← Writing archive

What the State Cannot Provide (first draft) — part 1 of 4

Source: public/assets/writing/humanstructures/What the State Cannot Provide (first draft).docx · humanstructures

What the State Cannot Provide

Why the goods we take for granted (like shoes) became universal, and why the ones we were promised (like healthcare) did not

Consider the following passage, from Murray Rothbard’s For a New Liberty (1973):

The libertarian who wants to replace government by private enterprises in the above areas is thus treated in the same way as he would be if the government had, for various reasons, been supplying shoes as a tax-financed monopoly from time immemorial. If the government and only the government had had a monopoly of the shoe manufacturing and retailing business, how would most of the public treat the libertarian who now came along to advocate that the government get out of the shoe business and throw it open to private enterprise? He would undoubtedly be treated as follows: people would cry, “How could you? You are opposed to the public, and to poor people, wearing shoes! And who would supply shoes to the public if the government got out of the business? Tell us that! Be constructive! It’s easy to be negative and smart-alecky about government; but tell us who would supply shoes? Which people? How many shoe stores would be available in each city and town? How would the shoe firms be capitalized? How many brands would there be? What material would they use? What lasts? What would be the pricing arrangements for shoes? Wouldn’t regulation of the shoe industry be needed to see to it that the product is sound? And who would supply the poor with shoes? Suppose a poor person didn’t have the money to buy a pair?”

The passage is usually read as a rhetorical device. Its point is taken to be that the market’s solution seems unimaginable when we are unfamiliar with it, but it is immediately obvious it has one. Rothbard is showing that the questions many accept as problematic when aimed at private provision are actually questions the market answers continuously, without any issue, across many sectors, such as shoes. Therefore, in many other industries (not just shoes) where we don’t currently have a good solution we should expect through market forces to create one, and one that is more efficient and robust besides.

This reading is correct as far as it goes. But the passage contains a second observation that Rothbard did not quite draw out, and which is actually the more interesting starting point for anyone interested in the functioning of their society. Notice that the fable depends for its force on the fact that shoes are, in our world, a settled case. We can imagine and scoff at the concept of public shoe provision because we live with a great multitude of private shoe companies that serve our needs, in a way centrally planned soviet shoe factories obviously failed to do so.

However, say the thought experiment were run on something we do not live with—if Rothbard had asked us to imagine private fire departments, or private coinage, or private drug certification or some of the same currently publicly provided ilk—the reader would have responded with similarly indignant questions he puts into the mouths of his imagined interlocutors, and then some. The fable is persuasive about shoes. But also proportionality to how unpersuasive it would be about everything else. This is worth dwelling on, because it changes what the example is really demonstrating. The standard reading treats the shoe case as representative: if you understand why private shoes work, you will understand by extension why private everything would work. However, it also relies on the fact that it is at once obvious to us: shoes could (and should) be free market provided by what we can observe of the world. That’s the whole point, and then the reader generalises to other domains. But will most readers actually look at shoes and apply the same logic to healthcare? To the Fed? To infrastructure planning? Most people don’t view them as nearly the same game.

The most interesting fact here is that the state has, in most developed economies, retreated from the direct provision of physical consumer goods. It does not make shoes. It does not run grocery stores, clothing factories, or electronics firms. Many tried, and then failed obviously. Where it once did (Britain’s nationalized industries of the 1950s and 60s, the command economies of the Eastern Bloc) it abandoned those industries, usually after visible failure. Meanwhile, it has expanded its role in domains that are harder to characterize: monetary policy, land-use regulation, healthcare provision, pharmaceutical approval, credit allocation, infrastructure planning. The list of things the state has entered or expanded its role in is more striking than the list of things it has left.

What’s the default explanation here? Simple. The state retreated from what it was bad at and concentrated on what it was good at. That makes sense. Except for the fact it’s not at all supported by the empirical evidence. The domains it occupies today are the ones where its performance is most contested and hardest to measure. Britain’s housing stock is older and more expensive than almost any developed country; its healthcare waiting lists run to over seven million. American monetary policy has presided over a redistribution of wealth toward wealthy asset holders that no legislature openly voted for. Drug approval imposes costs estimated in the billions per molecule and delays of roughly a decade. These outputs aren’t explained by the theory that this is a state that has concentrated its attention on what it knows it does well. At best, this is a state that provides what we can’t imagine the market could. Similarly to how in some alternate world we could not imagine how there could be any other equivalent to tax-financed shoes. But then, we are back to wondering what happened. If this is structurally similar to shoes, how come the state is involved in these areas, but not in shoes, or grocery stores, and so on?

A more honest explanation is selection by visibility. The state retreated from shoes because shoe failures are visible. You can see the empty shelves, compare the quality to imports, feel the wrong size on your feet. The political cost of failure is immediate and traceable. The state persists in monetary policy, zoning/planning, and drug approval because failures in these domains are not visible in the same way. Inflation takes years to surface and its causes are contested, and in almost any case one can argue that the Fed was the hero or the villain. Housing shortages are blamed on developers rather than on the planning regimes that stopped entrants to the market. A delayed drug will produce a patient who dies of the disease. Just another patient who dies of the disease (how can we tell if they could’ve been saved?), and very few actually count those deaths against the agency that delayed the drug. The state has not selected these domains because it is competent in them. It has settled into them because its incompetence is hard to see for both those on the inside, and those on the outside.

Rothbard’s shoe fable, on this reading, is not a general argument for private provision. It fails as a thought experiment in the domains the government still occupies. Very few Brits would read that paragraph and jump to the privatisation of the NHS. They’d simply consider it a category error. The indignant questions: who would supply them? in what quantities? at what prices? and what about the poor? These are the characteristic, sincere voices of defense in any sector where the state has been present long enough to make its absence unimaginable, where their failures are close to invisible. That the failures are in unseen costs, and the victories are in seen benefits.

I want to argue that this single variable, the visibility of failure in a given domain, does more work in explaining the shape of modern governance more than other candidates usually offer. It explains why governments grow monotonically rather than cycling. It explains why the most politically protected institutions are often the ones performing worst. It explains why some sectors are reformable and others aren’t, in ways that have nothing to do with the difficulty of the underlying problem. And it points to something I’ll get to state in full at the end: that markets, not states, have been the only force in modern history to make goods genuinely universal, and that what the state offers under the name of universalism is in most cases a legal promise of access to things it has prevented from existing in sufficient quantity for the promise to be honored. That if we want truly universal goods, or at least the closest possible approximation, we need to seriously reconsider public mechanisms of provision.

I. The Visibility Asymmetry

The point that political systems respond to what voters can see is old. Bastiat made it most clearly in 1850 with the concepts of the seen and the unseen. Buchanan and Tullock formalised it under the public-choice heading. Alex Tabarrok has recently named the specifics of a case which can generalise by his writings on the FDA’s invisible graveyard. Each is saying a version of the same thing: political costs in positive EV political domains are paid by visible effects, while invisible effects are absorbed silently by those who suffer them. What I want to do here is consolidate the observation into a predictive model that helps us understand why government is today in many ways the shape it is, particularly in how this leads to net harm to each of us who live in such a society.

Take the four outcomes any government decision can produce. The state can act, and act ‘correctly’ or ‘incorrectly’ (i.e. net positive or negative for the citizen body’s flourishing). It can decline to act, and the restraint can likewise be ‘correct’ or ‘incorrect’. Borrowing from statistics: Type I successes and errors when the state acts, Type II when it doesn’t. A regulation that prevents a real harm from occurring is a Type I success. The state has acted visibly, and it has acted in a way that is helpful. A regulation that causes harm greater than the one it prevented is a Type I error.The state has acted visibly, and it has acted in a way that is harmful. A regulation wisely declined is a Type II success. The state has acted invisibly (not acted), and it restrained itself in a way that is helpful. Some budget that should have been enacted but wasn’t is a Type II error. The state has not acted, and it has restrained itself in a way that is unhelpful. Four boxes that exhaust the space.

These four boxes are not symmetrically visible. Action is visible on both sides of its ledger. The bridge that stands, the drug that works, the program that delivers; and equally, the bridge that fell, the drug that killed, the program that failed. There is a referent. Voters can evaluate it. Journalists can investigate it. Opposition parties can campaign on it. Credit and blame attach to objects in the world.

But restraint… restraint rarely produces significant artifacts. A Type II success leaves nothing to point at: ordinary life continues, and ordinary life is never attributed to the regulation correctly declined. A Type II error is also invisible, but for different reasons. The cancer patient who would have been saved by a drug the regulator delayed by two years does not know which drug, does not know which months, does not know she is going to end up as another entry in Tabarrok’s graveyard. Her family attributes her death to the cancer, because the cancer is what they can see killed her.

So the political return on action seems positive or mixed: credit for Type I successes, blame for Type I errors, and across a career the balance tends positive, because the state exists in domains where the majority of the cost is unloaded across an, unorganised, unaware populace which might only occasionally cause backlash when macro-trends are analysed and even then, not on anyone specific, on a faceless and unaccountable mass; whereas the majority of the benefit is unloaded visibly to specific constituents. In many cases, because there is little communication between the government (government as multiple departments rather than a singular entity), there simply isn’t enough capability to even estimate net costs from the inside when most departments aren’t the ones gathering the tax, but spending it, or taking part in a chain of smaller organisations that collectively spend it. And then: the return on restraint. The return on restraint has to be zero or negative. No credit for Type II successes because they aren’t objects. Possible blame for Type II errors when the absence becomes salient after the fact. How many political careers have been built on the list of regulations the official chose not to write?

This requires no claim about the motives of the people inside government. They don’t need to be in a corrupt cabal, selfishly ambitious, or ideologically committed to big government. They need only respond to the local gradient of their environment, and the gradient runs in one direction. A bureaucracy made up of the best people you know, or the worst would similarly expand under this gradient.

Several features of modern political economy fall out of this directly.

Government grows monotonically. The standard explanations describe real mechanisms, but they share a puzzle: why do they only ever run one way? Why does capture always produce more regulation rather than less? The idea of visibility correctly shows us it is almost obvious that by the political economy of it, states can only fix themselves in the direction of more actions, not restraint.

Agencies expand their mandates after their original problems are solved. The FDA was created in response to the 1937 sulfanilamide tragedy. Modern pharmaceutical quality control makes a recurrence essentially impossible. Yet the FDA has grown continuously. What should it do? An agency that announced mission accomplished would lose its budget. An agency that found new problems gets to keep or expand it. Over decades the optimization produces an institution whose footprint has little relation to its founding justification, and which cannot be reduced even when individual reforms would be Pareto improvements, because each reduction is visible and each expansion compounded invisibly into the baseline.

Deregulation rarely sticks. A deregulated sector produces no visible successes attributable to deregulation: the cheaper flights and faster internet are experienced as ordinary consumption, not as policy outcomes. The first visible failure in a deregulated sector, however, gets read as deregulation’s failure. The S&L crisis, 2008, every airline bankruptcy: laid at the door of deregulation regardless of cause. The ledger punishes the deregulator coming and going. No credit for the market that now works, full blame for any market failure that would have happened under any regime.

Emergency powers become permanent. American income tax withholding, introduced as a wartime expedient, was never repealed. Wartime wage and price controls produced the employer-health-insurance system that still structures American healthcare today. Patriot Act surveillance provisions have been renewed continuously. COVID-era spending created baselines that subsequent budgets have not returned to. The reason is the same in each case: undoing an intervention is Type II action, invisible on the upside and blamed on the downside. The ratchet turns once.

The fifth feature is the consequential one, and the one Rothbard’s fable was actually illustrating. The asymmetry operates with different intensity in different domains. Where Type I errors are visible, where bad regulation produces shortages or defective products, the asymmetry is partially braked. The Eastern Bloc collapsed in part because empty shelves were legible to its citizens in a way that misallocated capital was not. British Leyland was abandoned because British consumers could see that British cars were bad and German cars were better. Visibility kept the brake on.

Where Type I errors are invisible or contested, the brake is gone. Monetary policy is the clearest case. When the Fed expands its balance sheet, the misallocation of capital produced is not visible in any single transaction. It is distributed across asset markets, delayed by years, and contested by serious economists on both sides. The Cantillon effect, by which new money enriches those who receive it first at the expense of those who receive it last, is a real phenomenon, but it does not produce identifiable victims with identifiable losses. The people who are poorer than they would have been under a different monetary regime do not know the counterfactual, and so they do not organize against the decisions that produced their relative impoverishment.

Land use is similar. A planning regime that prevents construction does not produce wrongly-sized shoes. It produces a housing stock smaller and more expensive than it would have been, but these are counterfactual facts, established only by comparing jurisdictions, and such comparisons are always contested by defenders of the local status quo. The error is hidden inside a counterfactual, and counterfactuals are politically weightless.

Pharmaceutical approval makes the pattern explicit. A drug approval that should not have happened produces, eventually, identifiable patients harmed by the drug. A drug approval that should have happened sooner produces patients who die of the disease they had, and their deaths are attributed to the disease. The FDA has institutionalized this asymmetry as a working principle, which is why its structural preference runs toward denial even when the science would permit faster action.

What these three cases share is a common structural feature: in each, the state has authority over an intangible. The price of money, the right to develop, the permission to sell a drug. Intangibles do not fail at the point of consumption the way shoes do. Their failures are diffused, delayed, and wrapped in counterfactuals. The state ratchets into domains whose Type I errors cannot be seen, because those are the domains where the ratchet has no brake.

Once you state it this way, a great deal of otherwise puzzling political behavior becomes legible. The strongest protections around an institution tend to be in place not because the institution is performing well but because its performance cannot be evaluated by the electorate. The NHS is more politically untouchable than a nationalized British shoe industry would be, not despite its performance problems but because its performance is harder to measure. The Federal Reserve operates at a level of insulation no agency responsible for visible consumer goods could sustain, not because its policies are more defensible but because its errors are less visible. The pattern is the structural product of a reward system that selects for opacity.

II. Why Shoes Are Easy and Money Is Hard

If the asymmetry were uniform across domains, you’d get government growth but not the particular shape modern government takes. The shape we actually see, a state that has retreated from physical goods and advanced into abstractions, requires the further observation that visibility itself varies. Some domains permit political systems to see failure almost in real time. Others permit failure only to be inferred decades later, through counterfactuals that defenders of the status quo can always contest. What separates the two?

Three things, and each of them separates shoes from money.

The first is the speed of the feedback loop. A failure in shoe production is observable within the period of a single purchase. The consumer who buys a wrongly-sized shoe knows immediately. The retailer with empty shelves loses customers that week. The manufacturer whose products fail loses orders next season. The signal travels from failure to consequence in days or months, through a chain that doesn’t require anyone to understand the underlying cause. Nobody needs to know why the shoe was wrongly sized; it is enough to observe that it was.

A failure in monetary policy travels through no such chain. The Fed’s decision to expand its balance sheet by several trillion dollars in 2020 and 2021 produced effects that became legible only partially, only contestedly, only over the following three or four years. The savers whose cash holdings lost purchasing power, the wage earners whose income lagged asset prices, the young buyers priced out of housing inflated by the cost of capital — none of them experienced the consequences as the consequences of monetary policy. They experienced inflation, unaffordable rent, a housing market they could not enter. Each of these was a real signal, but the signal did not propagate to the institution that generated it. The Federal Reserve doesn’t stand for election. It doesn’t lose customers. It doesn’t face return policies on its decisions. The feedback loop is either absent, or, where it exists through indirect political pressure, slow enough that causes are comprehensively contested by the time outcomes are felt.

The second feature is the presence of a real-time counterfactual. Consumer goods have counterfactuals built into the structure of the market. A Briton who wanted to know whether British shoes were good could compare them to Italian ones in the same shop. A Soviet citizen who wanted to know whether Soviet consumer goods were adequate could compare them, through informal channels if not legal ones, to the goods available in West Germany. The shoe that’s too narrow is too narrow relative to a standard that exists outside the system that produced it.

Many of the domains the state now occupies have had their counterfactuals suppressed. There is no parallel monetary system operating alongside the Fed against which its decisions can be benchmarked. There is no parallel British healthcare system operating alongside the NHS against which its outcomes can be compared without the comparison being dismissed as ideological. Cross-country comparisons exist, but they require analytical effort and trust in foreign data that most voters do not possess, and they’re easily deflected by appeals to institutional or cultural uniqueness. The counterfactual must be constructed, and construction requires expertise, and expertise is the property of the same class of people who defend the existing arrangement. The political system is left to evaluate the institution against itself, which is to say, against no standard at all.

The third feature is prices. Mises and Hayek made this argument in the calculation debate of the 1920s and 1930s, and it has not been improved on since. Mises showed that rational economic calculation is impossible without market prices for the means of production, which requires private ownership of capital goods. Without such prices, there is no cardinal measure for comparing alternative uses of resources; the planner is flying blind. Hayek’s later contribution was distinct: that the information needed to allocate resources well is dispersed, tacit, and held locally by millions of participants, and the price signal is the mechanism by which this knowledge gets aggregated into a form that coordinates their actions. The Soviet shoe factory could not calculate because no prices existed for its inputs, and could not learn what to produce because the knowledge held by retail workers, customers, and distributors could not be aggregated centrally. Both problems appear wherever prices are suppressed, and neither has a substitute mechanism anyone has yet devised.

Where prices operate, failure is legible even to people who don’t understand the underlying economics. The rising price of a good tells the consumer to economize, the producer to expand, the investor to enter. Errors correct themselves through the same mechanism that generated them. Where prices don’t operate, errors don’t correct. They compound, because the institution producing them has no way of knowing it is producing them. The Soviet planners who ordered the wrong quantities did not know they were ordering the wrong quantities. They received reports from retail outlets, but the reports arrived months late, were filtered through bureaucratic incentives to overstate demand or understate shortage, and could not be reconciled against any external benchmark. The system did not converge on the right answer; it drifted further from it.

Combined, these three features mean that visibility is a property of domains relative to the institutional regime operating in them, not of domains in the abstract. Shoe production under a market is a high-visibility domain on all three measures. Shoe production under a central planner is lower-visibility because prices have been abolished, but still higher-visibility than monetary policy because counterfactuals and feedback loops persist. Monetary policy under a central bank is low-visibility across all three: feedback is slow, counterfactuals are suppressed, and the object being managed is itself the thing that ordinarily generates prices.

The state’s footprint doesn’t reduce to a single category. Healthcare is a service. Monetary policy is an abstraction. Drug approval is a permission system. They don’t share a common position in the structure of production. What they share is that in each, the state has captured the layer at which the sector’s decisions get made, the mechanisms by which markets actually generate improvement over time.

There are three such mechanisms, related but separable. Hayekian price formation continuously aggregates dispersed information about scarcity and preference into signals that coordinate production. Where this is captured, the signal disappears; the NHS allocates £200 billion annually without any price at which to discover whether emergency capacity in Manchester should rise or fall next year. The specific failure is allocation failure: wrong quantities, wrong qualities, no internal means of correction. Kirznerian entrepreneurial discovery brings new providers in with innovations that drive productivity and quality. Where this is captured (through scope-of-practice laws, Certificate of Need regimes, accreditation gatekeeping), the organizational form of the sector freezes at the moment of capture, and the specific failure is technological stagnation: the Cuban taxi still runs, and it is still a 1957 Chevrolet. Competitive discipline is the third mechanism: providers who deliver poorly lose business to providers who deliver better, and the anticipatory threat of that loss keeps everyone from coasting. Where this is captured through legal monopoly or lock-in, existing providers coast, and the specific failure is quality degradation; the good gets worse over time because nobody has the incentive to prevent it.

Different sectors fail in different proportions. The Soviet shoe factory failed primarily on Hayekian grounds. American higher education fails primarily on competitive-discipline grounds, because Harvard’s applicants have no meaningful outside option for Harvard’s credential. British housing fails primarily on Kirznerian grounds. The NHS fails on all three at once, which is why its failure is the most complete.

This framing makes a sharper empirical prediction than the standard explanations for sectoral price divergence. Baumol’s 1967 cost-disease theory described the pattern Mark Perry has since documented in his widely circulated chart of American sectoral prices since 1975: hospital services, college tuition, childcare, and housing rising sharply in real terms; televisions, software, toys, and consumer electronics falling sharply. Baumol’s mechanism, that labor-intensive services experience rising relative prices as a growing economy bids up wages in sectors whose productivity hasn’t kept pace, is mechanically correct as far as it goes. But the interpretation laid on top of it, that healthcare and education are intrinsically labor-intensive and therefore destined to grow more expensive regardless of institutional arrangement, mistakes an outcome for a cause. Productivity growth in a sector is not a fact about the nature of the work being done. It is an outcome of the three discovery mechanisms. Where prices form, where entrepreneurs enter with better organizational forms, where competitive discipline forces existing providers to improve, productivity grows. Where these mechanisms are suppressed, productivity does not grow. The supposedly productivity-resistant sectors are productivity-resistant because they have been captured. LASIK surgery, by any reasonable definition a labor-intensive medical service performed by highly trained professionals, has seen its real price fall substantially over thirty years because the sub-sector escaped capture. Higher education has seen costs triple because it did not. Baumol described the outcome correctly. What produces the outcome is not the nature of the sector but the presence or absence of the discovery mechanisms.

Why is the state drawn to this layer rather than to final production? Because the discovery layer is where the decisions that matter actually get made. A regulator who sets safety standards for televisions gains limited power. A regulator who captures the payment mechanism for healthcare gains enormous power, because every decision about who gets treated, how, at what cost, and on what terms now runs through the political apparatus.

Capture at this layer is also stable as a political equilibrium. It simultaneously creates a constituency of incumbent providers whose revenue depends on the captured payer, eliminates the feedback mechanism that would have disciplined provider performance, and erases the counterfactual against which the capture could be evaluated, because the capture itself prevents the parallel system from existing. The politician who captures the discovery layer has, in one move, built a protected constituency, disarmed the opposition, and removed the evidence that would have condemned the capture.

It should be said that the politicians and officials acting on this gradient are rarely doing so cynically. Most of them are not calculating strategists hunting invisible-failure sectors to colonize. They are ordinary people who believe they are helping, and from inside the political system their actions look like helping. A politician who has not studied opportunity cost, who has not encountered Bastiat, who has not had occasion to think about what markets produce when left alone, will reliably experience the capture of a discovery layer as obvious moral progress. The constituency that benefits is visible to him. The counterfactual that would have revealed the cost is not. He acts on the gradient and feels virtuous, and the feeling is sincere. The selection pressure operates not against his moral intuitions but through them. A political culture that treats visible action as virtue and invisible restraint as indifference will produce politicians who genuinely believe action is what ethics demands, and those politicians will capture discovery layers in the full conviction that they are doing right. The conviction is not evidence against the framework. It is what the framework predicts.

The two case studies that follow illustrate what this produces. Coolidge shows what deliberate restraint looks like in a domain where the feedback features had not yet been fully dismantled. The NHS shows what the asymmetry produces when it has been allowed to compound for three-quarters of a century in a domain whose counterfactuals have been suppressed.

III. Coolidge, or What Restraint Looks Like

Calvin Coolidge is the closest thing the twentieth century offers to a controlled experiment in deliberate political restraint. He was not a libertarian in any doctrinal sense, and he would not have recognized the Austrian vocabulary in which his admirers now describe him. But he had a coherent theory of the presidency, articulated it consistently, and executed it with sufficient clarity over sufficient time that we can look at the record and ask what restraint actually produces.

The numbers first. Coolidge took office in August 1923 and left in March 1929. The federal budget was roughly $3.1 billion when he arrived; it was roughly $3.1 billion when he left. In nominal terms, across six years of strong growth, federal spending did not rise. The national debt, $22.3 billion in 1923, had fallen to $16.9 billion by 1929, a reduction of roughly a quarter, accomplished without crisis. The top marginal income tax rate was cut from 58 percent to 25 percent across the Revenue Acts of 1924, 1926, and 1928. Unemployment averaged under 5 percent through his term, closer to 3 percent by 1929. Real growth exceeded 4 percent per year. Consumer prices were effectively stable: inflation of 2.4 percent in 1923, 1.2 in 1925, 1.1 in 1928, with slight deflation in some intervening years.

These numbers should be read slowly, because they describe an outcome almost impossible to produce under any theory of governance other than the one Coolidge was following. A modern president of either party would regard any one of them as career-defining. Coolidge produced all of them simultaneously, did not claim much credit for most of them, and considered the whole enterprise unremarkable. If the federal government should go out of existence, he wrote, the common run of people would not detect the difference in the affairs of their daily life for a considerable length of time. That was not polemic. It was a description of what good government should feel like from the inside.

The mechanism was straightforward. Coolidge believed, and said publicly, that killing bad bills mattered more than passing good ones. He vetoed fifty measures, several of them twice. The most instructive veto was the McNary-Haugen bill, which would have raised domestic farm prices through a federal agency empowered to purchase surplus production and dump it abroad, financed by equalization fees on farmers. The bill had substantial agricultural-state support and passed both houses of Congress. Coolidge vetoed it in 1927 and again in 1928, with detailed messages each time explaining that central price-fixing through political agencies would produce overproduction, bureaucratic expansion, and eventual dependence. We should avoid the error of seeking in laws the cause of the ills of agriculture, he wrote. This mistake leads away from a permanent solution, and serves only to make political issues out of fundamental economic problems that cannot be solved by political action.