or nuclear power disaster, or both) can raise [gdp]{acronym-label="gdp" acronym-form="singular+short"} because of reconstruction *activities*, even if it actually wiped out *assets*.
[gdp]{acronym-label="gdp" acronym-form="singular+short"} also excludes environmental degredation or use of natural resources.

Lastly, [gdp]{acronym-label="gdp" acronym-form="singular+short"} falls short, because it records only activity with a market value;
economic activity in the household or family are not included, as are social costs or utility of market activities (externalities).

[^145]: GDP-related metrics fall short, again:
[gfcf]{acronym-label="gfcf" acronym-form="singular+short"} and consequently the gross savings rate does not include capital depreciation.
It also does not include real dissavings on goods without a market value, such as the population growth rate or environmental quality.

Consequently, gross savings rates may be overstated in existing accounts.

[^146]: Justifying a choice of the mixed economy does not mean that it can be reduced to a positive question.
Ends and means such as redistribution may, and maybe should, remain contested.

[^147]: [gdp]{acronym-label="gdp" acronym-form="singular+short"} also does not, as implied in distinguish between public consumption and public investment, but lumps both together in [G]{acronym-label="G" acronym-form="singular+short"}.

[^148]: Strictly speaking, this would be the case only for a perfect tax with a zero [dwl]{acronym-label="dwl" acronym-form="singular+short"}.

[^149]: Arguably, the past two centuries of growth in the West are to some extent due to the exploitation of fossil fuels.

[^150]: Preparing a Haig-Simons account on illiquid, public or intangible assets will be difficult.
As far as possible, accounting should rely on market prices, but will also have to rely on some planned [cba]{acronym-label="cba" acronym-form="singular+short"}.

[^151]: In my ongoing dissertation at [bigsss]{acronym-label="bigsss" acronym-form="singular+short"} on the pluralist and deliberative politics of taxation, I call this misunderstanding *Bastard Keynesianism*, gone badly awry.
I hypothesize that because people do not properly understand the Haig-Simons identity (and the economic cycle), they will erroneously assume that any tax on consumption will hurt the economy.
In truth, as I argue here, an economy can take on any trade-off between present and future consumption, if only it is phased in slowly enough.

This confusion, I argue, diverts people away from optimal and fair taxation and effectively curtails the democratic sovereign in pluralism.

[^152]: This distinction seems straightforward.
Yet, similar to bastard Keynesianism, much of public debate of finance and economicsis is marred by great confusion about these basic terms.
@McCaffery2005 explains how an inconsistent treatment of debt and capital gains opens up income taxation to "tax evasion 101".

[^153]: Credit and concomitant asset bubbles can arise for several reasons and according to competing theories, including beauty-contest-type [@Keynes1936], herding [@Banerjee-1992-aa] and excessive monetary expansion [@Stiglitz2010].
An authoritative history of financial crises is [@KindlebergerAliber-2005-aa].

[^154]: Conversely, credit crunches, in part, redistribute wealth from presence to the future.
By keeping the economy below its potential aggregate supply, credit crunches also waste capacity

[^155]: Conventionally, as @Stiglitz2010 points out, a default wipes out shareholders (equity finance) and makes creditors (debt finance) into the new shareholders.
In the 2007ff financial crises, as elsewhen, economic and political power often intervened to distribute the pain of default differently [@Stiglitz2010].

[^156]: @Coggan2011 recently told the history of all hitherto existing society as the struggle between debtors and creditors, to paraphrase [@MarxEngels-1848-aa].
According to both @Coggan2011 and [@Stiglitz2010], household debtors and governments assumed much of the burden during the 2007ff financial crises, causing, in part, the pursuant sovereign debt crises.

[^157]: Traditionally described as an (inward) move along the [@Phillips1958] curve.

[^158]: An upward shift of the [@Phillips1958] curve.

[^159]: In addition, the costs of later disinflation may be large.
Prescriptions for disinflation and expected costs differ.

[^160]: The second demographic transition (@Davis1945, restated by @Caldwell-1976-aa) delivered low, often below-replacement level [tfr]{acronym-label="tfr" acronym-form="singular+short"}.

[^161]: The greatest, widest-ranging market failure in the history of mankind, according to [@Stern-2006-aa].

### 3.0.3 Efficiency and Equity

Equity and efficiency are not a zero-sum trade-off. Sometimes, when the cake is sliced up more equitably, it may grow. Formulations of such positive-sum relationship between the two vary, and only some examples can be given here.

Extreme inequality in post-tax market outcomes may also conflict with the efficiency goal of full factor employment. Most (post)industrial societies have legislated socially accepted minimal incomes. This legislation takes the form of statutory minimum wages or unemployment benefits. Either way, an effective or de-facto price floor for wages is established.63

People with productivities below this price floor will not find employment, leading to a DWL of structural unemployment.64

Crucially, the level of minimal incomes depends on the broader configuration of tax incidence, indirect taxes in particular. When taxes are relatively flat and fall on labor or consumption, the real costs of living for low-income earners rise. This tax wedge can take the form of (flat or proportional, payroll-style) social contributions, VAT or a relatively flat or proportional personal income tax: either way, more money has to be made on the market for the same living standard (in Purchasing Power Parities, PPP).

The implications for tax design are twofold:

#### 3.0.3.1 Structural Unemployment as Endogenous.

The only genuine fix to structural unemployment is of course more education to lift people, or at least their children, out of their low standards of productivity. Better education, to be sure, will require fresh ideas — but also lots of public resources. Well-funded public education, in turn, is endogenous to a fiscal configuration.

#### 3.0.3.2 Structural Unemployment as Exogenous.

Even from a static perspective, assuming low-productivity workers to be exogenous, some measure of tax relief for would-be working poor may enhance welfare. An efficient tax system should reduce the burden on low-productivity earners, to keep their post-tax socially acceptable minimum wages as low as possible.

$des:low-price-floor$ A desirable tax has a minimal incidence on low-productivity earners to lower their cost of living.

### References

McCaffery, Edward J. 2002. Fair Not Flat - How to Make the Tax System Better and Simpler. Chicago, IL: University of Chicago Press.

Perrow, Charles. 1999. Normal Accidents - Living with High-Risk Technologies. Princeton, N.J.: Princeton University Press.

Merton, Robert K. 1936. “The Unanticipated Consequences of Purposive Social Action.” American Sociological Review 1.

Weber, Max. 1920. Economy and Society.

Stiglitz, Joseph E. 2002. Globalization and Its Discontents. London, UK: W. W. Norton.

Sinn, Hans-Werner. 2004. The New Systems Competition. Vol. 5. Munich, Germany: CESifo.

Scharpf, Fritz W. 1997b. “Economic Integration, Democracy and the Welfare State.” Journal of European Public Policy 4 (1): 18–36. https://doi.org/10.1080/135017697344217.

Zürn, Michael. 2000. “Democratic Governance Beyond the Nation-State - the EU and Other International Institutions.” European Journal of International Relations 6: 183–221.

Hicks, John Richard. 1937. “Mr. Keynes and the Classics - A Suggested Interpretation.” Econometrica 5.

Mills, C Wright. 1959. The Sociological Imagination.

Musgrave, Richard A. 1959. The Theory of Public Finance: A Study in Public Economy. Intern. St. New York, NY: McGraw-Hill.

Bordo, Michael D, Agnieska Markiewicz, and Lars Jonung. 2011. “A Fiscal Union for the Euro - Some Lessons from History.” Cambridge, MA.

Samuelson, Paul A. 1954. “The Pure Theory of Public Expenditure.” Review of Economics and Statistics 36 (4): 387–89.

Samuelson, Paul A., and William D. Nordhaus. 2005. Economics. 18th editi. New York, NY: McGraw-Hill Education.

Hayek, Friedrich A. 1931. Prices and Production. London: Routledge.

Lerner, Abba P. 1944. The Economics of Control. New York, NY: Macmillan.

Lange, Oscar. 1934. “Notes on the Determinateness of the Utility Function.” Review of Economic Studies.

Debreu, Gerard, and Kenneth J. Arrow. 1954. “Existence of an Equilibrium for a Competitive Economy.” Econometrica 22 (3): 265–90.

Coase, Ronald H. 1964. “The Regulated Industries - Discussuion.” American Economic Review 54 (2).

Hayek, Friedrich A. 1944. The Road to Serfdom. London: Routledge.

Friedman, Milton. 1962. Capitalism and Freedom. Chicago, IL: University of Chicago Press.

Gutmann, Amy, and Dennis F. 2004. Why Deliberative Democracy. Princeton, N.J.: Princeton University Press.

1. For this insight, I am indebted to Franziska Deutsch who first noted in 2011 that tax was an interesting case because it affected everyone, but most people knew little about it.

2. Ideational perspectives surely are important, and rationalist epistemologies not the only way to knowledge, but, as I explain in more detail in (p. ), first-order alternatives must be clarified first.

3. For this comparison, I am indebted to Maike Schulz, who — citing Eco — reminded me in 2013 that an ever elongating reading list need not be a bad thing.

4. For some social science that seems to physically placing “markets” in quotation marks, see, for example Beckert and Streeck (2012). In lieu of an economic, or social scientific explanation of failing or corrupted markets for sovereign debt, this rhetorical device works to distance the social scientist from these market messengers, as if their price signals were merely social constructs. There is, however, such a thing as objectively given, materially tangible, unsustainable debt that higher interest rates might merely communicate (Wihlborg, Willett, and Zhang 2010, 55). “Punctuation”, in any event, does not replace an explanation.

“The Single Market Review put (sic!) citizens, consumers and smes at the centre of policy-making.”
European Commission (2008, 3)

One wonders, at least, why, in addition to citizens, consumers and some (though not other) firms are also mentioned, when, in a functioning market, the latter two should be served only as proxies of the ultimate beneficiary, the citizen.

6. To mention just a few red flags, it is unclear what costs Postwar prosperity extracted from others (dependence or world systems theory), we do know whether or how Western affluence can be repeated without the same gigantuan carbon footprint and we worry whether broad-based growth in value-add is but a historical episode (cost disease).

7. For example, László Kovács (2004), then Commissioner for Taxation and Customs asked that tax rates must be allowed to differ 6-8% simply to make up for the remote location of some markets.

8. Capital inflows will be particularly effective in relatively poor economies. As these economies, supposedly, still lie far below the golden rule of saving (Solow 1956), returns on capital will be much higher than in richer economies where further (near-Solowian) capital deepening faces diminishing returns (for example, Barro and Sala-i-Martin (1995) or ibid. 1992 as cited in (Beckfield 2009, 3).

9. To remind readers that this is not, in fact, what the acquis currently stipulates: The overall structural and cohesion funds for 2007–2013 amount to no more than € 347 billion, less than 3 times the budget of the city of Berlin (€ 21 billion in 2009), or about 0.005 % of the budget of Germany (€ 1,164,000 billion in 2011). In addition, eu budget negotiations are still marred by juste retoure attitudes, with ms wanting to get paid out what they have received — the very opposite of a redistributive regime (for example, Begg et al. 2008).

10. It too, is neither new nor revolutionary: internally heterogenous mixed economies have always, with varying results, dealt with this question. In the US, much of the controversy about the role of the federal government boils down to the question of how much rich Texans should dole out to poor Louisianans. In Italy, the rift is between North and South (for example, Putnam, Leonardi, and Nanetti 1993)) and in Germany, initially between industrial and rural states, since 1990 between old and new Laender.

11. …the reach of which was only recently discussed in the 2012 US Supreme Court ruling on the Affordable Care Act.

12. For the german original, see footnote $fn:Offe-regress$. –>

13. Who knows? — Had they known about , and had they foreseen the coming wonders of electronic retail banking, they might have saved us all a lot of trouble.

14. For this insight, I am indebted to Franziska Deutsch who first noted in 2011 that tax was an interesting case because it affected everyone, but most people knew little about it.

15. Ideational perspectives surely are important, and rationalist epistemologies not the only way to knowledge, but, as I explain in more detail in (p. ), first-order alternatives must be clarified first.