design.rmd

# (PART) Methods {-}

# Research Design {#design}

Between these two 

At this intersection of ([pct]{acronym-label="pct" acronym-form="singular+short"}) taxation and (deliberative) democracy, two related research questions arise:

1.  [\[itm:resolve-better-tax\]]{#itm:resolve-better-tax label="itm:resolve-better-tax"}
    Can democratic rule resolve --- and if so, *how* --- any remaining disagreement about *a priori* desirable and doable taxation?

2.  [\[itm:prove-deliberative-democracy\]]{#itm:prove-deliberative-democracy label="itm:prove-deliberative-democracy"}
    Can deliberative democracy address --- and if so, *how* --- the vastly complex, macro-level abstractions facing our polities?

This dissertation investigates the intersecting set of these two research questions.

In one sense, deliberative democracy serves as the *method* to resolve *a priori* disagreement about taxation, and --- equivalently --- to test whether pluralist democracy may be partly to blame for the absence of better taxation.
Deliberative democracy is a good *method* for research question [\[itm:resolve-better-tax\]](#itm:resolve-better-tax){reference-type="ref" reference="itm:resolve-better-tax"} on taxation, both because I *axiomatize* it to maximize democratic legitimacy, and I *hypothesize* it to reveal some of the dysfunctions of pluralist aggregation plaguing tax choice (see below).

Conversely, taxation is a good *case* to test deliberative democracy in research question [\[itm:prove-deliberative-democracy\]](#itm:prove-deliberative-democracy){reference-type="ref" reference="itm:prove-deliberative-democracy"}, because it affects everyone, but is shrewd in the kind of macro-level abstractions which deliberative fora have so far avoided
[^1]
and which may strain the cognitive ability of deliberators.


1. Good deliberation is: Metaconsensus (or well-structured ordinal preferences)
2. Good deliberation is: Intersubjective Consensus (IC) or Intersubjective Rationality
   People who share the same *beliefs* and hold the same *values* express the same *preferences* and vice versa.

   1. As a result of deliberation, metaconsensus in *beliefs*, *values* and *preferences* will be higher. Factors will be fewer, and loadings higher.
   2. As a result of deliberation, intersubjective consensus will be higher. People who share the same beliefs about the economy, and hold the same allocative values will express the same tax preferences and vice versa.

This is *open* also to factors of fundamental disagreements about the role of rationality in democratic decisions (as compared to, say, power, practice).

   <!-- - explicates the *data*; in this case, Q-Sorts -->
   <!-- - *contextualizes* Q-methodology within existing literature, alternative formats, tradeoffs -->
   <!-- - *documents* q-sampling (of statements), condition of instruction etc. -->
   <!-- - *argues* how/why q-methodology is a fitting method for the (deliberative + tax) research question -->

<!-- add stuff from why q here -->

<!---
1. Knowledge Gain will be expressed in greater metaconsensus (and, equivalently), the lower preferences for "misunderstood" items.
2. "Tax misunderstandings" will appear as differences in factors and factor loadings before and after the deliberation.
3. Universal validity will be expressed in metaconsensus.
4. Attitude change *are not part of Q-methodology*.
5. Meta assessment (autonomy, competence and meaningful choice) *are not part of Q-methodology*.
6. Interaction effects (SES) *are not part of Q-methodology*.
7. Fundamental disagreements about the role of economics, or generally, rationality in policy choice (as opposed to, say, power, practice) will show up as factors -- if the Q sample of statements is good enough.
 -->

 <!-- 501:
 1. Choices (preferences regarding course of action) area *function* of
 2. beliefs
 3. desires (values)

 - (502f) Consensus
     + Normative consensus (on values)
     + Epistemic consensus (on beliefs; actions and cause and effects)
     + Preference consensus (what should be done)
 - (504f) Meta-Consensus
     + Normative meta-consensus ("shared recognition of the legitimacy of a set of values, but no agreement")
     + Epistemic meta-consensus ("agreement on the credibility of beliefs and their relevance to the question under deliberation")
     + Preference meta-consensus ("the character of choices across options" or structure of choices, ideally along a single issue dimension) --> -->


## Open Second-Order Questions

These hypotheticals flow, for the most part, from *a priori* knowledge.
They are grounded in reason, but not --- as hypotheticals can never be --- in *experience*.
I have, as much as possible, supported *a posteriori* assumptions with robust empirical findings --- but I cannot provide experiential evidence for the complete hypotheticals.

In welfare and tax, comprehensive *a posteriori* knowledge about hypotheticals is hard to come by.

Inframarginal reforms as those I suggest here strain the limits of economic models, all of which must be grounded in only *marginally* known behavioral data.
For example, we would need to know the price elasticity in the demand for luxury goods, or saving and agree on (as I have shown) irreduceably normative judgments of diminishing returns or an optimal savings rate.
Natural experiments are, of course, unavailable and the external validity of any laboratory setting will be strictly limited.
There is, in other words, not much else beside *a priori* knowledge that we might learn about welfare and tax.
My reasoning here is hopefully as widely agreeable as any comprehensive account of welfare and tax, but it will surely not go uncontested.
Absent empirical clarification, we must then resolve such disagreement in a democratic process.
Taxation, in other words, probably *cannot* be reduced to a first-order question, but it, too reverts to a second-order concern on how to democratically resolve any disagreement we might have about it.

In democracy, we have some, if limited *a posteriori* knowledge about the deliberative hypothetical I here suggest.
Deliberative democracy has been successfully implemented mostly in local settings on policies of limited scope [@Fung-2003-ac; @HendricksonTucker-2005-aa].
In their proposal for "Empowered Deliberative Democracy" @FungWright-2001-aa [17] explicitly suggest what underlies much of the current designs:

> 1.  A focus on *specific, tangible* problems
>
> 2.  Involvement of *ordinary people* affected by these problems and officials *close* to them
>
> 3.  The deliberative development of solutions to *these* problems.
>
Few attempts at deliberating regional issues of greater complexity have been made, including a 'Citizens Assembly' on electoral reform in British Columbia, Canada [confer @Ratner-2008-aa].

On the one hand, deliberative theorists as @FungWright-2001-aa [17] have criticized this limited scope of deliberative attempts, focused on the immediate, tangible and "confined to the realm of neighborhood and locale", as much out of necessity, as out of conviction [@Boggs-1997-aa 759].
The associated anti-authoritarian, anti-bureaucratic, anti-functional-differentiation impetus may, as [@Sousa-Santos-1998-aa] hopes for participatory budgeting in Porto Allegre, Brazil contribute to a "counterhegemonic globalization" by concentrating on special issues like land rights, urban infrastructure and drinking water in urban settings, but may by the same token remain "Bean n' Rice works" [@Sousa-Santos-1998-aa 479].
Faced, as our polities inevitably are, with macro-level abstractions, such cherishing of "human-scale democracy" above all else may, suspects @Boggs-1997-aa, move democracy "in a defensive and insular direction, laying bare a process of conservative retreat beneath a facile rhetoric of grassroots activism" [@Boggs-1997-aa 759].

On the other hand, political psychologists as [@Rosenberg-2002-aa] have problematized the cognitive demands deliberation poses.
Recent research in political psychology suggests, that --- contrary to the bounded rationality assumption --- imperfect human reasoning may not only stem from remediable cognitive scarcity, but may be developmentally determined.
@Rosenberg-2002-aa has suggested a threefold developmental sequence and typology of *sequential*, *linear* and *systematic* reasoning.
His empirical accounts suggest that if any, only systematic thinkers will be able to meet the cognitive demands for reasoned arguments, and egalitarian free speech of deliberative democracy.
Moreover, this cognitive competence was found not to be domain specific, and while people may regress to lower levels of competence under high loads or appropriate cues, they are unlikely to easily, if ever, achieve higher than developed levels.
"Structurally (more and) less developed reasoning adults" make "not only the adequacy of citizens' reasoning, but also their *equality*" a problem for deliberative democracy [@Rosenberg-2007-aa 12, emphasis added].

Few empirical work has been done to address both these ontological and empirical reservations towards deliberative democracy.
We know very little about if and how deliberative fora do in the face of macro-level abstractions and unequally limited cognitive ability of deliberators.

<!-- we need discrete choices for a CiviCon, otherwise we're all over the place, but it needs to be a choice that is relevantly abstract -->

<!-- %if you'd want to hide injustice, you'd best do it in a really complex system, like tax. -->

<!-- ### What's special about deliberating tax choice?

- Tax offers/demands very structured choices: there are essentially only 4x3 broad tax types (see tax table in drafts).
  Different from, for example, a deliberation on “racial tensions”, deliberating tax choice is more close-ended.
- Tax expertise is easily politicized, because:
	- Tax cannot be reduced to an empirical question, both because inframarginal data is not available, and because macroeconomics is incompletely understood.
	- Tax cannot be reduced to a consequentialist inquiry, because people (may) care about other normative claims, too (e.g. fairness).
- Tax cannot be easily (if at all) dumbed down to subsets of tax choice (e.g. VAT vs PIT), or subsets of abstractions (e.g. only “bastard keynesianism”) for two reasons:
	1. Any such reduction would violate the spirit of deliberation; citizens could not set their own agenda.
	2. Any such reduction may well fail to elucidate my hypothesized misunderstandings: ßI imagine “understanding tax” as a bucket, which, if leaky will drain to the lowest level. For example, if deliberators would discuss most aspects of taxation, but skip over the limits (!) of price controls, they may well opt for price controls in lieu of, or in addition to taxation. -->

## Hypotheses

Both research questions [\[itm:resolve-better-tax\]](#itm:resolve-better-tax){reference-type="ref" reference="itm:resolve-better-tax"} and [\[itm:prove-deliberative-democracy\]](#itm:prove-deliberative-democracy){reference-type="ref" reference="itm:prove-deliberative-democracy"} can be rolled up in one set of hypotheses:

## Methods

I test these hypotheses by subjecting people to a , pioneered by James [@Fishkin2009].
This 'Gold Standard' of deliberative fora [@Mansbridge2010 55], combines skillfully moderated small and plenary group discussions of randomly sampled citizens, rigorously vetted, balanced briefings and experts with pre-treatment, post-treatment and control group opinion surveys into a quasi-experimental design (as in , p. ).

\centering
![Method of the Deliberative Poll[]{label="fig:deliberative-poll-method"}](deliberative-poll-method){#fig:deliberative-poll-method width="1\linewidth"}

This design suits me well, because it combines the electoral part of democracy with the talking part [@ChambersKymlicka2002 308] and thereby offers the kind of quantitative data needed to test these hypotheses.
Both as a research method and as a second-order hypothetical, it also enjoys high external validity:
it delivers a collective choice and has been shown to work in actual policy making.

## Expected Results

### Hypotheses [\[itm:knowledge-gain\]](#itm:knowledge-gain){reference-type="ref" reference="itm:knowledge-gain"}.x are informed by prior research on the "Heuristics and Biases in Thinking about Tax" [@McCafferyBaron2003], supplemented by systematic misunderstandings specific to broad choices between income, consumption and asset taxation.
My hypothesized misunderstandings flow from anecdotal experiences I have had with laypersons and political leaders.
They also shine through in some of the welfare state research, and contrast with my synthesis of an (, p. ) and (, p. ).

### Hypothesis [\[itm:attitude-change\]](#itm:attitude-change){reference-type="ref" reference="itm:attitude-change"}, if confirmed, suggests that in fact, remaining *a priori* disagreements about tax *can* be resolved, and that hypothetical taxes really are desirable.

### Hypotheses
are related to public choice and opinion research in political science, showing that popular beliefs often violate [vnm]{acronym-label="vnm" acronym-form="singular+short"}-consistence and can fall prey to aggregation dysfunctions.
Opinions about taxation, in particular, may reveal anomalies at the individual and aggregate level, because tax is highly complex and demands highly structured choices.
Past research has shown that deliberation can alleviate preference structuration anomalies [@Farrar2003].
A [dp]{acronym-label="dp" acronym-form="singular+short"} on tax choice might replicate those findings and extend them to another policy area.

### Hypotheses [\[itm:interaction-effects\]](#itm:interaction-effects){reference-type="ref" reference="itm:interaction-effects"}.x inform the political psychology, both of tax and of deliberation.
If the above effects do not, in fact, interact with equity beliefs (hypothesis [\[itm:interact-equity\]](#itm:interact-equity){reference-type="ref" reference="itm:interact-equity"}, tax choice must be affected by something *other* than allocative preferences.
In (p. ), equity preferences are moves *along* a curve, and shifts or moves of curves reflect preferences orthogonal to equity.
If, as hypothesized, the above effects do not interact with equity preferences, deliberation may reveal (or bring about) popular understandings in line with the suggested institutional of (p. ):
If given some thought, people agree that they can shift the trade-off curve between, say, equity and efficiency outwards, and they prefer those institutions that do.
Hypothesis [\[itm:interact-ses\]](#itm:interact-ses){reference-type="ref" reference="itm:interact-ses"} is broadly related to the post-democracy thesis [@Crouch2004], by which late [oecd]{acronym-label="oecd" acronym-form="singular+short"} liberal capitalism and democracy not only disenfranchise the social contract, but do so at a socio-economic gradient.
The poor (and middle classes) may be too systematically confused to effectively act on their supposed self-interest, reinforcing a mutual crises of political and economic equality.
Lastly, hypothesis
is informed by the political psychology of deliberation, and the empirical research on unequal cognitive ability [@Rosenberg-2002-aa].
If, as hypothesized, people with greater cognitive ability will display greater changes along the aforementioned hypotheses, deliberations on tax, too, may have to consider such unequal abilities in their theory and practice.

More broadly, a [dp]{acronym-label="dp" acronym-form="singular+short"} on tax choice can inform several fields, including research on public finance and the welfare state, public choice and political economy.


<!-- In my research, I want to test whether ordinary people, if subjected to a deliberative process (similar to the Danish Consensus Conferences or a smaller BC Citizen Assembly), change their thinking, talking and decisions on taxation. I hypothesize, that in fact, they will change their preferences and understandings, revealing "enlightened" support for greatly more progressive taxes (including a post-paid, cash-flow-based progressive consumption tax). I have tried to provisionally translate these misunderstandings into a social choice account of amalgamated preferences (similar to your 2003 BJPS piece with List), but any empirical verification will need larger samples.

I have read some of your work and understand that you, too, find merely procedural standards for deliberative democracy too limited. I think that deliberation taxation, in particular, requires a more than procedural standard, both because it is highly complex and technical and because it is in those very abstractions that any more or less legitimate discourse ethic (Habermas 1984) would have to be found.

I find your suggested standard of "inter-subjective rationality" (with Niemeyer 2007 SPSR) very attractive; I would like to test a similar standard in the deliberation on taxation. -->

<!-- In terms of procedure vs substance in deliberative democracy - you are not so alone in seeking standards beyond procedure. Apart from Simon, I'm thinking of Gutmann and Thompson, especially their Why Deliberative Democracy? book; Joshua Cohen's essay on 'Procedure and Substance in Deliberative Democracy'; David Estlund on epistemic argument for deliberative democracy.  -->

<!-- Right now, I am struggling to spell out what a standard of intersubjective rationality may mean for deliberating tax choice (my area of interest). My position so far is that, to even test this standard I need 
a set of abstractions and remaining expert disagreement in which deliberators may -- or may well not -- find universal validity claims (borrowing Habermas 1984) (I have distilled a set of broadly consensual economic abstractions and disagreements in my chapters 4ff).
a relatively extensive, but rigorously balanced learning phase (akin to the BC Citizen Assembly on electoral reform), which people are introduced to (expert) concepts, alternatives and arguments, whose rationality they can then intersubjectively assert -- or refute. The trick must be to have such a learning phase without degrading deliberation into a classroom exercise in which the (economics) teacher knows best. -->

<!-- I continue to find the exchanges between social choice and deliberative theory very exciting; it seems like such a synthesis -- as you suggest with Christian List in the BJPS -- could reveal these disjunct strands of research to be concerned with two sides of the same, flawed coin that is pluralist democracy. If deliberation normatively implies an "epistemic conception of democracy" (Cohen 1986), social choice shows empirically and a-priori (Arrow, Condorcet, Gibbard-Satterthwaite ...), that absent some of such intersubjective rationality, unstable and dysfunctional aggregation will plague democratic rule. 
Unfortunately I have for now stepped back from applying such an analysis to taxation, but hope to do that later. I'd expect that pre-deliberative preferences on tax fall for aggregative dysfunctions (e.g. Condorcet cycles b/c not single-peaked), precisely as a function of collapsing distinct (value, belief ...) dimensions on tax into fewer, ill-formed dimensions, along some of the false tradeoffs and systematic misunderstandings of tax that I hypothesize. (e.g. "bastard Keynesianism" between more (consumption) taxation and less spending). I have also -- very provisionally -- argued that these misunderstandings and false tradeoffs translate into violations of von Neumann-Morgenstern rationality at the individual level.
I take the -- apparently uncommon and a little imprecise -- formulation of "amalgamated preferences" from David Miller (1992: 65, emphasis added), who defined them thus:
"I have suggested that the major reason apart from empirical error why preference orders are likely not to be single peaked is that the issue under discussion amalgamates separate dimensions of choice to which different voters attach different weights. I am claiming that it is a virtue of deliberative democracy (unlike, say, simple opinion-polling) that it will reveal this to be the case." 
-- Miller, D. (1992). Deliberative Democracy and Social Choice. Deliberative Democracy and Social Choice, 40, 54–67. 
Anyway, this, for now, is pie-in-the-sky for me.  -->


<!-- 'm currently wrestling with defining a substantive standard of argumentative quality, and justifying an appropriate deliberative format to test such a standard in deliberating taxation. Because taxation is such a complex topic, I think it requires a learning phase (akin to the BC Citizens' Assembly on electoral reform). I also think that the ontologically (ideologically?) loaded abstractions and inescapably political tradeoffs of taxation lend themselves well to a substantive standard of argumentative quality. In short, I think that a non-procedural view of deliberative democracy implies that to meaningfully deliberate taxation, some concepts and arguments should probably be considered at some point, on their own terms. I do not think that implies that people have to find these arguments compelling, let alone that they must prefer any given tax (that would collapse the exercise into a substantive dispute). I advocate in the proposal that we test robustly whether ordinary people do consider these arguments on their own terms, and, maybe, find some of them compelling (or not).

Is that a convincing argument (see attachment)?

Unfortunately, I came across your work on inter-subjective rationality only recently and I have not considered it in my proposal yet, though it seems applicable. I have so far rested my proposal on an -- admittedly cursory -- reading of Habermas, as well as Thompson's and Gutmann's work. However, I find your work more amenable to empirical verification, and admire the rigorous operationalizations, including Q-methodology.

Here are some preliminary thoughts on how your operationalization of deliberative outcomes might play out in taxation:
Meta-consensus
Epistemic: Agreement on the ontological status of tax: including, for example, a provisional and domain-constrained but pragmatic assumption of homo economicus (which may not be agreed to be an adequate, much less desirable description of human nature)
Preference: Agreement on some basic causal beliefs: including, for example, the non-trivial incidence of taxation (who ends up being burdened is often not the same person who pays the tax) and that by definition, only natural persons are moral subjects.
Normative: Agreement on some tradeoffs, including, for example, that given the above, taxes always present -- if differing -- trade-offs between efficiency, equity and sustainability (all of which are legitimate)
Inter-subjective rationality
People who a) believe that income taxation hurts incentives and b) value a progressive schedule should also c) prefer a progressive consumption tax.
..
This application is still fairly half-baked, and I'll need to work more on that.

For now, I'd appreciate your comment on my proposal, especially considering the role of learning phases and a substantive standard to which deliberation should be held.

I'm curious whether you see a connection, between meta-consensus, inter-subjective rationality and learning phases. It seems to me that learning phases designed to suggest bases for meta-consensus might be well-placed in deliberation, all the while avoiding a substantive stance in the issue at stake. -->


So, "who dunnit\"?
Because this dissertation is, at bottom, positivist, I cannot show what --- let alone *who* --- prevented a better tax, or a better democracy.
Non-events such as as land or consumption taxation and deliberative democracy are always causally underdetermined, just as history is always overdetermined.
For a detective, that is a bit of a let-down, but there is still work to be done.
I might not find the perpetrator, but at least, I can absolve wrongly accused democracy, and, during the hearing, spread word of the crime.

<!-- Your work on operationalizing meta-consensus and especially intersubjective rationality via Q methodology has grown on me, and I think it might work really well for taxation. It seems to be a more powerful approach than merely measuring attitude and knowledge change (Fishkin), mostly procedural "Deliberative Quality" (DQI Steenbergen, Bächtiger et al) or even preference structuration (Luskin et al) -- even though these may well be antecedents to (process) or functions of (structuration) intersubjective rationality. 

It's always been my intuition that there are different popular preferences in tax, stemming from similar (lip-serviced, muddled) beliefs and values -- and vice versa. For example, it would seem odd that someone, say, a self-identified libertarian, who is opposed to coercive allocation of material resources and knows that a lot of wealth is generated by extracting natural resources under the coercive protection of states would not also be in favor of taxing such rent-seeking (for example, through a land value or capture tax). However, if I try to apply such an exogenous standard of consistency to observed arguments, some social scientists get uncomfortable, maybe (rightfully) fearing a confusion of deliberative process and epistemic substance, or, in short: Who am I to say that such an argument would be inconsistent?

It seems a little odd that a practice founded on the hope (though not guarantee) of communicative rationality (Habermas) would shy away from critiquing the rationality of any observed argument, and it is disheartening to hear that your operationalization has not been universally welcome.

It seems to be that the appeal of your operationalization, especially the application of Q methodology is that it manages to uphold a standard of intersubjective rationality without reverting to some exogenous, substantive theory (even though there should be a place for that, too).

My immediate concern with my PhD project is that I face -- amongst other constraints -- a tradeoff between the size of the sample, and the length of the deliberation, especially any learning phases (akin to the BC Citizens' Assembly). I understand that Q methodology would need a pretty large sample to yield robust results, but I am not sure that with many people, but little time, much can be accomplished in terms of improving intersubjective rationality, especially on such a deeply technical issue (tax), rife with meta disagreement. 

Would you agree?

I am now setting out to exploring ways of adapting your research logic to a qualitative, small-n design, which seems  more appropriate for my means. If nothing else, your notions of meta-consensus and intersubjective rationality will be helpful in reformulating my hypothesis.

I'll keep you posted about my progress. -->

<!-- First, re the inclusion of more than one Q sort. I’ve always included value and belief statements into the same Q sort. This is partly because in practice it’s been impossible to absolutely demarcate between the two types of statements, particularly when they are drawn from real language. But even if it is possible to demarcate clearly between them, there is no reason not to do them at the same time. If you do want to conduct a separate analysis on each type of statement, you could simply reconstruct them into two separate Q sorts using their scores from the combined Q sort.  -->

<!-- might do the same with preferences -->

<!-- %here used to be \ref{fig:tax-with-misunderstandings}, now only in proposal

% \citep[334][]{Farrar2010}
	%The present study examines three important hypotheses about deliberation’s effects.
	%The first is that deliberation frequently alters policy attitudes (continuous dispositions towards policy alternatives) – not only at the individual level (‘gross change’) but in the aggregate (‘net change’).
%The second is that deliberation tends to bring policy preferences (ordinal rankings of policy alternatives) closer to single-peakedness, a help in avoiding cyclical majorities of the sort identified by Condorcet and Arrow.
	%The third is that both these effects tend to be stronger for less salient issues, on which less deliberation has already occurred.
	%Since very few issues are highly salient, these hypotheses together imply an important role for deliberation in shaping majorities and making them meaningful.

	%deliberation helps NOT by brining unanimity, but by changing the structure \citep[2]{List2007}

	%here is something important:
%Condorcet works for pairwise comparisons:
%here we got a problem
	% expanded by arrow:
%ANY voting method must violate one of four attractive properties of pairwise majority voting -- which is why it still matters this problem.
%(Fishkin, etal on meaningful democracy:
%1)
	%violations of independence of irrelevant alternatives open the door to agenda manipulation (Riker 1982), strategic voting (Gibbard 1973, Satterthwaite 1975),

	%deliberation increases ``meta-agreement'' on the dimensions of policy, which, in turn decreases the probability of cycling majorities \citep[9]{List2007}

	%also, say \cite[9]{List2007}:
%these should be the REAL dimensions.

	%\cite{List2007} show that you can get more Condorcet winners this way

	%genschel:
%you show that priming works.

	%social choice (Arrow, Condorcet, Gibbard-Satterthwaite), both in \citep[2]{Dryzek2003} and deliberative democracy -->


<!-- 1. Research Question [based on proposal] -->
<!--    - Explicates *research question* at the intersection of the literature on (subjective understandings of) tax, and (empirical research on) deliberative democracy -->
<!--    - includes literature review of tax misunderstandings/disagreements and empirical deliberative democracy -->
<!--    - Why tax is a good *case*; and more than a case. -->
<!--    - Why deliberation is a good *method*; and more than a method. -->
<!--    - Draft completed by **2014-01-31** -->

<!-- ===
why do we not have it?
===

%Genschel:
%"An idea whose time never came"

%add comment, footnote:
%this was first submitted as \ldots

%How Do We Get The Perfect Tax?

\section{Because It is Not Better}
%or:
%how to go further?

%First Order Questions of Social Change
%The first order questions are about the supposed superiority of the PCT.
%They include:
% Is the PCT really a more efficient, more equitable tax (quantitative ‘proof’)?
% This requires an extensive model of the economy, difficult empirical specification of key inputs (for example, elasticity in demand for luxury goods, saving), agreement on a number of contentious axioms (diminishing returns, positional consumption, optimal savings rate).
%This task is certainly beyond my training, arguably beyond the scope of a PhD thesis and possibly beyond the modeling powers of economics.


%consider the "Garbage Can model"/Kingdom for my tax.
%Maybe the time of my tax didn't come yet.
%But maybe, it's here now (sovereign default in Spain?!?).
%Read Kingdon.
%This is an idea that was developed in Kohler-Koch's idea seminar.

\subsection{Modeling}
%write a section on why modeling is futile.

% There is still considerable work to be done in specifying (cardinally) the schedule and base of a PCT regime, in its implementation and transition.
%This work depends, again, on a comprehensive model of the economy and bureaucratic detail.
%This grunt work is unlikely to yield analytically productive insights, and may best be confined to think tanks and tax administrations.
%Implementing and specifying the PCT, absent a (good) model of the entire economy will probably be a matter of trial and error.

\subsection{Specification}
%do i need this here?
% How would a PCT be (cardinally) specified and implemented?




%Second Order Questions of Social Change
%The second order question of social change is simply:
%Why don’t we have it?
%Possible answers (hypotheses) are:
% Progressive Taxation is subject to an international cooperation problem, akin to a Prisoner’s Dilemma.
%Because the PCT is progressive, it is individually rational for states not to implement it, even if it were collectively rational.
% This is a very interesting question that could potentially greatly enhance our understanding of the international political economy of taxation.
%It however, also requires a comprehensive model of the economy, and even international trade and finance to anticipate the hypothetical payoff of the PCT.
% Tax regimes are subject to great path dependency, they cannot be easily reformed.
% While a historical and tentative analysis of path dependency in tax regimes is possible and probably productive, the costs of reform can ultimately only assessed when a cardinal specification and implementation has been undertaken.
%Assessing the costs of the reform ex ante may be very hard, unreliable or impossible.
%Both a possible international cooperation problem and path dependency in tax regime choice are ultimately endogenous to broader questions of the political economy.
%The cooperation problem depends on the level at which the international political economy is governed.
%Path dependency depends on the domestic and international incidence of reform costs, in turn depending on the states ability to hand out side-payments.

% Ultimately, the PCT could be absent because people do not want it.
%The second order question of social change in tax regimes is:
%will people reject the PCT irrespective of the democratic process that governs their polity?
%If not:
%how do different democratic processes (pluralist, deliberative) constrain or even determine the politics and choice of tax?

%likewise:
%assumed theoretical link:
%if you want to hide injustice, than you'd best do it in a really complex system, like tax.

%what's the second order theory?

%I don't explain the elite-based (potential) reason for the non-reform.
%That, the second-order theory, would be someone elses job.
%discourse theory here?
Because It is Not Better
========================

Modeling
--------

Specification
-------------

Tax Competition
===============

Political science offers at least three rival perspectives to explain why we do not have the perfect tax, and common sense adds a third (page ).
I present these rival explanations here in a preliminary form.
Empirical testing and more specification must await further research.

Path Dependence: We Have Always Done This
-----------------------------------------

The path dependence perspective suggests that social systems generally resist change.
As specific institutions and solutions are adopted, their (re-)production enjoys economies of scale, and it becomes ever harder to alter or replace past arrangements [@Mahoney-2000-aa; @Pierson-2000-aa].
Once the QWERTY keyboard was introduced in 1878 and people had learned to use it, it was carved not only in plastic, but also in stone.
Other keyboard designs have never attracted a significant following given the strong network effects (using other keyboards) and large economies of scale (being really good at QWERTY) of the QWERTY design.
Incidentally, QWERTY was never designed to be particularly efficient, but to minimize the clashing of two neighboring typebars, a problem which is now long obsolete.

Much the same story can be told for the personal income tax:
it was originally implemented not on superior principle, but because of its intuitive appeal and practicality.
In a time before widespread retail, let alone electronic banking, the PIT was attractive because it could be withheld directly at the source of few, well-organized firms in the newly industrializing world.
Once adopted, the PIT was strengthened by network effects (other countries)
[^1]
and economies of scale (taxmen and taxpayers know it well).
Fundamental reform, again and again, failed.
As the (page ) became ever more apparent, it was supplemented or replaced by other problematic (regressive), but *compatible* taxes:
(page ) and increasing payroll contributions to social insurance.

Deficient Political Process: They Don't Get it
----------------------------------------------

An alternative, more cynical perspective suggests that while PCT reform is possible, it is never undertaken because of a confused and deficient political process.

On cursory inspection of the political debate in tax, this perspective is intuitive if appalling.
The public debate on tax, hardly, if at all corresponds to the abstractions governing tax.
When people and politicians demand more corporate taxation or fight for employer and employee parity in the financing of social insurance, at least one of them does not know what they are talking about[^2].

More systematic evidence for the inaptitude of voters come from psychology [@Converse-1970-aa], opinion research [@Delli-CarpiniKeeter-1996-aa] and recently, political psychology [@Rosenberg-2002-aa].

A look in the cookbook of a professional campaign strategist reveals a more fundamental malaise:
"the task of the campaign strategist is to find the easiest path to victory" [@Malchow2003 9].
If politicians and parties go for the low road, campaigning on topics and positions that are easy to explain[^3], tax will necessarily fall by the wayside.

Global Cooperation Problem: Race to the Bottom
----------------------------------------------

The international political economy perspective suggests that the PCT does not exist because of a widespread cooperation problem of high taxation.
Whether globalization, or, more accurately denationalization retrenches the welfare state, is a question widely discussed in political science.

#### The Missing [Normative Counterfactual](http://maxheld.de/2009/10/13/setting-goalposts/).

To know whether, and how much denationalization retrenches welfare states, political science would have to assess the welfare regime in a sufficiently large, prosperous and closed economy.
This of course, does not exist in the OECD world, and may not exist at all in reality.
I argue that the appropriate response, if only for methodological reasons, is not to ignore this problem, but to hypothesize a best case, normative hypothetical, where *cooperation is given*.

The PCT *is* that desirable and doable hypothetical that would be expected if cooperation were given.

...And a Little Conspiracy Theory {#sec:Conspiracy}
---------------------------------

The practical and "analytic muddle" that tax is [@McCaffery2005 862] raises suspicions beyond path dependence, political process and tax competition.
Current tax systems are not just poorly designed and poorly understood.
They are mis-configured and misrepresented in a very *specific* way.

Progression in the current fiscal configuration rests on the increasingly feeble and dysfunctional income taxation.
The PIT is under pressure from two sides.
On the one hand, its ugly backstops, the gift, estate and Corporate Income Tax are ever open to evasion while stifling economic activity.
On the other hand, income taxation, particularly of capital incomes is extremely vulnerable to tax competition.
As capital became increasingly mobile over the past decades, progressive PIT and high CIT rates became untenable.

As the PIT and its backstops crumbled as a source of revenue, states turned to the (seemingly) only alternative:
flat VAT-style, prepaid consumption taxes and proportional wage taxes or equivalent Dual Income Taxes.
These taxes were roughly compatible with the remnants of income taxation, did not require dysfunctional backstops and were relatively immune to tax competition.
VAT, Payroll and Dual Income Taxes all fall on relative immobile bases:
it is much more costly to move abroad for a worker with her family than for a mutual fund to change its portfolio.

For the past three to four decades now, governments have shifted their tax mix from PIT to VAT and payroll taxation (data for Germany and the UK in @Kemmerling2009 [11].
This shift comes at a hefty price:
as VAT and payroll taxation are proportional, the tax burden on low-productivity workers increases.
Present some socially acceptable minimum living standard, structural unemployment ensures, which in turn requires transfers out of general revenue.

The late developed welfare state is in a pretty bad fix.
Given costly unemployment, real dissavings and increasing public debt it is *structurally underfunded*.
If it raises the PIT, capital drains and growth falls.
If it raises the VAT and payroll taxation, *structural unemployment* rises or working poverty ensues [@Kato2003].

> *"This traditional view has generated an impoverished choice set for tax, consisting of a badly flawed status quo on the one hand and a flat consumption tax of some sort on the other.
> Under the guiding light of the traditional view, we are heading ever closer towards a flat wage tax."*\
> Edward J. @McCaffery2005 [812]

The status of quo of tax is not just arbitrarily bungled.
It is misconfigured to lead us to ever more proportional, if not regressive[^4], taxation of immobile sources.
It is misconfigured to lead us to a polity paralyzed by underfunding, to an economy of debilitating structural unemployment or heart-wrenching working poverty.
It is misrepresented to make us belief that our only choice in tax is between anti-growth PIT and proportional VAT or regressive payroll.

> *"The real and pressingly practical question for tax is not whether to have an income or a consumption tax, but what form of consumption tax to have.
> The stakes in this battle are clear and dramatic:
> the fate of progressivity in tax lies in the balance."*\
> --- Edward J. @McCaffery2005 [817]

The status quo of tax is bungled in a way that serves the interest of the beneficiaries of the status quo.
If you are rich, you are content with how the pie is sliced.
If you are rich, you do not depend as direly on underfunded public goods and risk pools:
you have your exit options (gated community, private health "insurance", boarding school).
If you are rich, and if your capital is mobile your luck does not depend even on growth in your home country:
you get to send your capital abroad, to whichever booming economy generates the greatest income.

This is not to suggest a historical materialism of tax.
Structuralist scapegoating is often analytically hermetic, sometimes rhetorically dangerous and always politically impractical.

But when a policy failure has such conspicuous distributive consequences, one has to wonder whether agency and interest make regressive taxation of immobile bases a *successful* failure[^5].
For legislators captured by, or serving rich special interest window dressing minimally redistributive and defunct tax policy may just be the professional thing to do.

How Could we Get the Perfect Tax? {#sec:HowToGetIt}
=================================

Given all these obstacles, getting the PCT will not be easy.
There is, however, one effect of the PCT which may act as a catalyst for its introduction and proliferation.

Recall that the PCT will replace all other revenue-generating taxes, including those on capital and corporate incomes.
In the short to medium run, a solvent government could even lift income taxation on foreign residents and foreign-owned corporations.
As a result, capital would rush into the PCT early adopter country, causing substantial harm to other income-taxing countries [@Dalsgaard2005 12].
*Every* capital owner abroad will be better off for *every* unit of capital they invest in the PCT early adopter.

Without international harmonization and cooperation, *home* residents will have an incentive to evade the PCT.
As argued in the above, they may either migrate altogether, or invest abroad and consume abroad out of their offshored capital incomes.
Either way, they will be subjected to income taxation abroad, but get to avoid the PCT at home.
Because capital owners still pay CIT, PIT or a withholding tax abroad, they are better off only if, and to the extent that they wish to consume their offshored capital incomes.
For many capitalists, this will be a (small) component of their overall net worth.

In the balance of these two investment flows, it appears that the PCT early adopter will gain significant capital inflows.
This, of course, is classic *beggar-thy-neighbor* policy.
In the context of the PCT, threatening other countries with zero income taxation may however serve as a catalyst for the introduction and proliferation of the PCT[^6].
If a critical mass of countries adopts the PCT and abolishes all income taxes, other developed trading partners may be forced to do the same.

This hypothesized dynamic requires much greater specification and econometric as well as game theoretical modeling.
If the trick works, the PCT carries within it not only the seeds to its own beggar-thy-neighbor proliferation, but also a promising recipe to overcome the international cooperation problem of progressive taxation.

Loose Ends {#sec:LooseEnds}
==========

This thesis generates pressing questions and promising avenues for further research.

#### Quantifying the PCT: Experiment or Model.

So far, I have *deduced* the PCT as the *ordinally* superior tax.
I have argued that *given* desirable and doable desiderata, the PCT emerges as the perfect tax.
I have not shown *inductively* that the PCT produces results in the real world that are normatively preferable.
I have argued that the PCT is better (*ordinally*), but not *by how much* (*cardinally*), both in terms of efficiency (welfare), equity (distribution) and sustainability (savings rate).

The PCT should be tested *inductively* and the superiority of the PCT should be *cardinally* quantified.

##### Inductive Test.

A straightforward *inductive* test of the PCT is not possible:
no large, rich country will implement the PCT just to "give it a shot".
Existing econometric research on the effects of taxation will also be inadequate:
market reactions to such a discrete, encompassing reform cannot be gauged from observations of incremental, continuous changes.
To name just one example, the price elasticity of excessive (positional) consumption is unknown at very high rates of taxation.
Instead of econometric research, a carefully crafted experiment in behavioral economics may serve better to verify the PCT inductively.

##### Cardinal Quantification.

Showing exactly *how much better* the PCT will make us in the real world hinges on the same problems as an inductive test.
Absent a *test run*, any attempt at (general equilibrium) modeling of the tax will suffer from underspecification.
Again, the price elasticity of excessive (positional) consumption is one of the key unknowns:
just how much revenue the PCT will raise at any given schedule, or how much positional waste it will curtail is not readily observable.

An economic model of rich economies with a PCT would still serve to at least roughly quantify the drivers of welfare and distribution.

Again, a behavioral experiment may be a more elegant, and powerful design to cardinally estimate the benefits of the PCT.

<!-- %why don't we have it?
	%Because it may not be better?
		%requires general equilibrium modeling, utility functions
		%depends on a number of axioms
	%how would it be specified, implemented
		%requires ge model, utility functions
		%modeling grunt work
	%there is an international cooperation problem!
		%requires general equilibrium model, utility functions, game theory
	%domestic, political dysfunctions prevent it!
	%maybe: path dependency, game theory, veto playing, principal agent
	%because people do not want it!
		%survey research, vignette research, experimental economics -->

#### Specifying the PCT.

As is argued in the above, the PCT still requires a lot of grunt work detail.
In particular, monetary dynamics of the PCT must be addressed, its formulae and schedule must be specified, an elegant design and administrative process must be drafted and a its implementation must be planned.
Specification of the PCT will depend on cardinal quantification of its effects and require extensive economic modeling as well as econometric data.

#### Informing the International Political Economy of Progressive Taxation.

If the PCT is indeed superior, cardinally and inductively, this verifiably desirable and doable hypothetical will put new questions to the literature on welfare retrenchment.
Given that a better configuration is possible, political science must answer whether denationalization and tax competition are to blame for the present misery.

Two formal approaches apply here.

##### Multi-Level Game Theory.

One is the multi-level game theory of international tax harmonization [@Scharpf-1997-aa].
Here, considerable effort should be devoted to understanding the strategic imperatives of governance both at the national and international level to adopt or avoid the PCT.
An econometrically informed modeling of the distributive effects *within* and *between* countries will be required to gauge this multi-level game[^7].

##### Veto Playing.

Closely related, international tax harmonization should be modeled as a veto-player problem [@Tsebelis-2002-aa].
Given the anarchic nature of the international realm (taking on a realist view), veto playing appears to be a reasonable approximation of international tax harmonization.
In particular, a modeled comparison between tax harmonization at the international and European level with their different veto rights may be instructive to estimate the overall significance of the model.

#### Modeling the Introduction of the PCT.

Lastly, the chain reaction dynamic hypothesized in the above will need deductive, as well as inductive verification and econometric quantification:
just how much costs will the PCT early adopter incur to the non-adopters?
And how quickly will these costs materialize?
How large would a critical mass of countries have to be to introduce and proliferate the PCT?
Which countries would be suitable to ignite the chain reaction? -->


Such a difference $\Delta$ between "raw" and "enlightened" thinking,
beliefs [@Caplan2007] and preferences [@Fishkin2009] is worth knowing for several reasons:

1.  There may be *agreeable*, doable but hypothetical taxes, which may (Pareto)-improve over current taxes [@Harberger1974] or may be fairer [@Rawls-1971] and more sustainable [@Solow1956], including a [@Mill1848; @McCaffery2002; @Frank2005a; @Seidman1997], a [@George1879; @Buiter1988] and a [@Friedman1962].
    By comparison, existing taxes appear unnecessarily wasteful, inequitable and unsustainable.[^oecd-taxes]
    <!-- COMMENT MN würde dieses Wissen schon so eingespeist? Bias? -->

    These suboptimal tax regimes may present democratic sovereigns with needlessly harsh tradeoffs between efficiency, equity, sustainability and other competing ends.
    The means of an intact mixed economy [@MusgThet1959; @Stiglitz2011] may consequently erode or even force retrenched welfare states into "permanent austerity" [@Pierson2002; @StreeckMertens2010] as it becomes harder for them to reconcile planned allocation and market exchange through taxation.

    Public confusion about taxation may be a contributing factor, or even a sufficient --- but not a necessary --- condition for the non-existence of these supposedly superior taxes.

2.  Progressivity in tax, especially, may hinge upon an enlightened, "new understanding of tax" [@McCaffery2005], as the historic mainstay of redistribution --- the [pit]{acronym-label="pit" acronym-form="singular+short"} --- withers away in the face of its inherent contradictions [@McCafferyHines2010].
    People may understand tax *systematically* different from how they would under (more) ideal speech, which may cause them to choose less effectively progressive taxes than they otherwise might, or otherwise inadvertently act even against their material self-interest.
    Additionally, the difference between raw and enlightened thinking about tax may, in itself, depend on the socio-economic status of citizens.
    Taken together, such socially structured and structuring differences in understanding tax may contribute to further, or perpetuate existing social inequality, even under nominal political equality.

3.  The aggregative and representative institutions of liberal-pluralist democracy may be partly outmatched by the vexing complexity [@Merton-1968-aa] and tightly concentrated interest [@Olson-1971-aa] of late market economies, such as in taxation, violating both their input and output legitimacy [@Scharpf1997] as well as basic democratic prescriptions for "enlightened understanding" [@Dahl-1989-aa].
    Empirical political sociology bears out such popular dissatisfaction, not with democracy it self, but with incumbents, enamored by special interest, and inattentive to the public good [@NyeJr.1997; @Norris2011; @PutnamPharr-2000-aa].
    If such a "post-democratic" constellation [@Crouch2004] were to plague the polity in its thinking about tax, it may well affect other policy areas too.

[^oecd-taxes]: In the [oecd]{acronym-label="oecd" acronym-form="singular+short"}-world, those are varying combinations of pre-paid, proportional consumption taxes such as [vat]{acronym-label="vat" acronym-form="singular+short"} and income taxes, often with different schedules for labor, capital and corporate incomes.
    Wealth taxes, often only on some asset classes, exist in some jurisdictions, but --- with the exception of local US property taxes --- are not a large revenue source.


## State of the field

This positive aspect of this thesis sits at the intersection of two fields of empirical research:

1.  Practical experiments with different deliberative fora on varying policy areas,
2.  Experimental and survey research about popular understandings of tax.

<!-- In the following, I briefly report key findings from these three fields and identify avenues for further research. -->

### Empirical Deliberation

#### Deliberative Experiments

Deliberative fora have been tried on a number of policy areas and places, including renewable energy in Texas (USA) [@LehrGuild-2003-aa], participatory budgeting in Porto Allegre [@CoelhoPozzono-2005-aa] (Brasil), nanotechnology [@Powell2008], electoral reform in British Columbia (Canada), policing and education in Chicago (Illinois) [@WarrenPearse-2008-aa], and city planning in Philadelphia (Pennsylvannia) [@Sokoloff2005].
They also vary in their institutional design [reviewed in @Fung-2003-ac].
These include several small-n, multi-day designs, such as Citizen Juries [@SmithWales-2000-aa] or Planning Cells [@Dienel-1999-aa], where a few randomly selected citizens hear expert witness and deliberate amongst themselves for a couple of days, all facilitated by trained, non-expert moderators and prepared by a balanced briefing book to draft a Citizen's Report or similar consensus recommendation.
The otherwise similar Consensus Conferences [@Einsiedel2000] are often (much) longer, feature extended learning periods and are typically held on highly technical issues.[^consensus_conferences]

[^consensus_conferences]: Consensus Conferences were first developed and hosted in Denmark by the Technology Assessment Board to collect informed popular input for regulating science and technology.

Deliberation has also been tried in large-n, shorter designs with a more quantitative methodology in [dp]{acronym-label="dp" acronym-form="singular+short"}s [@FishkinFarrar-2005-aa], where hundreds of randomly selected citizens receive issue books, deliberate in small, moderated groups and plenary sessions, hear expert witnesses and --- instead of reaching a consensus --- cast a secret vote or fill out a questionnaire, after the typically daylong deliberation concludes.
Often, these survey results are compared to answers provided by participants before the event, and/or to a control group of non-participants, allowing quasi-experimental within-subject or between-subject statistical inferences.[^drop_out]

[^drop_out]: Because participants can always drop out at will and will know when they are in the treatment condition, treatment cannot be strictly randomly assigned.

There is quite limited experience with long, large-n designs, including --- to my knowledge --- only the [@WarrenPearse-2008-aa] that spanned several months, and included several hundred participants who were offered stipends and accommodation, participated in extended learning sessions, attended public hearings and ultimately agreed to recommend a new electoral system (a [stv]{acronym-label="stv" acronym-form="singular+short"}) for the province [@Citizen-2004-aa].
<!-- MN true, but it failed? -->

<!-- TODO include tables from google doc. -->

Across these different designs and purposes, deliberative fora have been shown to improve knowledge of the subject matter in question, to change --- but not bifurcate --- beliefs [@Fishkin2009], to yield more orderly and orthogonal preferences structures [@Farrar2003] and to boost a sense of political efficacy or mutual trust even among disempowered citizens [@Karpowitz2009].

These encouraging findings on the capacity of deliberation to strengthen democratic rule notwithstanding, much of deliberative practice may be criticized for often focusing on the immediate, tangible and remaining "confined to the realm of neighborhood and locale" [@FungWright-2001-aa 17].
<!-- COMMENT MN das sind aber nicht die Hauptkriterien? -->
This includes not only the overtly "human-scale" democratic efforts [@Boggs-1997-aa 759] of participatory budgeting [@Sousa-Santos-1998-aa], city planning [@Sokoloff2005] or public stewardship of forests [@Cheng2005], but also the much-lauded [dps]{acronym-label="dp" acronym-form="plural+short"}, which have, even when they were about clearly large-scale issues such as the [eu]{acronym-label="eu" acronym-form="singular+short"} [@Fishkin2009] or renewable energy use [@LehrGuild-2003-aa] mostly shied away from the very structured choices and remote abstractions which policy must engage.
These [dps]{acronym-label="dp" acronym-form="plural+short"} have, for example, demonstrated that more people prefer [eu]{acronym-label="eu" acronym-form="singular+short"} enlargement and more people would be willing to pay a premium for renewables [@Fishkin2009 157], but did *not* ask --- and probably not deliberate on, either --- whether the [eu]{acronym-label="eu" acronym-form="singular+short"} was an [oca]{acronym-label="oca" acronym-form="singular+short"}, and if not, what fiscal complements might have to be implemented [for example, @Mundell1961], or by which process *competing* renewable energies should be chosen, especially if they depend on economies of scale and network effects [for example, @Krugman-1990-aa].
Clearly, these are only two haphazardly selected examples of abstractions ([oca]{acronym-label="oca" acronym-form="singular+short"}) and choices (infant industry regulation) relevant to these topics, but equally clearly, someone, somewhere in the democratic process has to make these calls.
<!-- %fishkin quote in the above is K2010 -->

Some of these more technocratic considerations may be rightly relegated to experts, but for others --- including the abstractions of tax --- this may not be possible or desirable as discussed in the above.
In terms of @Landwehr2010, [dps]{acronym-label="dp" acronym-form="plural+short"} may not be sufficiently "coordinative" to be fully deliberative, and not only because they do not enforce a collectively binding decision [-@Landwehr2010 377].
Absent abstractions, "coordinativeness" suffers because the policy options (for example, "faster [eu]{acronym-label="eu" acronym-form="singular+short"} expansion") in which preferences are expressed [-@Landwehr2010 375] are ill-defined (for example, *what kind* of european integration?).
If such expressed preferences are allowed to detach from ontologically given policy options (for example, currency union with our without fiscal complements), people can effectively "exit" from substantial agreement [@Landwehr2010 377]:
"deliberators" can just talk, which may still be desirable, but stretches the concept too thin [@Thompson2008 502].
<!-- %Or, in the words of \cite[2]{Moore2011} ``the broadly democratic idea that people ought to have equality of opportunity to contribute to deliberation on matters that affect them is undermined by the huge and manifold inequalities in knowledge that are necessary for the effective analysis, regulation and management of complex social and technological problems'' -->

To show its stripes --- or limits (!) --- deliberative democracy must be tried on such highly structured choices and abstract issues, too.
<!-- %COMMENT MN notwendigkeit der verbindung mit Entscheidungskompetenz? -->
So far, with the notable exception of some Danish (small-n) Consensus Conferences and the (large-n) [ca]{acronym-label="ca" acronym-form="singular+short"} [on its complexity, @Blais2008], such applications are lacking in empirical experiments with deliberation.
If deliberative practice continues to blank out governing abstractions and structured choices --- the very stuff of the modern world and its political rule --- it risks moving "in a defensive and insular direction, laying bare a process of conservative retreat beneath a facile rhetoric of grassroots activism" [@Boggs-1997-aa 759].[^drawing_delib]

[^drawing_delib]: A comically naive example of such retreat comes from Piombino, Italy, where schoolchildren were "involved" (sic!) in urban renewal of the town piazza:

    > "After school trips (...) they had to make drawings of how the piazza should look.
    > These often very colorful and joyful drawings were exhibited and also put on the website of the town."

    Praises @Steiner2012 [29]:
    "In this way, children learned a good lesson about practical politics with a very concrete case of policy-making" (sic!).
    These and other exercises in "involvement" may be pedagogically or otherwise valuable, but *emancipatory* deliberation they are not.
    In fact, the example of children seems awfully revealing as an operative metaphor for this kind of "deliberative" citizen:
    let them draw cute pictures, or busy themselves with otherwise inconsequential activities, so that they feel heard.
<!-- %in a different, more positive vein, \cite[4f]{Boswell2013} describes such deliberative designs as concerned with `` `low politics', where the stakes are lower, the glare of publicity less pronounced and the local knowledge of citizens more prized. -->
<!-- we could also be infavor of this, simply as a kind of societal glue -->

<!-- really add this as a completely new reason for my diss: the concern that deliberation might be watered down, professionalized to the degree where it becomes just marketing.
elisabeth clemens maybe? some other books at ssha -->


#### Academic Reflection

These and other deliberative fora have also received ample academic reflection.

One strand of positive research treats deliberation as the *independent* variable, investigating its effects (including @Bachtiger2005 on legislatures, @Hibbings2002 on power differentials or @Jackman2008 on better-grounded judgements).
<!-- %fishkin 1991, niemer 2002, Hansen 2004, Schneiderhahn und Kahn 2008 all do effects -->
In a similar vein, [@Steiner2004] studied in how far deliberatively reached decisions "correspond to criteria of social justice" [@Steiner2012 13], by which they seem to mean a utilitarian maxim [@Bentham1789] or the @Rawls-1971ian difference principle [@Steiner2012 95].[^steiner_v_rawls]

[^steiner_v_rawls]: @Steiner2012 presents a questionable summary of @Rawls-1971:
    > This principle states that in a political decision the most disadvantaged social groups should profit the most [@Steiner2012 95].

    He finds it expressed when a parliamentarian says:

    > The government's priority must be to help those who have the greatest difficulty in paying rather then spreading it evenly across the range.

    The difference principle actually states that "social and economic *inequalities* are to \[...\] be to the greatest benefit to the least advantaged members of society" [@Rawls-1971 266], a complex and conditional claim to moral validity that is characteristically degraded in both the operationalization and the coded snippet.

In other work, deliberation is the *dependent* variable, investigating its causes (including @Steiner2004 on institutional aspects or @Steiner2012 [13] on issue polarization).
Similarly, [@Landwehr2010] compare how the *discursiveness* and *coordinativeness* of parliamentary debates and citizen conferences on regulating embryonic stem cell research impacts resultant preference change, their proxy for deliberation.
Employing a more sophisticated and theory-grounded operationalization than most, @Landwehr2010 [377] analyze speech acts to code for arguing (for example, "to claim") and bargaining (for example, "to offer").
<!-- %developed in Holzinger 2001, also Niemeyer. -->

Lastly, researchers simply seek to operationalize deliberation.
For instance, [@Stromer-Galley2007] suggests a coding scheme, including "reasoned opinion expression" when people provide "a definition, a reason for holding the opinion, an example, a story, a statistic, or fact, a hypothetical example, a solution to the problem, further explanation for why the problem is a problem, an analogy, \[...\] or any further attempt to say what they mean or why they have taken the position that they have".

These academic reflections all try to abstract away the substantive issues at stake in the respective deliberative encounters.
Social scientists, here, as often, are loath to get caught in first-order conflicts over, say, school policy:
"\[...\] measuring the factuality or quality of reasoning was deemed unnecessary" [@Stromer-Galley2007 10].
This impulse is understandable, even commendable:
Social scientists are, by definition, in the business of explaining the second-order social conflict *over* normative and positive first-order questions [compare @GutmannThompson-2004-aa 12] and must generalize from the particular to build or test social theory.
Yet, since deliberation is a *discourse* ethic [@GutmannThompson-2002-aa 153], it is difficult to analyze without the actual, domain-specific discourse.
The operationalizations employed vary widely, ranging from the crudely statistical [@Steiner2012] [^count_interruptions] to the quick-and-dirty semantical [@Steiner2012] [^dqi_def] to promisingly speech-act linguistic [@Landwehr2010].

[^count_interruptions]: Summarizes @Steiner2012:
    "It is noteworthy that of the 1,027 speech acts, only 5 were interrupted by other participants, which indicates a calm and not boisterous atmosphere during the experiments" [-@Steiner2012 47].

[^dqi_def]: "\[O\]ur [dqi]{acronym-label="dqi" acronym-form="singular+short"} defined a sophisticated level of justification as having a certain complexity in the sense that an argument is justified with more than one reason and that these reasons are logically linked with the postulated outcome" [@Steiner2012 64].

The [dqi]{acronym-label="dqi" acronym-form="singular+short"} --- even in @Steenbergen2003's more exacting formulation --- and other such operationalizations "can be applied easily and reliably to a wide range of deliberative contexts" [@Steenbergen2003 22], but that very decontextualization makes it hard to adjudicate whether really, universal validity was discussed, let alone reciprocally acclaimed.
I share the uneasyness of "critical theorists and postmodernists" [@Steiner2012 11] with this measure of quality of deliberation --- though I do not think you have to be either to object, nor for the same reasons.[^postmodern_criticism_of_dqi]

[^postmodern_criticism_of_dqi]: [@King2009] criticizes that coded interpretations are presented as objective fact, rather than inter-subjective.

Rather than *how many* justifications were given in total, or *how often* people referred to counterarguments [@Steenbergen2003 29], I want to know whether these reasons corresponded in *substantive terms to those they disagree with*.
I provide a hypothetical exchange from deliberating taxation to illustrate my point:

<!-- TODO fix dialogue -->
**Foundationalist**
^[Inspired by the Philosophy of Robert [@Nozick1974].]
I believe that everyone should be entitled to incomes generated from uncoerced exchange with others.

**Social Democrat**
^[Not very inspired, and by no one in particular.]
Yeah, obviously we are not going to take away all of a person's income, they should be entitled to it, but maybe we'll take some of it because we need that for education.
That's not going to hurt them.
<!-- COMMENT MN social democrats/socialists will also always want to use taxation to disempower the powerful or the rich -->

**Georgist**
^[Inspired by the Philosophy of Henry [@George1879]]
I agree there may be something inviolable about owning the value added in free exchange with others.
But, Foundationalist, do you think that should also extend to owning unimproved land and natural resources?
It seems to me that *no one* has really created these values, and whoever reaps rents from them does so *only* because of the coercive powers that be.

Formally speaking and under [dqi]{acronym-label="dqi" acronym-form="singular+short"} coding, both Social Democrat and Georgist refer to Foundationalist's argument, yet their speech acts are of strikingly different quality.
Social Democrat, rather sloppily, reiterates the argument and proceeds to balance it with another good (education) and suggests a compromise.
We can see how the dialogue would proceed;
Foundationalist --- presumably deontologically bound to inviolable property --- would refuse to trade off property against opportunity.
<!-- %FIXME is maybe deontic the correct adjective some of the time? -->
Talking past one another, Social Democrat and Foundationalist may not have exchanged, let alone agreed on universal validity claims.
By contrast, Georgist --- maybe provisionally --- embraces the foundationalist's argument and points to an unexpected implication *in foundationalist* terms.
It is much harder to see how their dialogue would proceed;
presumably, Foundationalist would not --- ex ante --- have been open to large-scale land reform or taxation (as Georgism implies), and may not much warm to the idea.
Still, she may find it necessary to either clarify her position further, or to concede Georgist's point, because his may just be the "forcelessly forceful" better argument.
Foundationalist and Georgist are grappling with the formulations and implications of what *may* become a universally acceptable moral claim, and in so doing, have crept a little closer to communicative rationality.

Crucially, this kind of Habermasian adjudication cannot be done in the abstract;
researchers have to look for ideal speech in *substantive* terms as in this case, a foundationalist theory of property.

By abstracting away the actual arguments from a normative theory vested in their force, these researchers, as @Steiner2012 concedes, "take the position that large parts of the Habermasian version should be relaxed" [@Steiner2012 150].
For @Steiner2012, "the crucial point should be that participants in a political debate acknowledge that others also have reasonable arguments
\[...\]
If this is accomplished, the other side is humanized" [-@Steiner2012 150].
This may well be true and important research, especially for divided societies on which @Steiner2012 here reports, but it stretches the concept of *communicative action* too thin.
<!-- TODO refer to this is the kind of backstop argument for deliberation; it reduces the risk for civil war -->

To be sure, there may be such a loosely-defined thing as "deliberation" --- distinct from "debate"and "discussion" [@Landwehr2010 384] --- and we should study its antecedents and consequences, as much of the above-cited research does.
In this vein, deliberation raises positive questions:
What "middle-range theories" associate antecedents and consequences [@Mutz2008]?
To the extent that this deliberation is a normative project at all, it espouses a *consequentialist* outlook:
deliberation is good, because it leads to "more public-spirited attitudes;
more informed citizens;
greater understanding of sources of, or rationales behind, public disagreements;
a stronger sense of political efficacy;
willingness to compromise;
greater interest in political participation;
and, for some theorists, a binding consensus decision" [@Mutz2008 524].

Habermas prescription for communicative action, again, is different.
<!-- %COMMENT MN aber worüber wird bei H. delibieriert? über normen, nicht über policies -->
<!-- counter here: correct, but the only way to discuss norms in a meaningful way in the modern world is with regard to concrete policies; that's what the world offers us. Aalso, as we'll see, much of the below *is* actually about norms.-->
Communicative action is not a positive statement about causes and effects open to falsification, because it poses a *regulative* principle [@Kant1781], which "is unachievable in its full state, but remains an ideal to which, all else equal, a practice should be judged as approaching more or less closely" [@Mansbridge2010a 80].
Communicative action is also not a consequentialist ethic, but deontological, because "it has the force to legitimate" [@Habermas2008 147] --- or even teleological, because "reaching understanding is the inherent telos of human speech" [@Habermas-1984 287].
As such a normative political theory, communicative action cannot be operationalized into a set of treatments.
Just because you subtract power, inequality, disrespect [@Steenbergen2003], even "irrationality", or any other combination of procedural criteria from communication does not necessarily make it ideal speech --- even though the inverse may well be true, as these and other criteria may be *necessary* conditions for ideal speech.
Legitimating collective choice through ideal speech *is* the irreducible, final reason of @Habermas-1984' political theory whose achievement can only be found in substantive arguments.

Still, to thereby "simply define" deliberation as "intrinsically good" does not, as @Mutz2008 [524] suspects, render "empirical research irrelevant".
As @Steiner2012 [13] reminds us, the "level of deliberation" must be "measured and not merely assumed".
By definition, the normative impetus of communicative action cannot be falsified, but we may find that at any given time, any given people under any given circumstances, *fail* to meet Habermas' exacting standards [@Thompson2008 499].
Such failures can serve "as indicators of *contingent* constraints that deserve serious inquiry and \[...\] as detectors for the discovery of specific causes for existing lacks of legitimacy" (@Habermas2006 [420, emphasis in original], compare also @Thompson2008 [502]).
<!-- %Indeed, empirical researchers ``should focus on those features of the practice that directly relate to the fundamental problem deliberative theory is intended to address: In a state of disagreement, how can citizens reach a collective decision that is legitimate?'' \cite[502]{Thompson2008}. -->

If we accept this more demanding definition of deliberation as the venue in which our approximating of ideal speech is ascertained, as I here wish to do, our research designs must be focused on argumentative substance.
Others have done this, and have produced encouraging insights.
[@Ratner-2008-aa] interviewed [ca]{acronym-label="ca" acronym-form="singular+short"} members after the fact, and found that most members considered Habermas' validity claims honored --- especially compared to the proper tragedy that was the following referendum campaign in which many [ca]{acronym-label="ca" acronym-form="singular+short"} alumni participated.
While the analysis is based on interviews, not process data, it maintains the Habermasian standard and does not spare the reader qualitative data, including snippets recounting substantive disputes over electoral design.
[@GutmannThompson-2004-aa], too, reporting on deliberations over medical rationing, stick to their (quasi-Habermasian) standard, argumentative reciprocity, and offer examples of how it is met or violated in substantive arguments over health care.

Some may object that this approach collapses the procedural theory of democracy into a substantive theory of justice, and that it consequently requires researchers to take sides.
This neither need, nor ought to be so.
@Habermas-1984' *Communicative Action*, much like @Rawls-1971' contemporary *Theory of Justice* is neither purely procedural, nor substantive, but elegantly posits the *discursive conditions* under which substantive claims ought to be made.
The contrast between procedural and substantive theory, here, as elsewhere, is misleading:

> This kind of public philosophy would avoid the dichotomy that has come to dominate contemporary discussions of political theory, which poses a choice between basing politics on a comprehensive conception of the good, on the one hand, or limiting politics to a conception of procedural justice, on the other.
> We can and should avoid choosing either of these approaches exclusively.
> --- @GutmannThompson-2004-aa [90]

Consequently, researchers need not take sides, nor follow (dubiously legitimate) expert rule [@Blok2007; @Haas1992].
Emphatically, "alternative conceptions of the common good" [@Cohen-1989-aa 23] are possible and desirable.
Researchers must, however, familiarize themselves deeply with the issue to be deliberated, penetrate the discursive fluff and distortions to relentlessly unearth causal and moral disagreements which reasonable people may have over the issue.
Researchers can then analyze their rich, qualitative, ideally process data in terms of these expected disagreements, to assess in how far the observed deliberation lives up to the standards of communicative action.
Ideally, deliberators will struggle to formulate universal validity claims to express and resolve these very disagreements.
In short, researchers need not have a position on the issue, but they *must have a stake* in the abstractions, disagreements and uncertainties plaguing the issue.

Communicative action on taxation, in particular, must be analyzed in substantive terms for these reasons.
Tax is fraught with deep-seated causal and moral disagreement.
To test whether people in any given deliberation on tax really have exchanged universal validity claims to --- potentially --- resolve these disagreements, their actual communication must be compared to some ideal standard of expected arguments, tentatively identified by researchers.
<!--TODO  refer to validity-claims-tax table here -->

Table: Pluralist and Deliberative Democracy

|  | Pluralist Democracy | Deliberative Democracy |
|-----------------------|:---------------------------------------:|:---------------------------------:|
| *Knowledge* | Miracle of Aggregation Wisdom of Crowds | Schools as the Operative Metaphor |
| *Legitimacy* | Special Interests | (Alternative!) Common Good*s* |
| *Equal Participation* | Representation | Veil of Ignorance |
| *Speech* | Power | Communicative Action |
| *Preferences* | Pre-Social | Intersubjective |


<!-- NOTICE below 2 tables used to be one, but it is easier to read like this -->

Table: Hypothetical Examples of More or Less Valid Claims Concerning the Optimality of Taxation

|  | Comprehensibility | Truth | Truthfulness | Rightness |
|-----------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| `desirable-tax`, `tax-optimality` | ~~"A deadweight loss occurs if otherwise pareto-improving exchanges do not occur".~~"Taxes can sometimes prevent people from trading things on the market, which they otherwise would have exchanged to at least mutual benefit." | ~~"Taxes are anti-growth".~~ "Under some circumstances, taxes can depress market activity without raising revenue".  | ~~"Lower taxes benefit everyone, because growth trickles down."~~ "If and to the extent that people react to incentives, and that we can really compare their utility, perfect markets make some people better off without hurting others, compared to how things were before". | "~~No one would get any work done, if all their rewards get taxed away!"~~ "It seems that markets do a good job in solving some problems, and we can, to that extent and in those domains accept that people will also be influenced by threats and rewards |
| `doable-tax`, `tax-incidence` | ~~"The tax burden is distributed according to the relative price elasticities of sellers and buyers in any given taxed transaction."~~ "The ultimate burden of a tax may differ from who nominally pays it.,Taxes are almost always levied on some trade, and their ultimate burden is distributed among the traders according to how sensitive they are to price changes.,The less you can react to a price change, the more of a tax you bear." | ~~"With a Financial Transaction Tax (FTT), we can take back the money from those who caused the financial crisis."~~ "To the extent that a FTT discourages short-term trades, it does not raise any revenue.,To the extent that a FTT raises revenue, behavior does not change.,The burden of a FTT will be complexly determined by relative elasticities." | ~~"Employers already pay half of social insurance".~~ "Social insurance is a payroll tax and falls entirely on labor, no matter the nominal distribution.,The burdens are shared between employers and employees according to their relative price elasticities." | ~~"Big corporations should pay their fair share."~~ "Corporations are not moral subjects; they should not be liable for redistributive or general revenue taxes.,However, the rich people who hold a lot of stock in corporations should pay taxes on such income, or wealth." |


Table: Hypothetical Examples of More or Less Valid Claims Concerning the Incidence of Taxation

|  | Less Valid | More Valid |
|-------------------|---------------------------------------------------------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Comprehensibility | The tax burden is distributed according to the relative price elasticities of sellers and buyers in any given taxed transaction | The ultimate burden of a tax may differ from who nominally pays it.,Taxes are almost always levied on some trade, and their ultimate burden is distributed among the traders according to how sensitive they are to price changes.,The less you can react to a price change, the more of a tax you bear. |
| Truth | With a Financial Transaction Tax (FTT), we can take back the money from those who caused the financial crisis. | To the extent that a FTT discourages short-term trades, it does not raise any revenue.,To the extent that a FTT raises revenue, behavior does not change.,The burden of a FTT will be complexly determined by relative elasticities. |
| Truthfulness | Employers already pay half of social insurance | Social insurance is a payroll tax and falls entirely on labor, no matter the nominal distribution.,The burdens are shared between employers and employees according to their relative price elasticities. |
| Rightness | Big corporations should pay their fair share. | Corporations are not moral subjects; they should not be liable for redistributive or general revenue taxes. However, the rich people who hold a lot of stock in corporations should pay taxes on such income, or wealth. |


### Misunderstanding Taxation

Survey and experimental researchers have investigated how people misunderstand taxation and its abstractions.
[@McCafferyBaron2003], tracking the heuristics and biases research program [@KahnemanEtAl1982] suggest that voters fall short of @VonNeumannMorgenstern1944-rationality in their choice of tax:
they incorrectly believe that taxation has a trivial, "flypaper" incidence [@McCafferyBaron2004b], they favor (indirect) taxes not labeled or not visible, they fail to aggregate different taxes --- especially the proportional or regressive [payroll]{acronym-label="payroll" acronym-form="singular+short"} and [vat]{acronym-label="vat" acronym-form="singular+short"} --- confuse progression in absolute and percentage terms and inconsistently prefer bonuses over penalties.
Worried, [@McCafferyBaron2004] suspect that politicians may exploit these and other misunderstandings and that current tax systems may be suboptimal as a result of voter irrationality.
They also suggest that people may overestimate effective progression, and that tax regimes may, as a result, be less redistributive than voters intend and believe them to be [@McCafferyBaron2004].

In a similar vein, [@SausgruberTyran2011] show that people mistake the nominal incidence of a tax with its effective incidence on the buyers and sellers in a taxed market transaction.[^always_markets]

[^always_markets]: By definition, taxes in a market economy --- a pleonasm --- must *always* be levied as some potential market exchange occurs.
    This is straightforward for income and consumption taxes, where the earning and spending transactions are taxed, but also holds for wealth taxes and even other instruments of expropriation (such as inflation).
    Taxation always requires some real (or imputed) market valuation of the given tax base, which in turn necessitates an actual or hypothetical transaction between market participants.
<!-- %add reference to liquidity effects -->
In their laboratory experiment, participant buyers frequently tax sellers more than themselves, forgetting that --- under perfect competition, especially factor rent and price flexibility --- buyers and sellers will share any given tax on their transaction irrespective of the nominal incidence.[^tax-lse]

[^tax-lse]: This according to the [tax-lse]{acronym-label="tax-lse" acronym-form="singular+short"}.
<!-- %add reference to perfect competition conditions -->

Under the resulting "tax-shifting bias" [@SausgruberTyran2011], or, equivalently "flypaper theory of taxation" [@McCafferyBaron2003], participant buyers often choose suboptimally high taxes on sellers, which they end up paying (partly) themselves.
@SausgruberTyran2011 furthermore find that participants are less likely to tax-shift if they are given accurate information on tax burdens exogenously and if they have experienced its income-reducing effects in the laboratory marketplace.
Troublingly for deliberative democrats, they also find that (albeit minimally defined) initial deliberation does not reduce tax-shifting, but that it "makes initially held opinions more extreme rather than correct" [-@SausgruberTyran2011 164], corroborating concerns for (deliberative) group polarization [@Sunstein1991].

These are rigorously researched and disconcerting findings that have not received nearly the attention they deserve from welfare state scholars and other political economists.
These and other misunderstandings of tax may constitute partial, possibly sufficient --- but not necessary --- causes for changing tax regimes and constricted, or retrenched welfare states.
<!-- %note the absence of arbitrage mechanisms -->

These are also --- much like the sponsoring heuristics and biases program --- no definitive, comprehensive list of misunderstandings, but an evolving, loosely connected set of relatively low-level cognitive effects.
More research is needed:

-   how do people (mis)understand some of the broader and more complex issues bearing on taxation, such as economic growth, savings rates or government and market failures?
-   how --- if at all --- do the suggested low-level cognitive heuristics and biases filter up to the very structured choices of base, schedule and timing of taxation asked of the [oecd]{acronym-label="oecd" acronym-form="singular+short"} polities?

As these pressing questions are addressed to develop a coherent account of social change from misunderstanding taxation, research
designs as those by @McCafferyBaron2004 or @SausgruberTyran2011 may well face two kinds of limits:

1.  The complexity of higher-order considerations may overwhelm both designs.

    Laboratory models in behavioral economics as those by @SausgruberTyran2011 have the distinct advantage of directly testing the misunderstood causal effect in question (in this case, [tax-lse]{acronym-label="tax-lse" acronym-form="singular+short"}), and rendering participants with an immediate experience of the effect (here, lower incomes).
    As the causal effects in question become more complex --- for example, liquidity or employment decisions under different taxes --- laboratory models face ever harsher tradeoffs between internal and external validity [for a review and dissenting opinion, see @Jimenez-Buedo2010].
    Trivially, economy-wide choices --- for example, between a [pit]{acronym-label="pit" acronym-form="singular+short"} and [pct]{acronym-label="pct" acronym-form="singular+short"} --- are very hard to model in the laboratory;
    if they were any easier, economic policy would be a non-issue.

    The within-subject designs of cognitive psychology as those by @McCafferyBaron2004 also offer great internal validity when the heuristics and biases are low-level.
    Participants rate their agreement with, or the truth, of several statements they are presented with.
    Experimenters vary the context or wording of substantively identical statements to identify contradictions or mental shortcuts within the answers of individual participants.
    @McCafferyBaron2004 provide supposedly "de-biasing" information as within-subject treatments.
    The problem with higher-level concerns --- say, whether to tax income or consumption --- is that equivalent de-biasing treatments would eventually need to be an introductory economics course, covering at least the circular flow in the economy, some macroeconomics of aggregate demand and the Haig-Simons identity of income.
    As the required syllabus grows, it not only becomes impractical for participants with limited time and patience, but it also ignites a combinatorial explosion of necessary treatments to figure out just which (of supposedly many) insights from the de-biasing did the trick.

2.  More problematically yet for the deliberative-minded researcher, both designs cannot problematize, let alone resolve, any *remaining* substantive disagreements over taxation and its economic abstractions.
    Both behavioral economics and cognitive psychology stipulate that there is *a* ([vnm]{acronym-label="vnm" acronym-form="singular+short"}-)rationality, and merely chronicle whatever human deviations from it they find.
    Again as with the heuristics and biases research program in general, researchers and deciders must be aware of these deviations from rationality in tax, but these deviations do not add up to a theory, much less an account of social change in taxation or democracy.
    We know that @VonNeumannMorgenstern1944 say one (internally consistent) thing, and other homo sapiens mostly something else, but we learn little about the conditions of this divergence, let alone ways to reconcile it.

    Things get even thornier for a heuristics and biases research program of tax, when higher-level concerns are introduced and --- as they likely will --- turn out to be controversial.
    @McCafferyBaron2004 or @SausgruberTyran2011 posit expert rationality against their participants, and, whenever they differ, find their participants at fault.
    This works well enough for disentangling percentage and absolute tax burdens, and tolerably for demonstrating a tax-shifting bias[^flexible_prices],
    but leaves the researcher empty-handed when faced with genuine disagreement *between* experts, or challenges as the legitimacy of expertise.
    To be sure, there *may* well be *the* [vnm]{acronym-label="vnm" acronym-form="singular+short"}-rational, human-nature-compatible optimal taxation, that renders all other proposals regrettable misunderstandings --- but this hypothesized agreement and its communicative conditions must be tested, not axiomatized.
    The heuristics and biases research program can inspire a sociological account of misunderstood taxation, but it cannot bring it to its conclusion.
    When faced with inevitably controversial higher-level issues such as the choice between a [pit]{acronym-label="pit" acronym-form="singular+short"} and a [pct]{acronym-label="pct" acronym-form="singular+short"}, both cognitive psychology and behavioral economics will smack of expertocracy and remain fruitless because they allow no recourse to a second-order theory [@GutmannThompson-2004-aa 125] --- such as Habermasian deliberation --- to resolve disagreement.

[^flexible_prices]: Tellingly, @SausgruberTyran2011 already take pains to remind readers that [tax-lse]{acronym-label="tax-lse" acronym-form="singular+short"} only holds when factor rents and prices are fully flexible --- a condition that is as controversial as is it is difficult to verify, even after 60 years of macroeconomic debate about little else [@Wapshott2011].


## Research Questions

The deliberative experiments on taxation proposed here builds on these two strands of research as much as it probes its limits:


1. Taxation as a *Case*.
  <!-- does deliberation work -->
  A deliberative forum on taxation, including its structured choices and abstractions, adds another --- yet rare --- experiment on a technocratic and complex policy issue to the empirical record.
  From this perspective, deliberation is the research topic, and taxation is a *case* to test the limits and capacities of this prescription for political *rule*.[^tax_franzi]
2. Deliberation as *Method*.
  <!-- can it resolve the confusion and contradictions within tax -->
  Lastly, deliberating taxation will provide data to better understand possible misunderstandings of tax and, thereby, clarify remaining controversies over it.
  From this perspective, optimal taxation and popular deviations therefrom are the research topic, and deliberation provides the *method* to investigate, or even resolve these differences in a normatively attractive manner [@Rawls-1971; @Habermas-1984].
<!-- %add visualization of the above and below paragraph? -->

[^tax_franzi]: Taxation is a good case, because --- as Franziska Deutsch succinctly noted in 2011 --- it affects everyone, but most people know little about it.

These two research topics all fold into the aforementioned research question:
how will different people understand taxation differently, if they participate in a deliberative process *closer* to the ideals of communicative action than the status quo of representative democracy?


## Hypotheses

Both research questions, about 1) the remaining disagreements about tax and 2) the ability of deliberative democracy to cover abstract and structured issues can be rolled up into one set of hypotheses.

@Fishkin2009 suggests that deliberation increases knowledge, and changes preferences as a function of such knowledge gain.
Similarly, I hypothesize that people will gain in knowledge about taxation, including resolving some systematic misunderstandings of concepts and causal relationships, some variants of which are already documented [@McCafferyBaron2003; @Caplan2007].
I furthermore hypothesize that *as a function* of these knowledge gains, people will change their preferences on the desired base and schedule of taxation.
Additionally, deliberation will increase self-assessed autonomy and competence of participants, and they will find the presented alternatives of base and schedule more meaningful.
Learning interventions will further drive knowledge gain and improve self-assessments, but --- if it is sufficiently balanced --- *will not* interact with attitude change.
<!-- %Both research questions \ref{itm:resolve-better-tax} and \ref{itm:prove-deliberative-democracy} can be rolled up in one set of hypotheses: -->

1.  [\[itm:think-different\]]{#itm:think-different label="itm:think-different"}
    If ordinary citizens are given the possibility to deliberate welfare and taxation, they will think differently about it.
    Specifically,

    1.  [\[itm:knowledge-gain\]]{#itm:knowledge-gain label="itm:knowledge-gain"} *Knowledge Gain.*
        People will gain in knowledge.
        <!-- %COMMENT MN what if not? -->
        Specifically relevant for tax choice,
        <!-- %(\autoref{fig:tax-with-misunderstandings}, p.~\pageref{fig:tax-with-misunderstandings}), %requires that table to be in here -->

        1.  [\[itm:bastard-keynesianism\]]{#itm:bastard-keynesianism label="itm:bastard-keynesianism"} *Bastard Keynesianism.*
            People will understand that an economy can have an arbitrary (sub-@Solow1956) savings rate, and that --- equivalently --- if lowered slowly, aggregate *consumption* will not depress aggregate *demand*.

        2.  [\[itm:real-myopia\]]{#itm:real-myopia label="itm:real-myopia"} *Real Myopia.*
            <!-- %aka technocratic myopia -->
            People will understand that the future prosperity of an economy is, in part, determined by present *net* investment, and that --- equivalently --- *real*, not *nominal* (for example, [gdp]{acronym-label="gdp" acronym-form="singular+short"}) indicators track the savings rate of an economy.

        3.  [\[itm:bastard-hayekianism\]]{#itm:bastard-hayekianism label="itm:bastard-hayekianism"} *Bastard Hayekianism.*
            <!-- %aka government vs markets
            %make naming consistent -->
            People will understand that an economy can adopt an arbitrary government quota (if, possibly, at a cost), and that --- equivalently --- not *all* (but some) taxed resources are lost.
            Taxation is not exclusively or unconditionally a negative-sum proposition.

        4.  [\[itm:flyper-theory\]]{#itm:flyper-theory label="itm:flyper-theory"} *Flypaper Theory.*
            <!-- %aka tax-aversion -->
            People will understand that ontologically and empirically, only and always *natural* persons bear the burdens of taxation, and that --- equivalently --- the flypaper theory of tax incidence is false.

            This hypothesis is similar to, but not identical to @McCafferyBaron2003's findings of *tax aversion*, by which people mistakenly prefer taxes not labeled as such and/or indirect taxes and a *disaggregation effect*, by which people fail to integrate taxes on same bases.
            <!-- %might have to feature tax aversion separately, because that is already in the visualizatiuon -->

        5.  [\[itm:ordoliberal-mess\]]{#itm:ordoliberal-mess label="itm:ordoliberal-mess"} *Ordoliberal Mess.*
            <!-- %not hygiene, that's the opposite -->
            People will understand that Pigouvian and general-revenue taxes follow opposing logics, and that --- equivalently --- a Pigouvian tax *should not* raise revenue.

            This hypothesis is similar to, but not identical to @McCafferyBaron2003's findings of a *disaggregation effect*, by which people fail to integrate different taxes even on same bases.
            <!-- %this is (\cite{McCaffery2003}: 19ff)
      			%is this really right? -->

        6.  [\[itm:false-omnipotence\]]{#itm:false-omnipotence label="itm:false-omnipotence"} *False Omnipotence.*
            <!-- %clarify labeling
  			    %used to be negative some -->
            People will understand that taxation will often (if not always) cause unintended consequences in markets, or that --- equivalently --- taxation can cause a negative-sum [dwls]{acronym-label="dwl" acronym-form="plural+short"}.
            Taxation is not exclusively or unconditionally a zero-sum proposition.

        7.  [\[itm:false-omniscience\]]{#itm:false-omniscience label="itm:false-omniscience"} *False Omniscience.*
            Relatedly, people will understand that taxable income and wealth (but not consumption) cannot be measured independent of market-pricing, or that --- equivalently --- any taxation based on fiat evaluations may have unintended consequences in markets.

    2.  [\[itm:universal-validity\]]{#itm:universal-validity label="itm:universal-validity"} *Universal Validity*
        People will find the above abstractions meaningful and accept a universal validity of underlying causal and moral arguments.

        <!-- %			\item \label{itm:preference-structuration}
        %				\emph{Preference structuration.}
        %				Partly as a function of the above knowledge gain, people will have better-structured preferences over taxation.

        %				Specifically, and related to the cyclical preferences \citep[334]{Farrar2010} and other aggregation failures of pluralist democracy \citep[for example,][]{Condorcet1785,Arrow1963},

        %				\begin{enumerate}

        %					\item \label{itm:vNM-consistent}
        %						\emph{\glsfirst{vnm}-Consistency.}
        %						\emph{Individually}, people will have ordinal preferences that more closely resemble \gls{vnm}-consistency.

        %					\item \label{itm:single-peakedness}
        %						\emph{Single Peakedness.}
        %						\emph{In the aggregate}, people will have preferences that more closely resemble single-peakedness.

        %					\item \label{itm:orthogonal-dimensions}
        %						\emph{Orthogonal Dimensions.}
        %						\emph{In the aggregate}, people will have preferences that more closely resemble orthogonal factors.
        %				\end{enumerate} -->

    3.  [\[itm:attitude-change\]]{#itm:attitude-change label="itm:attitude-change"} *Attitude Change.*
        Partly as a function of the above effects
        people will prefer a PCT, conditionally supplemented by a WT, LVT and NIT over the present tax regimes of PIT, VAT and Payroll.
        <!-- %(as in \autoref{fig:tax-with-misunderstandings}, p.~\pageref{fig:tax-with-misunderstandings}), %not working -->

2.  [\[itm:meta-assessment\]]{#itm:meta-assessment label="itm:meta-assessment"} *Meta Assessment*

    1.  [\[itm:autonomy\]]{#itm:autonomy label="itm:autonomy"} *Autonomy.*
        People will consider themselves more effectively autonomous in discussing and choosing basic tax regimes.

    2.  [\[itm:competence\]]{#itm:competence label="itm:competence"} *Competence.*
        People will consider themselves more competent in discussing and choosing basic tax regimes.

    3.  [\[itm:meaningful-choice\]]{#itm:meaningful-choice label="itm:meaningful-choice"} *Meaningful Choice.*
        People will find the choices of tax base and schedule to be more meaningful to express their political autonomy.

3.  [\[itm:learning-interventions\]]{#itm:learning-interventions label="itm:learning-interventions"} *Learning Interventions.*
    A learning interaction targeted at one of the above-listed knowledge gains will further that knowledge gain.

4.  [\[itm:interaction-effects\]]{#itm:interaction-effects label="itm:interaction-effects"} *Interaction Effects.*
    The above effects will:

    1.  [\[itm:interact-equity\]]{#itm:interact-equity label="itm:interact-equity"}
        *Not interact* with people's equity beliefs, or --- equivalently --- people will not change their thinking about tax as a function of their allocative preferences.

    2.  [\[itm:interact-ses\]]{#itm:interact-ses label="itm:interact-ses"}
        *Interact negatively* with people's socio-economic status, or --- equivalently --- people with greater socio-economic status will change their thinking about tax the least.
        <!-- %				\item \label{itm:interact-cognitive-ability}
        %					\emph{Interact negatively} with people's cognitive ability, or --- equivalently --- people with less cognitive ability will change their thinking about tax the least. -->

    3.  [\[itm:interventions\]]{#itm:interventions label="itm:interventions"}
        *Intervention-interaction*

        1.  Of the above effects, meta assessments will *interact positively* with learning interventions.

        2.  Of the above effects, attitude changes will *not interact* with learning interventions over and above the effect mediated by knowledge gain.

        <!-- %\begin{landscape}
        % \begin{figure}[htbp]
        %    \begin{center}
        %	\includegraphics[width=1\linewidth]{tax-with-misunderstandings}
        %	\caption[The Vector Field of Taxes with Misunderstandings]{The Vector Field of Taxes with Misunderstandings}
        %	\label{fig:tax-with-misunderstandings}
        %	\end{center}
        %%	\input{./tex/tax-table-description.tex}
        %\end{figure} -->


Table: Variable Metaconsensus in Taxation (x3)

|  | Values | Beliefs | Preferences |
|-------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------|
| Subject | What makes a tax desirable? | What makes a tax doable? | Which tax do we want? |
| Existing Work | `axiology`, `desirable-tax` | `ontology`, `doable-tax` | `better-tax` |
| Example Statements |  *"People should pay taxes on what they earn on their wealth"* | *"A tax can be easily collected right at the source, in the firm"* | *"We should tax the income of corporations."* |
| Example Extracted Factors | Efficiency vs. Equity | Withholding Taxation | Corporate Income Taxation |
| Meta-Consensus | High(er) loadings on relatively few(er) factors. People understand that people can disagree on efficiency vs. equity. | High(er) loadings on relatively few(er) factors. People understand that one of the choices would be to tax at the source. |  High(er) loadings on relatively few(er) factors. People understand different taxes, and recognize it in different formulations. |
| Example Extracted Secondary Factor | Indirect Taxation (all columns) |  |  |
| Intersubjective Rationality | People who value wealth taxation, and think withholding taxes are easier to collect will prefer a Corporate Income Tax. (all columns) |  |  |
| Delta Before and After Deliberation | Tax Misunderstanding: Flypaper Theory (all columns) |  |  |


Table: Variable Metaconsensus in Taxation (x5)

|  | Values (specific) | Values (General) | Beliefs (General) | Beliefs (Specific) | Preferences |
|---------------------------|------------------------|----------------------------------------------------------------|----------------------------------------------------------------------|-------------------------------------------------------------|------------------------------------------------|
| Subject | What is desirable? | What makes a tax desirable? | What is doable? | What makes a tax doable? | Which tax do we want? |
| Existing Work | `axiology` | `desirable-tax` | `ontology` | `doable-tax` | `better-tax` |
| Example Statements | *"Growth is good"* | *"A good tax should encourage people to save for the future."* | *"Markets are a reasonable way to organize much of human activity."* | *"It is more feasible to tax things whose value is known."* | *"People should be taxed on what they spend."* |
| Example Extracted Factors | Consequentialist Ethic | Time-Discounted Stable Incomes | Pro-Market | Liquidity Taxation | Progressive Consumption Tax |


<!-- %%Cite Eliasoph on why we need a big ass issue; not a small scale issue.
%You're leading to despondence.
%Quote early passages of Eliasoph.
%Also argue with Eliasoph "Charles said …" that by all means trying to CHANGE things, and only discuss things you can change, has drawbacks.
%
%%Note how the DP on energy -- which, at first, seems similar -- is actually not.
%It asked, among other things, whether people would be willing to pay more on their monthly utility bills for wind \citep[loc.
%2013]{Fishkin2009}, but not more technical stuff than that.
%
%%Rosenberg 2007: Rosenberg has the radical inter-subjective formulation that I think I'm looking for:
%	%6: ``In this deliberative conception, equality and autonomy require each other.
%On the one hand, equality is a necessary precondition of autonomy.
%It is only in a cooperative exchange between equals that the self-expression and critical selfreflection required for the self-reflective construction of one’s understandings and interests is possible. -->
<!-- %%25: ``the potential for abuse [by facilitators] is real, and crafting an appropriate conceptual and institutional response will be difficult'' -->


<!-- A fitting deliberative format for my project...

1. Includes a strong learning component, structured roughly along my drafts -- but shorter, and in more accessible style.
	(It must not be a microeconomics text).
2. Relates these abstractions to personal experiences and intersubjective value judgments.
3. Prioritizes intensity (duration) over representativeness (sample size) under given constraints (time, money).
4. May involve me and existing material for the learning component, but someone else must moderate the deliberations.
5. Starts from a small-scale, face-to-face deliberation as a pilot study, both to guarantee concept validity and because I have no idea about deliberation- -->


<!-- #### Threats

1. SES-graded, unequal participation may remain/result, especially because tax tends towards academic abstractions.
2. Political apathy may remain/result, especially because fundamental tax reform cannot even be done at the national level and may face other hurdles.
	Participants will in no way have some actual decision-making power over the subject of their deliberation.
3. The deliberation might regress into a microeconomics class, where deliberators look for “right” or “wrong”, rather than empower their own voices and intersubjective understandings. -->


## Public Finance and Welfare State Research

If the above hypotheses (especially [\[itm:attitude-change\]](#itm:attitude-change){reference-type="ref" reference="itm:attitude-change"}) are confirmed, welfare research has a lot more to explain.
If under a legitimate democratic process, people resolve remaining *a priori* disagreements to select a *different* tax than we currently observe, welfare state research must provide a second-order theory of the observed (illegitimate) social choice of suboptimal taxation.
I cannot provide, let alone test that second-order theory of tax choice in this dissertation, but --- if hypotheses are confirmed --- can at least suggest that the democratic forum may be important.
Pluralism may not be the *only* reason why we have no better tax, but it might well be one of several reasons.


## Public Choice and Political Economy

Public choice has long specified how individual preferences can be aggregated, what it requires and how it can fail [for an overview, @Mueller].
Conversely, (classical?) political economy has formulated theories of how (material?) interest and power can corrupt and alter preference aggregation [for an overview, @Robbins1976].
Both those fields, too, can benefit from deliberating hypothetical taxes.

Public choice, for once, is mostly an *a priori* and often an insular enterprise, merely chronicling --- but not explaining --- the inconsistent preferences and structuration anomalies it faces.
For example, portfolio theory clearly differs from observed loss (not risk!) averse strategies [@Kahneman2011], but we know relatively little about how culture or institutions mediate this gap.
Deliberation is one (possibly attractive!) institution to narrow the gap between *a priori* public choice and *a posteriori* policy preferences.
The hypotheses tested in this [dp]{acronym-label="dp" acronym-form="singular+short"} on taxation problematize the institutional link between optimal and actual choice.

Public choice is also hard to bring to bear on real-life political choices, because it deals in such high abstractions and the very consistency in preferences it demands are often subject to (second order) controversy.
It can be difficult to argue how any particular policy may be explained by, or differ from a theory of public choice, as long as there is disagreement over possible alternatives and ontological concepts.
This dissertation synthesizes and develops consistent preferences on taxation, and puts them to a deliberative test.
If successful, it may provide an insightful, practically relevant case study of policy and help resolve the controversy over preferences.

Conversely, (classical) political economy can gain from a [dp]{acronym-label="dp" acronym-form="singular+short"} on taxation, because it, too, facing disagreement over alternatives and material ontologies (@Keynes1936 vs @Hayek1931, as in @Wapshott2011), too often reverts to second-order theory, sometimes getting lost in an unproductive muddle of ideas *and* interests, rather than one *or* the other as independent variable.
[^2]
This dissertation develops a --- hopefully --- agreeable material ontology for the mixed economy (), and synthesizes a desirable and doable alternative which is then subjected to a specific, legitimate second-order process:
deliberation.
By comparing cognition --- or ideas (!) --- as a *function* of this second-order process, and by relating it to socio-economic outcomes and conditions of deliberators, this dissertation also tests one specific theory of the political economy.
Based on transparent, indeed quite orthodox assumptions, I hope to show that material interest may, through whichever agency or structure, have found a clever way to corrupt democratic rule without resorting to crude power.
If confirmed, my hypothesis suggest that --- for whatever reason --- people systematically misunderstand tax in a way that favors the rich, and more generally, the benefactors of the status quo.

## Conclusion

In Schumpeter's
thundering words, taxation and democracy *are* the two sides of social contract (, p. ).
They, above much else, balance contradicting social integration and individuality in modernity and moderate its key antagonistic institutions:
states and markets.

\centering
![Taxation and Democracy[]{label="fig:tax-democracy"}](tax-democracy){#fig:tax-democracy width="1\linewidth"}

Surely, such a broadly defined research agenda risks hubris, but it also promises to formulate and maybe, tentatively answer some important questions for the social sciences.
I may well --- figuratively --- be biting off more than I can chew, and it sometimes shows (for example in ).
All I can do is to tentatively mull over a mouthful at a time, and spit out again that which I cannot address.
This --- forgive the metaphor --- *rumination* takes a lot of pages to write and read, but at least, it is a relatively transparent way of reasoning.
Maybe, in the spirit of Umberto Eco's *anti-library* [as recounted in @Taleb2007], explicating what I do not know or cannot address in this dissertation will make for better questions.
[^3]
Still, to digest anything in full, I must concentrate.
Counterintuitively, the often economic or even technical abstractions of taxation and public choice provide a great guide here.
This mostly *a priori* knowledge, too, requires a lot of pages to write and to read, but, akin to enzymes in the digestive tract, it helps me break down complex issues to ever simpler, fewer and clearer questions (as in and ) that can be handily metabolized into a new, and hopefully original argument that I put to the test.

Admittedly, too, underlying this research are normative convictions about political and economic opportunity, and even human destiny.
Somehow, such a prescriptive impetus has become unacceptable in much of (German) social science on the welfare state.
I must indeed avoid "politics-based evidence making" (The Economist, 2012)
or any other conflation of what should be, with what is.
However, even from a merely positive perspective, and even to readers who may not agree with these convictions, the hypotheticals they raise pose analytically worthy questions:
If there is indeed another, superior fiscal configuration on which we could democratically agree --- but do not --- this absence needs explaining.

If anything, finding that which needs a social explanation, must be a worthwhile attempt for the social sciences.

It, too, is how social sciences can contribute to human progress, by *politicizing* the status quo.

> "Politicization is the realization that established social norms, social practices, and social relations are contingent rather than sacrosanct, that things could also be different and that citizens, individual and collectively, have political agency by means of which alternatives can be explored and implemented.
> This recognition that things could also be different has always been the igniting spark of the emancipatory-progressive movements, and politicization has always been their key strategy."\
> --- @Bluhdorn-2007-aa [313]


[^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.

<!-- %Second Order Questions of Social Change
%The second order question of social change is simply:
%Why don’t we have it?
%Possible answers (hypotheses) are:
% Progressive Taxation is subject to an international cooperation problem, akin to a Prisoner’s Dilemma.
%Because the PCT is progressive, it is individually rational for states not to implement it, even if it were collectively rational.
% This is a very interesting question that could potentially greatly enhance our understanding of the international political economy of taxation.
%It however, also requires a comprehensive model of the economy, and even international trade and finance to anticipate the hypothetical payoff of the PCT.
% Tax regimes are subject to great path dependency, they cannot be easily reformed.
% While a historical and tentative analysis of path dependency in tax regimes is possible and probably productive, the costs of reform can ultimately only assessed when a cardinal specification and implementation has been undertaken.
%Assessing the costs of the reform ex ante may be very hard, unreliable or impossible.
%Both a possible international cooperation problem and path dependency in tax regime choice are ultimately endogenous to broader questions of the political economy.
%The cooperation problem depends on the level at which the international political economy is governed.
%Path dependency depends on the domestic and international incidence of reform costs, in turn depending on the states ability to hand out side-payments.

% Ultimately, the PCT could be absent because people do not want it.
%The second order question of social change in tax regimes is:
%will people reject the PCT irrespective of the democratic process that governs their polity?
%If not:
%how do different democratic processes (pluralist, deliberative) constrain or even determine the politics and choice of tax?

%likewise:
%assumed theoretical link:
%if you want to hide injustice, than you'd best do it in a really complex system, like tax.
%
%\subsubsection{Research Question b):
%Why Hasn't it Been Implemented Yet?} Research question b) is then, based on the developed ``perfect tax'', more approximately empirical.
%In the absence of it being implemented in any case, it asks why it has not.
%
%The usual suspect here is a lack of international cooperation.
%Its plausibility will be compared to that of other explanations, namely:
%\marginpar{\textsf{This section needs an overhaul, and \emph{a lot} more reading to make the alternative explanations MECE.}}
%\begin{description}
%	\item{\textbf{Ideology}} Lack of support for such a broadly and boldly redistributive scheme.
%	\item{\textbf{Path Dependency or Historical Institutionalism}} Typological work applies here --- in how far is the tax compatible or not with the institutions and mindsets of established types, for example \cite{BeramendiRueda-2007-aa}.
%	\item{\textbf{Deficient Political Process, Public Deliberation or Discursive Institutionalism}} Literatures on the political process, as well as, on discursive institutionalism \cite{Cerami2009a}.
%	\item{\textbf{International Cooperation Problem}} See above.
%\end{description}

%or addition to science

%First, developed welfare states, in spite of their unprecedented prosperity, appear to be increasingly constrained in their ability to raise revenue and redistribute it as their legislators see fit.
%The pertinent literature on globalization and the welfare state (for an overview see Genschel 2004) asks whether the welfare state has retrenched, and for which reasons.
%Frequently operationalized only as spending (Swank 2005) or decommodification (cf.~Esping-Andersen 1990), that question is incomplete.
%Some of the latently affirmative analyses (Swank 2005) seem to be not so much optimistic about what the welfare state can do, as they seem to be minimal about what it should do:
%income replacement.
%To critically appraise the development of the welfare state, we must instead ask whether legislators are able to raise revenue and redistribute while maintaining growth, factor clearing, low inflation and a positive savings rate.
%We must ask whether legislators were better able to strike this balance of “opportunity and prosperity” (Obama for America, 2007) in the past, or whether there could be a better balance, and then investigate why we do not have better policy.
%Today, welfare states cannot have the cake and eat it.
%They can’t even do one of the two.
%Instead, welfare states are faced with a twin crisis that is mutually reinforcing:
%one of structural unemployment, and one of structural underfunding.
%Structural unemployment is caused by high effective price floors for labor, at or below the socially acceptable minimum income (transfers, minimum wage) in a society, but above the labor productivities of a segment of the workforce.
%Price floors are further raised by an increasingly proportional (or even regressive) tax falling on labor incomes.
%Structural underfunding is caused by a state that is increasingly unable to raise sufficient revenue necessary for the provision of risk pools, public and common goods, as well as to finance its allocative goals.

%Second, my research interest emerges from a sense of utter disconnect between political debate and the abstractions and interests governing the political economy.
%This dis-connect is evident in misleading sloganeering (‘Mehr Netto vom Brutto’, more net out of gross income), widespread superstitions (‘employers [sic!] pay half of social insurance’), bastard Keynesianism (‘consumption is good!’) or redistributive smoke grenades (‘we should tax companies’).
%At a deeper level, I wonder whether pluralist interest and electoral re-presentation can still be reasonably assumed to yield efficient and equitable policies, in an ever more complex world marred by cooperation problems.

%I don't explain the elite-based (potential) reason for the non-reform.
%That, the second-order theory, would be someone elses job.
%discourse theory here? -->

<!-- %There is a vibrant and innovative community of scholars and activists who host and moderate deliberative fora of varying designs, particularly in the US.
%Organizing successful deliberations is at least as much an art as a science and requires extensive experience.
%In addition, I must prevent my analytical and normative preoccupation with a particular reform proposal (a postpaid PCT) from exerting undue influence on the deliberation.
%The deliberative must not be set up to confirm the PCT.
%Both hypothesis a) and b) must be falsifiable in this design.
%A postpaid PCT must be presented as one of the (few) fundamental choices in redistributive design.
%My prior work (Held 2010) on the postpaid PCT can serve as a survey of the trade-offs and desiderata (see summary above) in redistributive design.
%These may serve to organize the expected transformation of previously “amalgamated” (Miller 1992:
%4ff) preferences into distinct, single-peaked dimensions.
% The trade-offs and desiderata are also inputs for questionnaire design and may be used to draft briefing materials.
%This quasi-experimental design requires great methodological care and experience.
% Ideally,  I will be able to co-host, or investigate a Deliberative Poll ® on redistributive design in Germany or the US, with help from James S.
%Fishkin and the Stanford-based Center for Deliberative Democracy.
% This would ensure rigorous institutional design, peer-reviewed balance and experienced implementation, but requires substantial external (to BIGSSS) funding.
%Should I be unable to secure such funding, some scaled-down version will also be possible. -->