Named literature: Wicksell, Finanztheoretische Untersuchungen (1896). Arrow, Social Choice and Individual Values (1951). Black, "On the Rationale of Group Decision-Making" (1948); The Theory of Committees and Elections (1958). Downs, An Economic Theory of Democracy (1957). Buchanan and Tullock, The Calculus of Consent (1962). Olson, The Logic of Collective Action (1965); The Rise and Decline of Nations (1982). Tullock, "The Welfare Costs of Tariffs, Monopolies, and Theft" (1967). Demsetz, "Information and Efficiency: Another Viewpoint" (1969). Niskanen, Bureaucracy and Representative Government (1971). Stigler, "The Theory of Economic Regulation" (1971). Krueger, "The Political Economy of the Rent-Seeking Society" (1974). Nordhaus, "The Political Business Cycle" (1975). Buchanan and Wagner, Democracy in Deficit (1977). Brennan and Buchanan, The Power to Tax (1980). Shepsle and Weingast, "Structure-Induced Equilibrium and Legislative Choice" (1981). Acemoglu and Robinson, Why Nations Fail (2012). Persson and Tabellini, Political Economics (2000).
The previous chapter showed where strict rationality breaks down: real people are loss-averse, present-biased, and frame-dependent, and the rational-actor model survives as a benchmark rather than a description. This chapter runs the opposite experiment. It takes the rational, self-interested agent at full strength and walks it out of the marketplace and into the voting booth, the legislature, and the government bureau — and the analytical payoff turns out to be large. Where behavioral economics found the rational actor too strong a description of individual choosers, public choice found it had been applied too narrowly: economics modeled buyers and sellers as self-interested maximizers and then, at the boundary of the market, switched to modeling the state as a benevolent planner who wants only the social good. Public choice asked what happens when that switch is refused.
The move is best stated in James Buchanan's own phrase: politics without romance. The romantic theory of the state treats public officials as a class apart — legislators who legislate for the public interest, regulators who regulate for the common good, bureaucrats who administer without an agenda of their own. The economist's theory of everything else treats people as agents who respond to the incentives they face. Public choice is what you get when the second theory is extended over the first: politicians are people who want to win elections, bureaucrats are people who want larger budgets and easier lives, voters are people who weigh the cost of being informed against the negligible chance their vote decides anything, and interest groups are people who will spend real resources to capture a favor from the state. None of this asserts that public actors are venal. It asserts only that they are the same kind of agent economics models everywhere else, and that the analytical consequences of that assumption are worth taking seriously.
Public choice is the application of economic method — rational self-interest, methodological individualism, and exchange — to political and collective decision-making: to voters, legislators, bureaucrats, and the rules that bind them. The phrase to fix is its contrast with social choice. Social choice, the tradition that runs from Kenneth Arrow's Social Choice and Individual Values (1951), asks how a society ought to aggregate the preferences of its members into a single collective ranking — it is the search for a well-behaved social-welfare function. Public choice rejects that framing at the root. There is no social mind doing the choosing and no social-welfare function waiting to be discovered; there are only individuals, each pursuing private ends, who under some sets of rules reach mutually beneficial agreements and under others do not. Politics, on this view, is not the aggregation of preferences into a collective will. It is exchange — the same category economics uses for the market, now applied to the trades, bargains, logrolls, and constitutional compacts through which individuals pursue gains from collective action. The aggregation paradigm asks what the right social ranking is; the exchange paradigm asks what bargains rational individuals will strike, and under what rules those bargains leave everyone better off.
The substrate public choice diverged from is worth seeing, because the divergence is the founding act. Arrow's impossibility theorem (1951) established that no rule for aggregating individual preferences into a consistent social ordering can satisfy a small set of reasonable conditions at once — no dictatorship, unanimity respected, choices unaffected by irrelevant alternatives, all preference profiles admitted. The aggregation project, pursued at full rigor, runs into a wall. Duncan Black's median-voter result (1948, 1958) is the partial escape: if voter preferences over a single issue dimension are single-peaked — each voter has one most-preferred point and likes options less the farther they sit from it — then majority rule selects the position of the median voter, and the impossibility is dodged. Anthony Downs's An Economic Theory of Democracy (1957) turned that geometry into a theory of party competition: two vote-seeking parties, each wanting only to win, converge toward the median voter, because any platform away from the median can be beaten by a rival that moves toward it. The rational voter, the convergence of parties, and the fragility of the result when preferences stop being single-peaked are the apparatus public choice inherited — and the interactive below lets you drive each piece.
Downs supplied something else the tradition would build on: the recognition that the rational voter faces a problem the romantic theory never noticed. Voting is costly — it takes time to register, to inform oneself, to turn out — and the benefit a single vote confers is its chance of being decisive, which in any large electorate is vanishingly small. The strictly rational citizen, weighing the cost of voting against the negligible probability of casting the deciding ballot, might rationally not vote at all, and would certainly not invest heavily in becoming informed. This is the famous paradox of voting (that anyone rational votes) and its companion, rational ignorance (that rational voters stay uninformed), and both fall directly out of taking the self-interested-agent model seriously in the polling booth. They are not cynical observations about apathy; they are predictions of the same model that predicts the convergence of parties, and they matter because an electorate that is rationally inattentive is an electorate that organized, concentrated interests can work around — which is the thread the next two sections pick up. The point to carry forward is that the apparatus is unified: the voter, the party, the legislator, and the interest group are all the same maximizing agent, and the surprising political results come from following that one assumption into each role.
Two parties, each wanting only to win, chase the votes. Press Run convergence and watch both platforms march toward the median voter — the Downsian result. Then flip the issue space to two dimensions, or make the preference distribution non-single-peaked, and watch convergence break: the platforms orbit and never settle, because outside the single-peaked one-dimensional case there is no median for them to find. The fragility is the lesson, and you produce it yourself.
Figure 14.1 (interactive). Two vote-seeking parties iterating toward the electorate's median under single-peaked one-dimensional preferences — and failing to settle in two dimensions or under polarized preferences. Run the convergence; flip the dimension and the preference shape and watch the median disappear.
A party away from the median can always be beaten by a rival that steps toward it, so in a single-issue contest both parties march to the middle and stop. That is why two-party systems feel like they converge. The result is brittle: add a second issue and there is no single middle to converge on — any platform can be beaten by some other — so the chase never ends. Single-peakedness on one dimension is doing all the work, and the moment you remove it the tidy median result collapses into the cycling Arrow warned about.
The Arrow impossibility is worth meeting directly, because it is the wall the aggregation paradigm runs into and the reason Buchanan walked away from it. Set three voters against three options and let majority rule decide each pairwise contest, and the social ordering can cycle: a majority prefers A to B, another prefers B to C, and a third prefers C to A, so there is no option a majority would not overturn. This is the Condorcet paradox, the simplest case of Arrow's general result — majority rule, applied to certain preference profiles, produces no consistent collective ranking at all. The explorer below lets you build the cycle from rankings you choose, then find the profiles (the single-peaked ones) that escape it.
Set three voters' rankings over three options {A, B, C}, then step the pairwise majority votes. With the default profile a Condorcet cycle closes — A beats B, B beats C, C beats A — and no option commands a stable majority. Press Find a stable profile and the rankings reset to a single-peaked configuration where a Condorcet winner exists. The impossibility is something you build, not something you are told.
Figure 14.2 (interactive). Three voters, three options, and the three pairwise majority contests. The default profile cycles (no Condorcet winner); single-peaked profiles produce a winner. Choose the rankings; step the votes; watch the cycle close or break.
Majority rule is supposed to deliver "what the group wants," but on the wrong preference profile it delivers a loop: whatever wins can be beaten by something else, forever. There is no social-welfare function hiding behind the votes for the rule to reveal. That emptiness — the absence of a coherent collective will to aggregate — is exactly what convinced Buchanan to stop asking what society ought to rank and start asking what bargains individuals will strike under the rules they agree to live by.
With the social-choice substrate behind it, the founding text can be read for what it built rather than what it rejected. James Buchanan and Gordon Tullock's The Calculus of Consent (1962) is the constitutional charter of the tradition. Its central move is the two-stage analysis at the heart of constitutional political economy. There is a difference, the authors insist, between choosing the rules of the political game and choosing within them. The choice of rules — the constitution, the voting threshold, the division of powers — happens at the constitutional stage, behind a veil of uncertainty about which particular interests one will later have. The choice of policies under fixed rules happens at the post-constitutional stage, where interests are known and the bargaining is over spoils. The two stages call for different standards. At the constitutional stage, where no one yet knows whose ox will be gored, the benchmark is unanimity: a rule everyone would consent to is a rule that, by construction, makes no one worse off than the alternative of no rule at all. This is the Wicksellian root the authors openly acknowledge — Knut Wicksell's 1896 insight that a tax-and-spending decision is genuinely justified only if those who pay would voluntarily agree to it, which makes approximate unanimity the test of a legitimate fiscal bargain.
Why, then, is unanimity not the operating rule? Because reaching it is ruinously expensive. The Calculus of Consent's most durable analytical contribution is the trade-off between two kinds of cost that any voting rule incurs. The external costs of a rule are the losses imposed on those who are outvoted — the harm a decision does to the dissenters who must live with it. As a rule moves toward unanimity, external costs fall, because fewer and fewer people can be forced to bear a decision they reject; at full unanimity external costs vanish, because nothing passes without everyone's consent. The decision costs of a rule are the resources spent reaching agreement — the bargaining, the holdouts, the time. These rise as a rule moves toward unanimity, because each additional consent that must be secured gives one more person the power to extract concessions, and the last holdout can hold the whole bargain hostage. The optimal voting rule for a class of decisions minimizes the sum of the two. Unanimity is the benchmark precisely because it zeroes out external cost; it is not the operating rule precisely because it maximizes decision cost. The interactive below lets you watch the optimum move as the two costs trade off.
Slide the required-majority fraction from simple majority toward unanimity and watch the two cost curves trade off: external cost (the harm to those outvoted) falls, decision cost (the cost of reaching agreement) rises, and the U-shaped sum locates the optimal rule at its minimum. Then change the kind of decision — how costly dissent is, how costly bargaining is — and watch the optimum move: a constitutional rule lands near unanimity, a routine rule near simple majority. That the best rule varies by decision class is the Calculus of Consent's point.
Figure 14.3 (interactive). External cost (falling in $k$), decision cost (rising in $k$), and their U-shaped sum, with the optimal required-majority $k^*$ at the minimum. Slide $k$; change the decision class and watch $k^*$ move from simple majority toward unanimity.
Unanimity sounds ideal — nobody is ever forced to accept a decision they reject — but the price is that any one person can block everything, so nothing gets done. Simple majority is cheap to operate but lets a bare majority impose serious harm on the rest. The right rule is neither: it depends on what is at stake. Constitutional questions, where being outvoted is catastrophic, warrant near-unanimity; routine questions, where the harm of losing is small and the cost of haggling is the real worry, warrant simple majority. There is no single best voting rule, only a best rule for a class of decision.
Write external cost as $E(k) = a(1-k)^2$ (falling to zero at $k=1$, where no one can be outvoted) and decision cost as $D(k) = b\,k^2$ (rising convexly in $k$, since each additional consent that must be secured is harder to get than the last). The interest cost of the rule is $C(k) = E(k) + D(k)$, minimized at the interior $k^*$.
Minimizing $C(k)$ gives the clean result $k^* = a/(a+b)$ (clamped to the $[0.5, 1]$ range of feasible majority rules). The slider on dissent harm scales $a$; the slider on bargaining cost scales $b$. Raising $a$ (constitutional stakes) pushes $k^*$ toward unanimity; raising $b$ (hard bargaining) pulls it back toward simple majority. There is no single best rule, only $k^* = a/(a+b)$ for a given class of decision — which is exactly the Calculus of Consent's point.
The founding move, then, is a refusal and a reconstruction. It refuses the romantic state and the aggregation paradigm that goes with it — the search for a social-welfare function that Arrow had shown leads nowhere stable. It reconstructs politics as exchange among rational individuals, governed by rules those individuals would consent to behind a veil of constitutional uncertainty. The formal apparatus — the median-voter theorem stated as a proof, the optimal-voting-rule model derived from utility, Arrow's theorem proved in full — belongs to the economics textbook's microeconomics and to its chapter on institutional economics; the intellectual claim is the one this chapter carries. Once the benevolent planner is dropped, the rest of the tradition is the working-out of what self-interested agents do in each political role: in the interest group, the rent contest, the bureau, and the legislature. The next section takes the role that produced the tradition's single most portable result.
You have just met the founding move — the romantic state dropped, politics modeled as exchange. This is one of three converging lineages a coherent right-economics rests on.
Public choice is one of three lineages that converge into a coherent right-of-center economics: the Austrian tradition (knowledge and prices), the macro counter-revolution (rules over discretion), and this one — the political-economy counter-revolution that dropped the benevolent-planner assumption. The Buchanan-Tullock founding, Olson's collective-action asymmetry, Tullock's rent-seeking, and the government-failure-symmetric-to-market-failure prior together supply the case that "the state will fix it" is not a free move. The full three-lineage argument, and where the public-choice strand sits in it, is the walkthrough's load; this chapter supplies its lineage.
Here is a puzzle the romantic theory of democracy cannot answer. A sugar tariff transfers a few hundred million dollars a year from American consumers to a handful of sugar producers. The producers are few and the consumers are tens of millions; in a system that counts heads, the many should defeat the few every time. They do not. The tariff persists across administrations of both parties. The same pattern recurs across the policy landscape: trade protection, occupational licensing, farm subsidies, tax carve-outs — narrow producer interests beat broad consumer interests with a regularity that majority-rule arithmetic says should be impossible. Why do diffuse majorities lose to concentrated minorities?
Mancur Olson's The Logic of Collective Action (1965) answered the puzzle and, in doing so, produced the public-choice tradition's most portable single result. The answer is the free-rider problem applied to organization itself. A collective good — a tariff that raises every sugar producer's price, a piece of legislation that benefits every member of a group — is, by definition, available to all members of the group whether or not they helped secure it. A rational member therefore reasons: my contribution to the lobbying effort is costly to me, the benefit if it succeeds is shared by everyone regardless of what I do, and my individual contribution is unlikely to be decisive. So I free-ride, and let others bear the cost. The trouble is that every member reasons the same way, so the collective good is under-provided or not provided at all. The group fails to organize not because its members are irrational but because each is acting perfectly rationally.
The decisive twist is that this failure gets worse as the group gets larger and more diffuse. In a small group — three sugar refiners, say — each member's contribution is a large share of the total, each can see whether the others are pulling their weight, and the benefit to each from the good's provision is large enough to justify acting even if others free-ride. Olson called such a group privileged: it will tend to provide its collective good. In a large group — millions of sugar consumers — each member's potential contribution is a negligible share of the total, no one can monitor anyone else, and the per-person benefit is so small that no individual finds it worth acting on. Olson called such a group latent: it will not organize on its own, however large its aggregate stake. The result is the concentrated-benefits, diffuse-costs asymmetry — the master diagnostic of the entire tradition. A policy whose benefits concentrate on a small group and whose costs spread thinly across a large one will be politically supplied even when it reduces total welfare, because the small group organizes and the large one cannot. The sugar producers each gain millions and form a privileged group; the sugar consumers each lose a few dollars a year and remain a latent one. The few beat the many because the few can act and the many cannot.
There is a middle case, and Olson's name for it explains how large groups ever organize at all. An intermediate group — larger than a privileged few, smaller than a latent mass — may organize if some member's stake is large enough to make unilateral action worthwhile, or if the members can monitor and pressure one another. But the general escape from the free-rider trap, and the one that explains why some large groups are organized while most are not, is the selective incentive: a private benefit, available only to those who contribute, bundled with the collective good. A labor union that offers members a grievance service, a discounted insurance plan, a professional credential; a farm bureau that sells members crop insurance and equipment discounts; a professional association that controls licensing — each ties a private, excludable reward to membership so that the individual has a reason to join and pay dues quite apart from the collective good the dues also finance. The lobbying is, in effect, a by-product of an organization that exists to supply private benefits. This is why effective interest groups are so often organized around something members want for themselves, and why purely public-interest causes — clean air, lower taxes for everyone, honest government — are chronically under-organized relative to the breadth of their support: they have a collective good to offer and no private one to bundle with it. The asymmetry is not that people do not care about diffuse goods; it is that caring does not, by itself, generate the organization that turns numbers into political weight.
Slide group size up and watch the equilibrium contribution rate collapse: a small group provides its collective good, a large diffuse group free-rides itself into providing nothing. That collapse is Olson's latent-group result, and it is the engine of "concentrated benefits beat diffuse costs." Then switch on a selective incentive — a private reward tied to contributing — and watch provision return even at large size. That is Olson's own solution to the problem he posed.
Figure 14.4 (interactive). Equilibrium contribution rate as a function of group size in a public-good contribution game. Provision collapses as the group grows (the latent-group result); a selective incentive restores it. Slide group size and watch the collapse; toggle the selective incentive and watch provision return.
In a small group your contribution visibly matters and you reap a real share of the benefit, so you pitch in. In a huge group your contribution is a drop in the ocean, nobody is watching, and the benefit comes to you whether you pay or not — so you don't, and neither does anyone else, and the group that should win on numbers wins nothing. Olson's escape is the selective incentive: bundle a private perk with membership (a magazine, a discount, a professional credential) so that contributing pays off for the individual regardless of the collective good. That is why effective lobbies are organized around something members want for themselves, not just the cause.
The asymmetry is the result a non-economist remembers, because it reorganizes how political outcomes look. Once you have it, you stop asking "what does the public want?" and start asking "who is organized, and who is not?" — and the second question predicts policy where the first does not. It explains why regulated industries so often capture their regulators, why trade liberalization is politically harder than its aggregate benefits suggest, why tax expenditures accumulate and rarely get repealed, why farm support survives in countries where farmers are two percent of the population. Each is a case of a concentrated, organized interest extracting a benefit from a diffuse, unorganized public that pays a little each and never musters the collective will to resist.
The asymmetry also reframes a familiar complaint about democracy. The usual lament is that special interests corrupt a system that would otherwise serve the public — that money, lobbying, and influence-peddling are pathologies grafted onto an essentially healthy body. Olson's logic says something more unsettling: the dominance of concentrated interests is not a corruption of the democratic process but a structural feature of it, present wherever benefits concentrate and costs diffuse, and it would persist even if every actor were perfectly honest and every campaign perfectly clean. No bribery is required for a sugar tariff to survive; all that is required is that sugar producers find it worth their while to organize and sugar consumers do not. This is why reforms aimed at the symptoms — campaign-finance limits, lobbying disclosure, ethics rules — tend to disappoint: they target the visible machinery of influence without changing the underlying asymmetry of organization that gives concentrated interests their advantage. The result a non-economist remembers is also, on reflection, the most pessimistic claim in the tradition, because it locates the problem not in bad actors who could be removed but in the incentive structure of collective action itself. Recognizing that does not counsel despair — selective incentives, transparency, and institutions that lower the cost of broad organization all push back — but it does insist that the problem be diagnosed correctly before it can be addressed, and the diagnosis is structural.
Olson's later The Rise and Decline of Nations (1982) ran the logic forward across historical time, and the extension reaches its macro-historical payoff. If small concentrated groups organize where large diffuse ones cannot, then over a long period of political stability a society accumulates distributional coalitions — interest groups, cartels, lobbies, professional bodies — each of which captures favors, blocks competition, and rigidifies the economy in its own narrow interest. These coalitions are slow to form and, once formed, hard to dislodge; they grow denser the longer a society goes without the upheaval (defeat in war, revolution, occupation) that sweeps them away. The result is institutional sclerosis: stable, undisturbed societies gradually clog with accumulated special-interest privilege and lose dynamism, while societies whose old coalitions have been destroyed grow fast in the aftermath. Olson used the thesis to explain a striking pattern of the postwar decades — the rapid growth of defeated and occupied Germany and Japan, whose prewar coalitions had been broken, against the slower growth of victorious Britain, whose dense web of established interests survived intact. The free-rider result, scaled up to economic history, becomes a theory of why nations stagnate.
If the logic of collective action supplied the tradition's most portable result, the Virginia school — the Buchanan-Tullock program centered first at the University of Virginia and later at George Mason — supplied its sharpest set of tools. Each is an instance of the same move: take a political role that the romantic theory treats as a servant of the public interest, and model its occupant as a rational maximizer instead. The result is one apparatus applied across six roles, and the sharpest instrument of the set is Gordon Tullock's.
Tullock's "The Welfare Costs of Tariffs, Monopolies, and Theft" (1967) re-measured the social cost of a monopoly, and the re-measurement is the sharpest single move in the Virginia toolkit. The textbook cost of a monopoly is the deadweight loss triangle — the trades that do not happen because the monopolist restricts output to raise price. Arnold Harberger had estimated this triangle in the 1950s and found it surprisingly small, a fraction of a percent of national income, which seemed to suggest monopoly was not worth worrying about much. Tullock's insight was that the triangle is only part of the cost. A monopoly worth having is a prize, and rational agents will spend real resources competing to win it — lobbying for the protective tariff, litigating for the exclusive license, donating to the politicians who grant the franchise, staffing the offices that defend the privilege. Those resources are a pure social loss: they produce nothing, they merely transfer an existing rent from one pocket to another, and in the competition to capture the transfer they are burned up. This is rent-seeking — the expenditure of real resources to capture an artificially created rent (a monopoly, a tariff, a subsidy, a license), as distinct from profit-seeking, which earns a return by creating something of value. The distinction is load-bearing and it is the easiest thing here to get wrong: profit-seeking builds a better product and is the engine of a market economy; rent-seeking builds a better lobbying operation and is pure waste. The social cost of the monopoly is the deadweight triangle plus the resources dissipated competing for the rent — the Tullock rectangle, the area Harberger's triangle left out. Anne Krueger's 1974 paper, which coined the term "rent-seeking" and estimated its magnitude in India and Turkey, found the dissipated rectangle could run to several percent of national income — an order of magnitude larger than the triangle Harberger had measured. The interactive below lets you watch the rectangle fill.
The social cost of a monopoly is the familiar deadweight triangle plus the Tullock rectangle — the resources burned competing for the rent. Slide the number of contestants chasing the rent and watch the dissipation rise toward the rent's whole value; slide the rent's value and watch both areas scale. The slider also reveals the Tullock paradox: at small or asymmetric contests, total spending falls below the rent — under-dissipation a static "rectangle = rent" picture hides.
Figure 14.5 (interactive). The social cost of a rent: the deadweight triangle (the monopoly distortion) beside the Tullock rectangle (resources dissipated competing for the rent), with total dissipation $V\,(n-1)/n$ rising toward $V$ as contestants enter. Slide the number of contestants and the rent value; watch the rectangle fill and the under-dissipation paradox at small $n$.
A monopoly licence worth a hundred million dollars is worth almost a hundred million dollars of lawyers, lobbyists, and campaign donations to whoever can capture it — and all of that effort produces nothing, it just decides who pockets the rent. That burned-up competition is the real cost of protection, and it dwarfs the textbook triangle. The twist the slider shows: when only a few players can compete, or one has a big head start, the contest under-dissipates — the winner keeps part of the rent rather than spending it all away. The waste is real and large, but how much of the rent gets burned depends on the structure of the contest.
In the standard symmetric Tullock contest, $n$ risk-neutral contestants compete for a rent $V$, each winning with probability proportional to its outlay. The unique symmetric Nash equilibrium has each spending $V(n-1)/n^2$, so total dissipation is $V(n-1)/n$, which rises monotonically toward $V$ as $n \to \infty$ and equals zero at $n=1$.
Total social cost $= $ Harberger triangle $+$ Tullock rectangle. The rectangle is the dissipation $V(n-1)/n$; the under-dissipation (the Tullock paradox) is the gap $V/n$ between the rent and what gets spent, which shrinks as the contest crowds. Asymmetric contests dissipate even less. The formal derivation is the economics textbook's institutional-economics chapter; this is the result, not the proof.
The same maximizing move, applied to the next political role, gives the budget-maximizing bureau. William Niskanen's Bureaucracy and Representative Government (1971) modeled the public bureau not as a faithful executor of the legislature's will but as an organization run by officials who want what officials rationally want: larger budgets, more staff, higher status, easier work. The bureau enjoys a decisive advantage in the bargain with its legislative sponsor — it is a monopoly supplier of its service, and it knows far more about the true cost of providing that service than the legislators who fund it. A rational bureau exploits that information asymmetry to over-supply: it presents its output as a take-it-or-leave-it package at an inflated budget, and the sponsor, unable to price the service independently and unwilling to do without it, funds more of it than the public interest would warrant. Government failure, on the Niskanen model, has a supply side as well as a demand side: not only do organized interests demand inefficient policies, but the agencies that supply public services have their own incentive to over-provide.
The bureau is not the only public body that comes to serve interests other than the public's. Regulatory capture is the result when the demand side of government failure and the supply side meet: George Stigler's "The Theory of Economic Regulation" (1971), written from Chicago rather than Virginia but pointing at the same target, argued that regulation is typically acquired by the industry it regulates and operated for that industry's benefit. The logic is pure Olson. The regulated firms are a small, concentrated, intensely interested group with everything to gain from favorable rules — entry barriers that keep out competitors, price floors, rules written to the incumbents' technology — while the public the regulation nominally protects is a large, diffuse, rationally inattentive group. The agency is staffed by people whose careers run through the industry, who depend on the industry for the information they regulate with, and who face concentrated pressure from one side and none from the other. The prediction is not that every regulator is corrupt but that the structural incentives bend regulation, over time, toward the interest that is organized to bend it. Stigler is Chicago, not Virginia, and the capture story has a home in both traditions; what public choice added was the collective-action mechanism that explains why capture is the expected outcome rather than an occasional scandal. The convergence of the Chicago capture literature and the Virginia public-choice apparatus into a single account of how organized interests work the machinery of the state is one of the places where the broader counter-revolution's strands run together.
Two more roles complete the apparatus. In the legislature, log-rolling — the trading of votes across bills — is the politics-as-exchange paradigm at its most literal: a legislator who cares intensely about a dam in her district and little about a bridge in another trades her vote on the bridge for support on the dam. Vote-trading lets intensity of preference register, which simple head-counting cannot, but it also lets coalitions assemble bundles of projects that no majority would support item by item, each project passing on the strength of trades rather than merit. And among voters, Downs's category of rational ignorance closes the loop: since a single vote almost never decides an election, the rational voter invests almost nothing in becoming informed, which is exactly why concentrated interests — who have every incentive to be informed and organized — find a diffuse, inattentive electorate so easy to work around. The macro-policy instance is the political business cycle (William Nordhaus, 1975): an incumbent who wants re-election has reason to stimulate the economy before an election, accepting the inflation that follows after the votes are counted, distorting stabilization policy to the electoral timetable. Six roles — the rent contestant, the bureaucrat, the legislator, the vote-trader, the ignorant voter, the incumbent — and one model running through all of them: the rational self-interested agent, placed in a political position, doing what the incentives of that position reward. The formal models of the rent contest and the budget-maximizing bureau live in the economics textbook's institutional-economics chapter; the unifying intellectual claim is that they are one apparatus, not six separate findings.
The political business cycle deserves a moment of its own, because it shows the apparatus reaching all the way into macroeconomic policy. Nordhaus's 1975 model takes the incumbent's incentive seriously: voters reward a strong economy at election time and have short memories, so an office-holder who can influence monetary and fiscal policy has reason to engineer a pre-election boom and let the inflationary hangover arrive after the votes are counted. The result is a macroeconomy whose fluctuations track the electoral calendar rather than the needs of stabilization — expansionary policy timed to the polls, contraction deferred until safely after them. Whether the effect is large in practice has been debated for decades, and disciplined central banks and rational, forward-looking voters both blunt it; but the model's importance is less its empirical magnitude than what it demonstrates. It shows that the same self-interest assumption that explains why a sugar tariff survives also generates a prediction about the timing of recessions, which means the public-choice apparatus is not a special-purpose tool for studying lobbying but a general method for finding the incentive structure inside any political decision. That generality — one assumption, carried into role after role, each time producing a result the romantic theory could not — is what made the apparatus worth absorbing, and what the sections that follow watch it do.
You have just seen rent-seeking and capture as the cost of protection. This is the government-failure case against industrial policy at strength.
The sharpest case against industrial policy is the rent-seeking one: a subsidy or tariff meant to nurse an infant industry creates a rent, the industry organizes to defend it, and the infant becomes a politically powerful adult who lobbies to keep the protection long after the market-failure justification has lapsed. The capture and rent-seeking apparatus this section walks is the lineage behind that critique; the walkthrough holds it against the case for mission-oriented industrial policy. The protection outlives the problem it was meant to solve.
You have the public-choice toolkit for capture and rent-seeking. The minimal-state tradition uses it to ask whether state capacity gets used for the public good.
The minimal-state tradition engages state capacity at its strongest sceptical point: the public-choice and Chicago argument that the capacity to intervene is also the capacity to be captured. Stigler-Becker regulatory capture and the Buchanan-Tullock skepticism about whether state power gets used for the public good are the lineage this thread routes here; against the developmental-state case that capacity is what makes growth possible, they are the warning that capacity is dangerous precisely because organized interests will turn it to their own ends.
Everything to this point has been apparatus. Here is the claim the apparatus was built to land. Economics had, by the middle of the twentieth century, a fully developed theory of market failure. The Pigovian tradition — Arthur Pigou and the welfare economics that followed — established the cases in which a competitive market, left alone, produces an outcome that departs from the efficient one: externalities (the factory that pollutes a river imposes costs it does not bear), public goods (the lighthouse that everyone uses and no one will pay for), monopoly (the firm that restricts output to raise price), information asymmetries. In each case the market, judged against the efficient ideal, fails — and the failure was taken to establish a case for government action to correct it. This is the half of the comparison economics built. For decades it was the whole of the comparison, and the gap is what public choice exists to fill.
The gap is this. Market-failure analysis compares a real market — with its externalities and its monopolies — against a frictionless efficient ideal, finds the market wanting, and infers a case for government to step in. But it compares the remedy against nothing at all. It assumes, silently, that the government doing the correcting is a benevolent planner who wants only efficiency and faces no incentive problems of its own. Drop that assumption — and dropping it is the single move the whole tradition turns on — and the comparison changes completely. The government remedy is not produced by a benevolent planner. It is produced by the same rational, self-interested agents the chapter has been modeling: vote-seeking legislators responsive to concentrated interests, budget-maximizing bureaucrats with private information, rent-seekers competing to capture the regulation, incumbents on an electoral clock. The remedy, in other words, is subject to government failure — the systematic ways political and bureaucratic provision departs from the efficient outcome, arising from the same maximizing logic applied to political actors — just as the market is subject to market failure. The two failures are symmetric. The interactive below makes the symmetry something you produce by your own act of dropping the planner.
On the left, the market-failure world economics built: toggle an imperfection (an externality, a public good, a monopoly) and watch the market outcome depart from the efficient ideal. On the right, greyed out, the government remedy — held behind a single hinge labeled "benevolent-planner assumption: ON." Flip the hinge OFF and the government panel comes alive: now toggle each political imperfection and watch the remedy depart from the efficient outcome by the very same maximizing logic. One assumption, dropped, makes the second half of the comparison appear — and you drop it yourself.
Figure 14.6 (interactive). The market-failure / government-failure mirror. The market panel departs from efficiency under its imperfections; the government panel is gated behind the benevolent-planner hinge and, once the hinge is dropped, departs from efficiency under its imperfections by the same logic. Flip the hinge and watch the right panel materialize; toggle each imperfection; switch the intuition/formal framing.
"The market failed, so the government should fix it" is only half an argument. It compares a real, flawed market against a perfect government that exists only in the model. Drop that perfect government — replace it with the actual one, run by people with elections to win and budgets to grow and lobbyists at the door — and the remedy turns out to have failures of its own, by exactly the same logic that produced the market's. The honest comparison is not market-versus-ideal; it is one real, flawed institution against another, with the question being which flawed outcome is less bad. That is the comparative-institutional discipline, and it is the whole point.
The nirvana fallacy (Harold Demsetz, 1969) is the error of comparing an actual institution against an idealized alternative that no real institution can attain. Formally, market-failure analysis establishes $W(\text{market}) < W(\text{first-best})$ and infers a case for intervention — but the relevant comparison is $W(\text{market})$ versus $W(\text{actual government remedy})$, not versus the unattainable first-best.
Comparative-institutional analysis replaces the ideal benchmark with the best feasible alternative: choose the institution that maximizes welfare over the set of attainable arrangements, each with its own failures. The benevolent-planner assumption is exactly the move that sets $W(\text{government}) = W(\text{first-best})$; dropping it restores $W(\text{government})$ to a real, flawed quantity to be compared, not assumed optimal.
The symmetry has a name for the error it corrects. Harold Demsetz's "Information and Efficiency" (1969) called it the nirvana fallacy: the mistake of comparing a real, imperfect institution against an idealized alternative that does not exist, rather than against a real, imperfect alternative that does. Market-failure analysis that stops at "the market fell short of the ideal, therefore government should act" commits the fallacy — it holds the market to the standard of nirvana and the government to no standard at all. The discipline the symmetry installs is comparative-institutional analysis: evaluate an actual market, with its failures, only against an actual government remedy, with its failures, and choose the less imperfect of the two real options. Never measure either against a frictionless first-best that no institution can reach.
It is worth being clear about what the symmetry does and does not establish, because the move is easy to over-read in either direction. It does not show that government always fails, or that markets are always to be preferred, or that intervention never improves on the unaided market. That would be the mirror-image error — holding the government to the standard of nirvana and the market to none. A pollution externality is a real market failure; an emissions tax may genuinely improve on it even after the tax is set by a legislature responsive to the polluters and administered by an agency with its own incentives. The symmetry establishes something more modest and more useful: that the question is empirical and comparative, not settled by the diagnosis of market failure alone. Whether a particular intervention helps depends on the magnitude of the market failure, the severity of the government failures the remedy would carry, and the comparison between the two actual outcomes. Sometimes the market failure is large and the government failures small, and intervention wins; sometimes the market failure is small and the political failures large, and it loses; and the only way to know is to do the comparison honestly rather than to assume the answer from the diagnosis. What public choice removed was not the case for government action but the automatic case — the inference that ran from "the market is imperfect" straight to "therefore the state should act," skipping the question of whether the state's action would be any better.
This is the tradition's central methodological contribution, and stating it plainly is worth the space. Before public choice, "the market fails, so the government should fix it" was a valid inference — the second clause followed from the first because the government was assumed to do its job. After public choice, the inference is broken: the first clause establishes only that the market is imperfect, and whether government action improves matters depends entirely on whether the actual government remedy, with its actual failures, does better than the actual market, with its. "The state will fix it" stopped being a free move in policy argument. It became a claim that has to be earned, case by case, by showing that the comparative-institutional balance favors the remedy — which sometimes it does and sometimes it does not. The market-failure side of this comparison is the economics textbook's chapter on market failures and externalities; the government-failure side and the comparative-institutional framework are its institutional-economics chapter. The structural point — that economics built half the comparison and public choice built the other half — is the one this chapter owns.
The discipline cuts in both directions, and seeing it cut against the deregulatory conclusion is the surest sign that the symmetry is an analytical tool rather than an ideological one. Consider a natural monopoly — a water utility, a power grid — where a single firm can supply the whole market more cheaply than several can. The market failure is real: an unregulated monopolist would restrict output and overcharge. The comparative-institutional question is whether a regulatory remedy, with its capture risk and its information problems, does better than the unregulated monopoly. Sometimes the honest answer is yes: a regulated utility, imperfect as the regulator is, may still beat an unconstrained monopolist gouging captive customers. The comparative-institutional discipline does not prejudge that; it insists only that the comparison be made between the two real options rather than between the real monopoly and an imaginary perfect market, or between the real market and an imaginary perfect regulator. A serious advocate of intervention should welcome this, because it is the only framework in which a case for intervention can actually be won rather than assumed — and a case won by comparing real outcomes is far more robust than one that rests on the silent premise that the government doing the correcting is a benevolent planner. The tradition's gift to policy argument is not a thumb on the scale against government; it is the demand that the scale have both pans.
You have just seen the symmetry: government can fail as surely as a market can. That is the public-choice voice in the market-efficiency debate.
When someone answers "markets are inefficient, so government should act," the public-choice reply is the comparative-institutional one: the government remedy is run by the same self-interested agents, so it has failures of its own, and the case for intervention has to clear the comparative bar, not just point at a market imperfection. In the market-efficiency walkthrough this is the secondary against-voice — the government-failure mirror standing beside the Hayekian knowledge argument. Markets fail; so do the governments that would fix them; compare the two real outcomes.
You have the government-failure apparatus. Turned on the regulators and the discipline itself, it sharpens the pre-2008 story.
The government-failure machinery applies as sharply to the regulatory and academic institutions of the pre-2008 period as to any market. Regulators captured by the firms they oversaw, agencies with their own incentives, a discipline whose institutional rewards favored certain models — the comparative-institutional lens reads the run-up to the crisis as a failure of the institutions meant to police the market, not only of the market itself. The public-choice apparatus is sharper here than the market-failure machinery alone, because the failures were on the remedy's side of the mirror.
Public choice arrived in the world coded right, and the coding was not an accident. Its results pointed in a consistent political direction. If government action is as prone to failure as the market, the case for deregulation strengthens. If diffuse majorities are systematically beaten by concentrated interests, the New Deal picture of government as the tribune of the public against organized capital looks naive. If discretionary stabilization invites the political business cycle, the case for binding rules over discretion strengthens. And if democracies over-spend because the benefits of programs are concentrated and salient while the costs are diffuse and deferred, the case for constitutional limits on taxing and spending follows. These conclusions arrived in the 1970s and 1980s alongside the macroeconomic counter-revolution — Friedman's monetarism, Lucas's rational expectations, the new classical assault on Keynesian stabilization, treated in the counter-revolution chapter. The two movements were allies. One dismantled the case for discretionary macro management; the other dismantled the benevolent-planner assumption underneath the whole interventionist program. Together they supplied much of the intellectual architecture of the era's deregulatory, market-liberalizing turn, the politics of which — the stagflation that discredited the old consensus, the deregulation episodes, the neoliberal turn — economic history narrates in its chapter on stagflation and the neoliberal turn.
The alliance was more than a coincidence of conclusions; it was a division of labor against a common target. The Keynesian-welfarist consensus of the mid-century rested on two assumptions the two movements attacked from different sides. It assumed that the macroeconomy could be fine-tuned by a benevolent and competent government using discretionary policy — the assumption the macro counter-revolution dismantled, by arguing that rational agents anticipate policy and that discretion produces worse outcomes than rules. And it assumed that, where markets fall short, a benevolent and competent government can be relied on to correct them — the assumption public choice dismantled, by modeling the government as the same self-interested machinery as everything else. Strip out both benevolent-planner assumptions and the interventionist program loses its automatic justification on two fronts at once: the macroeconomic case for active stabilization and the microeconomic case for corrective regulation both now have to clear a bar they had previously been allowed to assume. The deregulation of the late 1970s and 1980s — airlines, trucking, telecommunications, finance — drew on exactly this combined argument: that the regulatory apparatus served the regulated industries more than the public (the capture story), that the market failures it nominally addressed were smaller than claimed, and that the cure had become worse than the disease. Whether each particular deregulation was wise is a separate question, and some look better in hindsight than others; the point here is that the intellectual case for the turn was built from public-choice and counter-revolution materials, which is why the two are remembered as a single right-coded moment even though their formal apparatus was entirely distinct.
The alignment is what makes the historiography contested, and the contest deserves to be put at full strength before any verdict. The critique runs like this. Public choice did not discover that government fails; it manufactured a vocabulary in which government failure is the default expectation and market failure a footnote, and it did so in a political moment that wanted exactly that conclusion. The tradition's conclusions were congenial to a deregulatory, anti-government politics, and its institutional home — the Virginia and George Mason programs, the funders who supported them, the think tanks that amplified them — was of a piece with the broader neoliberal project. The constitutional-constraint program was not a neutral finding but a political agenda: the proposal for a balanced-budget amendment, the campaign for constitutional limits on the franchise of the state, the Leviathan model of Geoffrey Brennan and James Buchanan's The Power to Tax (1980) — which models the state not as a flawed servant but as a revenue-maximizing monster to be chained by constitutional rules — is a normative crusade dressed in the language of positive science. On the strongest version of this reading, public choice is ideology disguised as theory: it takes the contestable political claim that the state is dangerous, encodes it in a model, and then presents the model's output as a scientific result. Nancy MacLean's Democracy in Chains (2017) pressed a version of this case hard, casting Buchanan's program as a stealth project to insulate property and markets from democratic majorities — though that book drew substantial scholarly rebuttal from historians and economists (Phillip Magness, Michael Munger, and others) who documented serious errors in its reading of the archival record, so the controversy is itself contested and should not be imported as settled.
The critique is serious and the better part of it is correct. So the call this chapter makes is not to dismiss it but to draw a distinction the critique runs together. There are two things inside public choice, and they have different epistemic status. The first is the positive machinery: the rational political actor, the collective-action asymmetry, rent-seeking, the comparative-institutional discipline. This machinery is sound. It is also portable across the political spectrum, which is the decisive evidence that it is not merely right-wing ideology. The same collective-action logic that explains why sugar producers capture a tariff explains why polluters capture an environmental regulator and why incumbent firms capture the agencies meant to check them — conclusions a progressive reaches as readily as a libertarian. Rent-seeking analysis is used by the left to attack corporate welfare and regulatory capture by industry; the comparative-institutional discipline is exactly what a serious advocate of government action should want, because it tells you when intervention will actually improve on the market rather than assuming it always will. The machinery became training-level — taught in every political-economy course, used across the field's ideological range — precisely because it is analytically productive independent of the politics.
The second thing inside public choice is the constitutional-reform program: the balanced-budget amendment, the Leviathan-chaining constitutional limits, the Buchanan-Wagner argument in Democracy in Deficit (1977) that Keynesian deficit finance removed the discipline that once forced democracies to balance their books and so unleashed a permanent bias toward deficits that constitutional rules must restrain. This is a normative project. It rests on value judgments — about how much the state should be constrained, about whose interests constitutional rules should protect, about the relative weight of the harms of over-government and under-government — that no model produces and no theorem settles. It is a politics, and a respectable one, but it is not a finding. The error the critique makes is to read the program as discrediting the machinery; the error the tradition's defenders make is to launder the program as though the machinery proved it. Neither follows. The positive analysis of how political actors behave is one thing, and it is sound and ideologically portable; Buchanan's proposals for what the rules should therefore be are another, and they are a normative argument that stands or falls on its value premises, not on the science. Distinguish the finding from the program, and the contested historiography resolves: the machinery is the durable contribution, the reform program is a separable politics, and the conflation of the two is what generates the heat on both sides.
You have just met the constitutional-reform program. Its fiscal-discipline strand is the public-choice case that democracies systematically over-promise.
Buchanan and Wagner's Democracy in Deficit (1977) is the public-choice fiscal-unsustainability lineage: democracies over-promise because the benefits of spending are concentrated and immediate while the costs are diffuse and deferred onto future taxpayers who do not yet vote, so the political market is biased toward deficits unless a constitutional rule restrains it. The walkthrough holds that diagnosis against the case that the welfare state's books can be balanced by ordinary means; this section supplies the lineage behind the over-promise claim.
The public-choice pole of the method-layer debate — intervention carries capture risk — sits beside the case that the state can be mission-oriented.
One pole of the markets-versus-states debate is the public-choice and Chicago position that any power to intervene is also a power to be captured, so the comparative-institutional question — will this real government do better than this real market? — has to be answered case by case rather than assumed. Buchanan's constitutional-design tradition supplies one answer: constrain the state's discretion at the constitutional stage. This section carries that pole's lineage; the walkthrough holds it against the mission-oriented-state case.
Public choice is no longer a school, and the reason is that it won. A school exists as a distinct camp when its central claims are contested and its methods are its own; it dissolves when those claims become assumptions and those methods become defaults that everyone uses without attribution. That is what happened. The rational political actor, the collective-action asymmetry, rent-seeking analysis, the comparative-institutional discipline — these are no longer the property of a Virginia-school camp arguing against a mainstream. They are training-level apparatus, taught in graduate political economy across the ideological range, deployed by economists and political scientists who could not say which school they belong to because the question no longer means anything. The tradition's success is measured by its invisibility: when an idea becomes the water everyone swims in, it stops being recognizable as a position.
The dissolution is visible in what an educated reader now takes for granted. That regulators can be captured by the industries they oversee; that programs with concentrated beneficiaries and diffuse costs are politically durable far out of proportion to their merits; that "the government should fix it" is a claim to be tested rather than a conclusion to be assumed; that public officials respond to the incentives of their offices like everyone else — none of this reads, today, as a partisan thesis. It reads as common sense, the kind a journalist deploys without knowing it has a source, the kind a graduate student absorbs before learning whose work it came from. That is precisely the mark of an absorbed tradition: its claims have migrated from the contested foreground into the unexamined background, from things one argues for into things one argues from. A school that has to defend its premises is alive as a school; a school whose premises have become everyone's starting point has dissolved into the discipline, and the dissolution is the victory.
The terminus has a name in political science: positive political theory. Where the early public choice of the 1960s applied economic reasoning to politics in broad strokes, the rational-choice institutionalism that grew out of it — Kenneth Shepsle and Barry Weingast's work in the 1980s — turned the apparatus on the fine structure of institutions. Their central concept, structure-induced equilibrium, answered a puzzle public choice had inherited from Arrow: if bare majority rule cycles endlessly, why do real legislatures produce stable outcomes? The answer is that institutional structure — committee jurisdictions, agenda rules, the sequence in which proposals are considered, the gatekeeping power of who decides what comes to a vote — constrains the chaos of pure majority rule into a determinate result. The cycling that Arrow's theorem and the Condorcet paradox predict is real in principle and tamed in practice by the rules of the institution, and studying how the rules do the taming is what positive political theory does. This is the public-choice apparatus grown into a mature, technical sub-discipline that has no need to call itself public choice.
The apparatus also flowed forward into the modern political economy of institutions, where it became an input to a larger account. Daron Acemoglu and James Robinson's work on extractive versus inclusive institutions — the argument of Why Nations Fail (2012) that the deep cause of the wealth and poverty of nations is whether their institutions are organized to extract rents for a narrow elite or to provide broad incentives to invest — runs on machinery public choice supplied. Rent-seeking is what extractive institutions are for; collective action is why narrow elites can capture the state against the diffuse majority; the comparative-institutional question is the one the whole research program asks. Persson and Tabellini's Political Economics (2000) made the marriage formal, building the median voter, the rent-seeking interest group, and the budget-maximizing politician into general-equilibrium models of policy determination. Public choice did not vanish into these programs; it became their foundation. Where the institutional tradition picks up the question public choice took as given — not "how do political actors behave under the rules?" but "where do the rules come from, and why do they persist?" — is the subject of the institutionalist-tradition chapter, which carries the new institutional economics from Coase and North through to Acemoglu and Robinson. Public choice and that tradition converge on the same modern-political-economy terminus from two directions, and the convergence is the seam between the two chapters.
What survived the absorption, and what did not, is worth stating precisely, because the survival is uneven in exactly the way the chapter's verdict predicts. The positive machinery survived whole and became foundational: the rational political actor, the collective-action asymmetry, rent-seeking, the comparative-institutional discipline, the structure-induced taming of cycling — all of it is now built into the toolkit of political economy, formalized in the Persson-Tabellini models, deployed in the Acemoglu-Robinson account, taught without attribution. The early tradition's grander ambitions fared less well. The strongest version of the deductive constitutional project — the hope that one could derive, from the model of the self-interested political agent, a determinate set of constitutional rules that rational individuals would unanimously consent to behind the veil — turned out to underdetermine the rules it sought, because the choice among feasible constitutions depends on value judgments and risk attitudes the model does not fix. And the most polemical applications, the ones that treated every act of government as presumptively a rent-extraction, did not survive contact with the comparative-institutional discipline the tradition itself supplied, since that discipline insists the presumption be tested rather than assumed. What absorbed, in other words, was the analytical core that earns its keep across the political spectrum; what fell away was the part that was politics rather than science. That selective survival is the cleanest possible evidence for the verdict that follows.
The institutional recognition came in 1986, when James Buchanan received the Nobel Prize "for his development of the contractual and constitutional bases for the theory of economic and political decision-making." The citation honored the constitutional political economy of the Calculus of Consent — the founding move, twenty-four years on, certified as a permanent contribution to the discipline. By then the machinery it launched was already dissolving into the mainstream it had once stood outside. The verdict the chapter lands is the one its absorption earns: public choice was a genuine and durable methodological achievement, not the ideological artifact its critics charge, because its positive core proved portable across the political spectrum and became default apparatus, while the constitutional-reform program Buchanan attached to it remained a separable normative project that the science neither proved nor required. The tradition stopped being a school because its methods became the way the mainstream analyzes political institutions — which is the most complete victory an intellectual movement can win, and the hardest to see, because nothing is harder to notice than an assumption everyone shares. The era-cluster on the intellectual-history timeline that surrounds this tradition — the counter-revolution figures it allied with, the modern institutional thinkers it flowed into — shows the lineage's neighbors even though the public-choice nodes themselves are not yet on the graph, a gap the timeline's coverage of this school has still to close.
Named literature: Wicksell, Finanztheoretische Untersuchungen (1896). Arrow, Social Choice and Individual Values (1951). Black, "On the Rationale of Group Decision-Making" (1948); The Theory of Committees and Elections (1958). Downs, An Economic Theory of Democracy (1957). Buchanan and Tullock, The Calculus of Consent (1962). Olson, The Logic of Collective Action (1965); The Rise and Decline of Nations (1982). Tullock, "The Welfare Costs of Tariffs, Monopolies, and Theft" (1967). Demsetz, "Information and Efficiency: Another Viewpoint" (1969). Niskanen, Bureaucracy and Representative Government (1971). Stigler, "The Theory of Economic Regulation" (1971). Krueger, "The Political Economy of the Rent-Seeking Society" (1974). Nordhaus, "The Political Business Cycle" (1975). Buchanan and Wagner, Democracy in Deficit (1977). Brennan and Buchanan, The Power to Tax (1980). Shepsle and Weingast, "Structure-Induced Equilibrium and Legislative Choice" (1981). Acemoglu and Robinson, Why Nations Fail (2012). Persson and Tabellini, Political Economics (2000).