Steven Teles on Kludgeocracy
What this covers
Steven Teles of Johns Hopkins joins Russ Roberts to examine what he calls kludgeocracy — a diagnosis of American governance as fundamentally broken by unplanned complexity rather than by size. The conversation centers on how constitutional features like separation of powers and veto points function as obstacles to reform, not merely to new legislation. Teles argues that veto points act as "Veto Points as Toll Booths," where politicians extract payment in the form of preserving existing arrangements. The result is a system wired to pile new programs atop old ones rather than to eliminate what no longer works, leaving behind layers of regulation, tax expenditures, litigation, and mandates that hide government's true cost and reach.
The discussion ranges across how this complexity serves the interests of both left and right. Teles traces how constraints on visible spending pushed government demand into less measurable channels—regulation, private litigation, tax shelters—creating Hidden Taxation that distorts private markets without appearing in the budget. He examines specific cases: how the filibuster made the Affordable Care Act more kludgey, how the parallel retirement-account system overwhelms individual decision-making, and how Concentrated Benefits, Dispersed Costs (drawing on Mançur Olson's logic) locks in policies like sugar quotas. Teles argues that liberals should embrace simpler, more transparent programs to make their case plainly, while conservatives should recognize that diffused, hidden government is actually harder to remove than transparent programs. The conversation also sketches when reform becomes possible—through Punctuated Equilibrium in Politics, where media scandals and focusing events can suddenly dislodge seemingly entrenched arrangements.
Teles argues that America's central governance problem is not the size of government but its complexity ('kludgeocracy') -- an unplanned accretion of overlapping programs preserved by veto points and entrenched interests -- and that both liberals and conservatives have reasons to prefer simpler, more transparent governance.
- Veto points in Congress and federalism function as 'toll booths' that extract payment in the form of preserving existing arrangements, biasing the system toward layering new programs atop old ones.
- Complex, hidden forms of government (regulation, litigation, mandates, tax expenditures) impose real but invisible costs and are harder to measure and reform than direct taxing and spending.
- Conservatives' success in constraining taxing and spending merely pushed government demand into less measurable channels, while liberals' kludgey programs hide government's hand from beneficiaries.
Constitutional veto points function as toll booths extracting prices for passage rather than genuinely constraining government action.
- The original constitutional design (separation of powers, filibuster) is well-suited to a people with modest expectations of the federal government, but once public expectations shift to a mature set of government functions, that design works opposite to libertarian intentions: any maturing government will develop new functions, and a desirable one destroys old programs while adding new ones, but the American system is wired for creation and bad at destruction, which is precisely why it produces kludge.
“our system is wired up for creation and not very good at destruction, and that's why you get kludge”
- The filibuster does more than stop action -- it adds roughly ten extra senators who each must be paid off to get past the toll, and because those marginal senators are precisely the ones who want to protect existing interests (as in the ACA's 'cornhusker kickback'), it makes it especially hard to tear up entrenched arrangements from the roots; the same dynamic would block Republican tax reform like Dave Camp's proposal.
“One thing the filibuster does, it doesn't just stop action. Right? It means now you've got 10 extra people all of whom have to get paid off, or bought off, in order to get past the toll.”
- The separation of powers and multiple committee veto points function less as devices to stop government action and more as 'toll booths' where actors extract a price for passage -- and that price is usually preserving existing arrangements, since toll-takers are tied to interests invested in the status quo, which means the constitutional design that slows government on the way up also makes it nearly impossible to undo or redesign existing programs.
“these operate less like veto points and more like toll booths, where the people who are at these particular veto points are naturally interested in just stopping things. Usually what they are interested in is extracting a toll”
- American institutions create separation of powers whose original theory was that it would constrain government by creating veto points that keep government from acting.
“American institutions as we all learned when we were boys and girls, creates separation of powers, and that separate of powers, a lot of the original theory behind that is that it would constrain government”
- Much of the mid-20th-century expansion of the social welfare state had to be routed through state governments rather than done centrally because Southern senators who controlled Congress insisted states run programs so they could ensure black people did not receive welfare payments, while urban political machines similarly demanded to be 'cut in on the action' -- a feature of federalism and separation of powers that produced the complicated forms government took during its expansion.
“Southern Senators controlled Congress; and they insisted, they said, Okay, you can have your bigger government but you are going to have to cut, you are going to have to make sure South Carolina runs it so we can make sure that black people don't get welfare payments”
Government kludge arises from layered programs, hybrid public-private arrangements, and diffused responsibility that obscure accountability.
- Much of what government does is kludge -- adding new programs on top of existing ones because it is too hard to eliminate what already exists, producing complex systems that are hard for citizens to understand and that obscure who is responsible when problems arise.
“A lot of what government today does is an effort in the absence of being able to get rid of a lot of stuff we've already got, to simply add new things on top of what we already have”
- Most privatizations do not make a function genuinely private but retain an expectation that the firm performs a social function while ownership shifts, which leads to pervasive regulation and a 'faux private sector' of contractors (like military and DC consulting firms) whose primary or sole buyer is government; rather than becoming lobbies for further privatization, these contractors become lobbies for continued socialization of the function so long as they capture the benefits.
“those private contractors become a lobby for the continuing socialization of the function, so long as they are the ones who end up getting the benefits”
- American federalism is not 'layer-cake' (clean division of functions) but 'marble-cake' interpenetration, where the federal government funds programs like Medicaid, education, and environmental regulation but sends pervasive rules down to states, producing a compliance mentality that discourages innovation because actors constantly fear landing on the wrong side of a complex rule.
“leads to a kind of compliance mentality, because there are so many complicated rules that are being applied from the federal government down to the states that people are constantly worried that they are going to get on the wrong side of it”
- Big complicated multi-layered systems are themselves forms of taxation that don't pass through the federal budget but impose costs on private actors, making goods more expensive -- hidden forms of extraction that are hard to mobilize against because, unlike direct taxing and spending, we cannot measure who pays or who benefits.
“big, complicated multi-layered systems are in fact forms of taxation. Now they don't go directly through the Federal government budget. But they all end up imposing costs on private actors... But they are hidden forms of extraction that are hard to mobilize against.”
Complexity and opacity matter more for governance quality than the size of government itself.
- The important distinction in evaluating government is not big vs. small but complex/opaque vs. simple/transparent; policy can be simple and large (Social Security) or complex and hard to hold accountable (private tax-sheltered retirement accounts), and the complexity axis matters more than the size axis for assessing governance quality.
“I would distinguish that complexity/simplicity axis from the big-vs.-small government axis”
- Unlike the federal budget where taxing and spending are known to the penny, the distributive effects of governing through private litigation (e.g., the Americans with Disabilities Act) and regulation (e.g., the minimum wage) are subject to wildly varying estimates, so we have almost no reliable sense of how much is being redistributed or who actually benefits and pays.
“the only form of government that we actually know--literally down to the penny, we know exactly how much government is spending... People have incredibly wide ranges of a sense of how much we are actually spending through the American Disabilities Act”
- By pushing down effectively on the visible supply of government (taxing and spending), conservatives and libertarians have not reduced the demand for government but merely constrained one form of its supply -- so like pressing down on a down pillow, the demand pops up elsewhere through litigation, mandates, and regulation (e.g., carbon-tax opposition leading to complicated EV subsidies and light-bulb regulations); the better strategy is to loosen up on the measurable form (taxing and spending) while constraining the forms we cannot measure or control.
“like a down pillow, you push it down in one place, the feathers all push up somewhere else. And I think that's essentially what's happened with some of conservative and libertarian efforts to control government”
- Liberals should dislike kludge because hidden, complex forms of government (subsidy, regulation, market structure) lead citizens -- especially the middle class -- to experience government assistance as private and not recognize their dependence on government, so simpler, more visible programs would make clear how much social welfare depends on a muscular role for government.
“the actual hand of government in helping them pay their mortgage, helping them send their kid to college, etc., etc., these are all things that people experience, especially the middle class often experience as private, because the hand of government is hidden”
Both liberals and conservatives should pursue simpler, more transparent programs and avoid hybrid corporatist arrangements.
- Conservatives should dislike kludge because when government is everywhere and nowhere -- penetrating the tax code, private litigation, and private enterprise -- it becomes like trying to stab gelatin to the wall: those who want to remove government from particular areas cannot get hold of a target diffused across hidden mechanisms.
“it becomes hard, actually, for people who want to get government out of particular areas, when government is sort of everywhere and nowhere it's like trying to stab gelatin to the wall. You just can't do it.”
- Liberals should stop negotiating with themselves by cutting private interests in on the action and instead pursue simple, transparent programs (e.g., Medicare for all rather than the ACA's exchanges), making the social-justice case openly to the public and accepting they may lose -- because they will be blamed anyway when the complexity of hybrid programs makes them look bad.
“My plea to liberals I think would be to be willing to go big or go home. That is, to say that we're going to stop kind of negotiating with ourself and trying to figure out ways to do government by cutting private interests in on the action.”
- Education could be funded simply through a national voucher financed by income tax with people choosing among private providers, which would dissolve the distributive problems created by thousands of property-tax-funded jurisdictions and the complex apparatus of federal grants and adequacy lawsuits that currently attempt to solve those problems.
“We ought to have straight up vouchers funded through income tax and then people get to, you know, would get to choose in a market of private providers”
- The murky middle ground between public and private provision generates corruption and rule-of-law problems; functions should be either straightforwardly government (funded through taxation, delivered through spending) or straightforwardly private, and if government is going to act it should both 'steer and row' rather than create hybrid corporatist arrangements.
“if government is going to do something, steer and row; and if it is not willing to steer and row, it ought to leave it to the private sector”
Entrenched interests mobilized by existing programs resist reform because prospective beneficiaries remain unmobilized and attention capacity is scarce.
- Many special-interest arrangements are actually vulnerable to public exposure, but the political system has limited 'parallel processing capacity' -- it has room for only about one focusing event or scandal at a time -- which is why entrenched arrangements survive even though each could be defeated if it captured sustained attention.
“The real problem is our political system has limited parallel processing capacity. There's only so many things that it can handle in attacking existing interest group arrangements at any one time.”
- The parallel private retirement system of IRAs, 401ks, and Roth IRAs forces individuals into a volume of investment decisions that overwhelms their decision-making capacity, drives them to costly financial advisers and active funds that typically underperform index funds, and imposes real but private costs in time and anxiety.
“all the people who are busily reading magazines trying to get themselves up the learning curve about all these different mutual funds which almost always turn out to give them worse results than simply an index fund would, that's all real costs on people's time and anxiety”
- Interest groups usually form not in anticipation of government action but in response to it (e.g., the elderly organized as a consequence of Social Security), so the existing universe of interest groups is the mobilized force of those who already benefit from government and is dedicated mainly to protecting existing gains, creating a durable bias against undoing programs because prospective beneficiaries of new alternatives remain unmobilized.
“interest groups usually form not in anticipation of government action, but in response to government action. That is, the elderly were not really organized before Social Security. They got organized as a consequence of Social Security.”
- Major destructive reforms occur not from single-day stories but when an issue becomes a sustained media feeding frenzy (e.g., nuclear safety scandals in the 1970s, corporate tax-avoidance stories before the 1986 tax reform), which requires a policy entrepreneur to frame the scandal as a structural problem and officials waiting for their moment to act.
“the thing that really distinguishes those moments when you really get the capacity for major destructive force are when those things become more than just a single day story, but they become a full-on scandal and sort of media feeding frenzy”
Reforms that destroy losers while simplifying tend to endure, unlike those that merely simplify without dismantling supporting interests.
- While US special-interest arrangements like sugar subsidies are real, they are mild compared to corporatist systems in France and Japan where farmers effectively control the government; the US achieved meaningful deregulation in the 1996 Freedom to Farm Act (even though subsidy clawbacks didn't last, the deregulation did) that France and Japan never matched.
“if you look at the United States, we did actually get the Freedom to Farm Act... back in the 1990s. And that was problematic; it wasn't perfect, but France and Japan have never gotten anything with that much deregulation”
- The same veto points that gum up the system and stop action are also entry points where reformers can start something going (e.g., a congressional hearing) without first securing universal agreement, introducing dynamism that produces long periods of stability punctuated by sudden wrenching changes that put seemingly indestructible interest groups on the back foot.
“all of those veto points that I mentioned that stop action from occurring are also entry points. They are places where you can start getting something going.”
- Reforms that simplify by destroying the interests that supported the prior corporatist arrangement tend to stick (airline deregulation bankrupted high-cost carriers like Eastern Air Lines that had wanted regulation), whereas reforms that merely simplify without destroying losers (like the 1986 tax reform) get clawed back once attention moves on and the defeated interests return.
“airline deregulation worked in part because it literally destroyed the strongest supporters of airline regulation. Look at the first people who went to the wall and got bankrupted by airline deregulation: it was airlines like Eastern Air Lines that had the worst cost structure.”