A new Federal Reserve study by Jason P. Brown, Elior D. Cohen, and Alison Felix looks at the effects of marijuana legalization.  Here is the abstract: We analyze the effects of legalizing marijuana for recreational use on state economic and social outcomes using difference-in-differences estimation robust to staggered timing and heterogeneity of treatment. We find moderate economic gains accompanied by some social costs. Post-legalization, average state income per capita grew by 3 percent, house prices by 6 percent, and population by 2 percent. However, substance use disorders, chronic homelessness, and arrests increased by 17, 35, and 13 percent, respectively. Early legalizing states experienced larger economic gains yet similar social costs, implying a potential first-mover advantage. Tyler Cowen discusses this study in Bloomberg: The researchers used appropriate statistical controls, but there is some question about causation vs. correlation. At the very least, it seems highly likely that state GDP went up: A state with legal marijuana can sell it, including to users in other states. Selling marijuana is a new business, and like any new business, it boosts the local economy. Due to the replication crisis in the sciences, it’s sensible to remain cautious about this sort of research.  But in this post, however, I’ll assume their findings are accurate. Let’s start with the fact that the estimated gains in income are huge.  To a non-economist, 3% may not sound very large, but it is.  The US defense budget is roughly 3% of GDP, and you rarely see people describe defense spending as small.  In contrast, the legal marijuana industry is tiny, well below 0.2% of GDP in California.  Therefore, this large a rise in income cannot plausibly be attributed to the direct effect of adding legal pot to a state’s GDP.  Instead, marijuana legalization seems to have produced some strong positive externalities—some combination of making workers more productive and adding to the number of workers.  If true, that’s a finding that we should be “shouting from the rooftops”. Tyler has mixed views on pot legalization, and in his Bloomberg piece he mostly emphasizes the negative: It would be hard to use this latest research paper to persuade people that additional drugs should be legalized as well. And I would not be surprised if some governments decided to end their experiments with marijuana legalization. Unless you are a responsible user, how exactly does it make you better off? Looking only at the practical issues, what is the case for legalization? Well, the study says it leads to higher incomes.  Yes, that seems unlikely.  But then why cite the study? The strongest argument for pot legalization is that it is cruel to send people to prison for selling or consuming pot.  After legalization in California, the number of people imprisoned for marijuana offenses fell dramatically.  On the other hand, the black market has not gone away, and thus the criminal justice benefits have been far less than they should have been.  That’s partly because pot remains illegal at the federal level, and this substantially increases the cost of doing business.  In addition, states have adopted legalization in such a way as to encourage the continuance of a black market.  There’s nothing special about marijuana that would make it more susceptible to a black market than are toasters or tee shirts.  The black market is almost entirely caused by burdensome regulation.  (Contrary to popular wisdom, taxes are not the main problem.)  The government may wish to restrict sales to people below a certain age and ban driving while under the influence.  Otherwise, it’s not obvious why there should be any regulation of pot production and distribution. So what would we expect from complete pot legalization?  Here are my guesstimates:  1. Some increase in a state’s population, but probably less than 2%. 2. No significant change in per capita productivity or income. 3. Some increase in both total usage and problem usage. 4. A substantial decrease in crime and punishment, much bigger than what we have observed thus far.  The black market would be almost completely ended, except for resale to underage teens.  (We’d have an even smaller black market than for cigarettes, which have higher taxes than pot.) (Note that for points #1 and #2 I am actually more pessimistic than the Fed study.  I believe they overstate the economic gains.) It is interesting to compare this list to the effects of alcohol legalization.  I suspect that alcohol has a much more negative impact on productivity than does pot.  It also seems likely that there is more problem usage of alcohol than pot, and that the health costs are greater. If society were serious about banning “bad things”, it might make more sense to start by banning alcohol.  Of course that experiment was tried, and the effects were roughly consistent with the pros and cons discussed above.  Banning alcohol reduced both consumption and problem consumption, and led to a big increase in crime and punishment.  The latter is a clear negative from prohibition, whereas the former is ambiguous.  Many people enjoy consuming alcohol, while heavier users suffer from some pretty severe consequences. I suspect that both the gains and losses from pot consumption are a bit lower than for alcohol. When I look at proposals to ban products such as alcohol, tobacco and pot, I see one massive negative consequence (more crime and punishment), and then some other effects that are hard to judge.  In 1933, the US public rejected alcohol prohibition, and is now beginning to develop the same view of pot prohibition—the policy has uncertain gains and huge losses. PS.  When I say “complete” pot legalization, I mean legalization at the federal level combined with state laws that are not more burdensome than the laws for selling beer.   (0 COMMENTS)

In April 1971, President Nixon was worried. Inflation had fallen from an annual rate of 6.2% in February 1970 to 4.2%, but unemployment, 3.5% in December 1969, hadn’t been below 5.9% since October 1970. Nixon’s ratings were tumbling; since the New Year, his approval rating had fallen from 56% to 49% and against his likely opponent in the 1972 presidential election, he had had flipped from leading 43-40 to trailing 47-39. He had to act. An avid sports fan, Nixon pulled an old favorite from his playbook. “Among the sharpest recalls of [Nixon’s] experience,” Theodore H. White wrote, was:  …the campaign against John F. Kennedy in 1960, and how the economy affected that campaign. As early as 1959, Nixon, then Vice-President, recognized the political danger as the second Eisenhower recession began. He had pleaded, early, in the Cabinet for an easy-money, pump-priming policy to get the economy moving before the election of 1960. His only ally had been economist Arthur Burns, but the Eisenhower administration had waited until late spring to loosen credit. By then it had been too late, for pump-priming requires long lead-time; and Nixon had been compelled to campaign against Kennedy with unemployment rising all across the country in the fall. He had lost. He did not want to repeat that experience in 1972. Now, time was running against him once more. ‘‘I’ve never seen anybody beaten on inflation in the United States”, Nixon said, “I’ve seen many people beaten on unemployment.’’ In one of Nixon’s famous secretly taped conversations, Arthur Burns, nominated by Nixon and now Federal Reserve chairman, told the president in February 1971: In my view the monetary authority…has laid the foundation for recovery…What is holding back the economy now not any shortage of money but a certain shortage of confidence. If we flooded the banks even more than we have I think you could have awful problems in 1972 and beyond.  But with unemployment stuck near 6%, Nixon continued pressuring Burns to ease monetary policy. “We’ve really got to think of goosing it…late summer and fall of this year and next year. As you know there’s a hell of a lag,” Nixon told Burns in March, but Burns replied that “To drive interest rates lower would run the risk of accelerating an international monetary crisis.” Changing tack, in July, Nixon discussed a possible vacancy on the Federal Reserve Board with Office of Management and Budget director George Shultz: I’ve told [Treasury Secretary John] Connally to find the easiest money man he can find in the country. And one that will do exactly what Connally wants and one that will speak up to Burns…and Connally is searching the god damn hills of Texas, California, Ohio. We’ll get a populist Senator [sic] on that Board one way or another…If you know of someone that’s that crazy let me know too…I want a man on that board that I can control. I really do. Basically that Connally can control. “To further pressure Burns,” Burton A. Abrams writes, “Nixon told his close advisors, John Ehrlichman and H. R. Haldeman, to leak a story through Charles Colson” – all key figures in the Watergate scandal – “about a recommendation to expand the Federal Reserve Board.” This “packing” would undermine Burns’ authority. Haldeman was also to leak that Nixon was considering legislation to curb the Federal Reserve’s independence.  On November 10, Burns folded, telling Nixon “Look, I just wanted you to know that we are reducing the discount rate today.” In December, Schultz told Nixon “[Burns] agrees that the money supply should now go up.” Later, Nixon urged Burns, “The whole point is, get [the money supply] up. You know, fair enough? Kick it!”  He did. Unemployment drifted down to 5.3% on election day. Nixon was reelected in a landslide, though this had as much to do with the Democrat’s missteps as anything else.  Wage and price controls announced in August 1971 muted inflation temporarily. But in September, Milton Friedman had warned Nixon that price controls “might be able to hold [inflation] down at least through the election…After this, you’ll have a great upsurge in inflation.” This soon materialized. The year-on-year inflation rate rose from 1.4% on election day to 4.9% when Watergate forced Nixon’s resignation in August 1974. One commodity after another surged against the dollar; soybeans, wheat, and finally oil, this latter, a consequence of inflation, being frequently mistaken as a cause even today. Unemployment was on its way up again, to a peak of 9.0% in May 1975. ‘Stagflation’ was here.  The American economy experienced the “awful problem” Burns and Friedman had warned about and the great inflation of the 1970s would not subside until the policies which caused it were reversed by Paul Volcker. Its origins lay not in the Arabian oilfields, but in Richard Nixon’s White House.           John Phelan is an Economist at Center of the American Experiment. (0 COMMENTS)

The arrest of a 19-year-old suspect who had allegedly planned a terrorist attack on a Taylor Swift concert in Vienna carries some lessons in view of what Auberon Herbert, a disciple of Herbert Spencer, wrote in 1894. The suspect, an Austrian citizen “with North Macedonian roots,” who is reported to have sworn allegiance to Islamic State, apparently intended to attack the crowd with knives and explosives. (See “Teenager Confesses to ‘Isis’ Attack Plot Against Taylor Swift Concerts,” Financial Times, August 8, 2024; and “Taylor Swift Terror Suspects Planned to Use Bomb-Filled Car at Concert, Authorities Say,” Wall Street Journal, August 8, 2024). Auberon Herbert was a British individualist-libertarian anarchist, although he called his philosophy “voluntaryism” and rejected the anarchist label. European anarchists of his time were mostly communists and often terrorists who blew up things to precipitate the revolution. In his Contemporary Review article, “The Ethics of Dynamite” (reproduced in The Right and Wrong of Compulsion by the State, Liberty Fund, 1978), Auberon Herbert argued that the “dynamiters” (the terrorists) were not really against government: Perhaps I ought at once, for the benefit of some of my friends who are inclined a little incautiously to glorify this word “governing” without thinking of all that is contained in it, to translate the term, which is so often on our lips, into what I hold to be its true meaning: forcing your own will and pleasure, whatever they may be, if you happen to be the stronger, on other persons. … Dynamite is not opposed to government; it is, on the contrary, government in its most intensified and concentrated form. … [Dynamite] is a purer essence of government, more concentrated and intensified, than has ever yet been employed. It is government in a nutshell, government stripped, as some of us aver, of all its dearly beloved fictions, ballot boxes, political parties, House of Commons oratory, and all the rest of it. How, indeed, is it possible to govern more effectively, or in more abbreviated form, than to say: “Do this—or don’t do this—unless you desire that a pound of dynamite should be placed tomorrow evening in your ground-floor study.” It is the perfection, the ne plus ultra, of government. Speaking of the terrorists, he harangued governments of his time: Here is your own child. This is what your doctrine of deified force, this is what your contempt of human rights, this is what your property in men and women leads to. This was at a time when Western governments were much less powerful than today’s. For sure, the latter are still far from beating at that game Isis and many other tyrants in the world. Still, we might reflect on a Senate hearing about another matter related to Taylor Swift but mainly about the government’s dirigiste crusade against so-called monopolies and market inefficiencies—a claim that, coming from the government, is hard to take seriously (“Senate Hearing on Ticketmaster’s Taylor Swift Meltdown: Five Takeaways,” Wall Street Journal, January 25, 2023): Live Nation Entertainment Inc. faced questions from lawmakers in Washington on Tuesday, in a Senate hearing stemming from Ticketmaster’s botched ticket sales last year for Taylor Swift’s coming tour. … Senators from both political sides of a divided Congress came together to criticize Ticketmaster, with Democrats and Republicans calling for a re-examination of Live Nation’s market power. Individual liberty and private property are required to prevent social strife or Leviathan. Let people who like Taylor Swift free to patronize her concerts, if she or others can finance them. Those who don’t like Taylor Swift or don’t like fun just have to abstain. And let people who want to go to Taylor Swift’s concerts buy their tickets from whoever organizes these concerts or acts as an agent of the organizers or has tickets to resell. (0 COMMENTS)

How should we think about the problem of unanticipated consequences? And what are the implications for the possibility of unintended consequences regarding top-down, technocratic policy initiatives that aim to mitigate targeted social problems?  For example, I’ve occasionally heard it argued that we shouldn’t be too worried about unanticipated consequences of interventions, because unanticipated consequences don’t have to be bad. They might be good! Albert Hirschman made this claim in his book The Rhetoric of Reaction, where he advanced two claims – the idea that “purposive social action” leads to adverse unintended consequences only “occasionally,” and that “it is obvious that there are many unintended consequences or side effects of human actions that are welcome rather than the opposite.”  In his book Power Without Knowledge, Jeffrey Friedman argued that Hirschman’s case falls flat on both points. To start, “Hirschman’s first claim is a generalization of naïve technocratic realism. It tacitly appeals to the reader’s agreement that if we tally up our first-order assessments of technocratic wins and losses, technocracy comes out ahead, begging the epistemological question by assuming the reliability of these tallies.” Given that the ability to accurately tally such things is the very point under dispute, trying to resolve the dispute by appealing to those tallies would indeed be a textbook case of question-begging.  The second claim Hirschman makes might provide a basis for defending technocracy, but Hirschman fails to adequately defend it, Friedman argues: To counteract worries about the adverse unintended consequences of technocracy he would have had to contend that the unanticipated consequences of technocrats actions will tend to be beneficial, not merely that they may be beneficial. Thus he would have had to argue not that “there are many unintended consequences or side effects…that are welcome,” but that, even though policymakers may be ignorant of the side effects of their actions, something or other ensures that these effects will be more welcome than unwelcome overall. This claim would not be naively realistic, as it would gesture toward a second-order factor or factors that might explain the on-balance beneficial valence of unintended consequences. However, since Hirshcman does not specify what this factor or factors might be, it is hard to imagine how the claim could be supported, saved through a quasi-religious providentialism.  That is, Friedman argues that if one wants to salvage the argument in favor of technocracy in situations where technocrats lack what Friedman called “type 4 knowledge” – knowledge that the costs of a technocratic policy (consisting of both the costs of implementing the solution as well as any unanticipated and unintended costs) will not be higher than the costs of the initial problem – merely pointing out that unanticipated outcomes could in principle be beneficial is simply inadequate. One would need to provide some positive grounds for believing that unintended consequences will have an overall tendency to be beneficial.  In his book, Friedman simply adopts the fairly modest premise that “while the tendency of unintended consequences might be either more harmful than beneficial or more beneficial than harmful, we do not know which is the case…The question, then, is whether our ignorance of the valance is more damaging to epistemological criticisms of technocracy or to defenses of it.” He argues that the simple fact of uncertainty is fatal to the argument for technocracy, and to say otherwise “would fly in the rationalist face of technocracy, for it would license the adoption of policies that – like policies pulled from a hat – are justified not by knowledge, but by hope.” Appealing to the mere possibility that unintended consequences might be beneficial as a defense of technocracy actually rebuts the argument in favor of technocracy. Friedman left the question of how to judge the valence of unanticipated consequences unexamined – his case didn’t depend on making a positive case that the valence will be neutral or even negative. But I want to look a step further than Friedman did – do we have reason to think that valence of unintended consequences will tend to be positive, neutral, or negative? And on what basis would we examine such a claim? Friedman argues (correctly, I believe) that we need to make a second-order argument on this issue. A second-order argument is one that focuses on systemic reasoning about the workings of a system, rather than first-order arguments where one attempts to tally up points on a case by case basis. For example, one might argue that government operates inefficiently compared to market activity by first-order means, perhaps pointing out that building a public restroom consisting of a “tiny building with four toilets and four sinks” cost the taxpayers of New York City over two million dollars, while by contrast “privately managed Bryant Park, in the middle of Manhattan, gets much more use and its recent bathroom renovation cost just $271,000.” But the same article also makes a second-order argument about the systematic differences under which state and private enterprises operate, arguing that since “government spends other people’s money, it doesn’t need to worry about cost or speed. Every decision is bogged down by time-wasting ‘public engagement,’ inflated union wages, and productivity-killing work rules.” So we can distinguish between the first order argument (examining specific cases) and the second order argument (comparative institutional analysis). Thus, the article uses a first-order case as an example of government being wildly wasteful and inefficient in what it does, and also offers a second-order argument for why this sort of disparity is systemic rather than random. In my next post, I will be considering a second-order argument about the valence of unintended consequences, and whether we should expect them to have a tendency to be positive, neutral, or negative.    (0 COMMENTS)

The claim in the title of this post might sound obvious. But I encounter lots of people that think import tariffs are taxes on imports, but not exports. In fact, taxes discourage both imports and exports, to a roughly equal extent.Just to be clear, tariffs might reduce imports a bit more than they reduce exports if they led to a smaller budget deficit. I doubt there’s anyone who believes that has been the case for the tariffs imposed in the US by recent administrations.Another misconception is that tariffs on imports reduce exports only in the case where other countries retaliate. That’s not true. Tariffs move the exchange rate to a position where exports are likely to fall roughly as much as imports decline, even if there is absolutely no explicit “retaliation” from other countries.The intuition here is that exports are the way we pay for imports. If you tax one side of a transaction you will reduce both sides, just as a tax on gasoline reduces both the sale of gasoline and the purchase of gasoline.  It makes no difference whether the tax is imposed on buyers or sellers.To be sure, the quantity of imports and exports of goods can differ if countries are also exchange financial assets in trade.  The trade balance (technically current account balance) is national saving minus national investment.  But unless tariffs reduce the budget deficit (which is negative saving), they are not likely to increase the trade surplus, or reduce the trade deficit.   Bloomberg has an article discussing how US farmers are losing market share to Brazilian farmers: An aging rural population is the latest strike against a country that’s been losing its agricultural dominance for years. That standing has been a crucial source of political power, including crucially with China, the biggest agricultural importer. But US-China relations frayed during Donald Trump’s trade war, allowing Brazil to take the place of some US supplies. Already the top exporter of soybeans, Brazil may now be on pace to overtake the US in corn exports, too. As the US’s agricultural trade deficit widens to a record $32 billion in fiscal 2024, households will find themselves at increasing risk of supply-chain disruptions and price spikes when far-flung disasters hit. A loss in agricultural export competitiveness is exactly what you’d expect when a country adopts higher tariffs.  It has nothing to do with “US-China relations”, and everything to do with the real exchange rate. None of this means that tariffs are necessarily a bad idea.  Rather this analysis suggests that it would be a mistake to move toward protectionism under the assumption that tariffs only reduce imports, whereas in fact tariffs reduce both imports and exports, and by a roughly equal amount.  PS.  The same applies to an export tax, which also reduces imports.  For the same general reason, a policy regime that combines a uniform import tariff with an equal export subsidy is pretty close to a pure free trade regime, as the two policies roughly offset.  Thus countries like South Korea were far less “mercantilist” during their high growth phase than many people claim. (0 COMMENTS)

Was it divine intervention or simply the fact of being right that makes the infamous “Letter from Birmingham Jail” timeless, profound and even life-changing for some readers? If you haven’t met Dwayne Betts in an earlier EconTalk episode, get ready for Russ Roberts’s phenomenal friend and guest. Betts is so present in the moment and affected by the beauty, truth, and humility of the great Martin Luther King that his voice sometimes cracks answering Roberts’s questions. In this episode, he shares moments from his own history and the effect King’s work has had on him. Betts’s 9 years in prison and remarkable journey since uniquely qualify him as the King family’s choice author for the introduction to Letter from Birmingham Jail (The Essential Speeches of Dr. Martin Luther King). We hope this conversation stirs thoughts about freedom in you. Please share a thought or insight in the comments below.     1- Both Roberts and Betts have appreciated King’s great speech in different ways upon revisiting it. Roberts calls it a love letter to justice in his (King’s) country. As you pause and read “Letter From Birmingham Jail,” by Martin Luther King Jr., what do you notice that perhaps you didn’t remember?   2- How does Betts argue that King’s urgent letter, a response to criticism of his nonviolent protesting, honored the eight clergy critics?    3- Betts states, “I feel like it’s much more challenging to name what the side of justice looks like,” referring to the difficulty of arguing with conviction on contemporary topics. To what extent do you agree with this statement and, with what examples would you explain?   4- “Turning regrets into feathers” versus “Economics explains everything except justice”.  John Rawls (not Robert Nozick) is in The Freedom Library in 340 prisons. Confronting the sense of “nobodiness,” which conversation strand would you want to pursue over dinner, and why? (0 COMMENTS)

Should the Fed placate the markets? Yes and no. Let’s start with the no. Today’s Bloomberg has a piece by Mohamed El-Erian with the following title and subtitle: The Fed Should Resist Placating MarketsThe central bank needs to avoid being rushed into another policy mistake by making an emergency interest-rate cut. The Financial Times has a similar piece by Barry Eichengreen: The Federal Reserve will not let markets dictate a rate cut Stock moves are not a reliable signal of looming economic downturn I mostly agree with both commentators.  The Fed should not be in the business of trying to prevent big moves in the stock market, and yesterday’s 3% decline in the S&P500 was not even a particularly large move.   (Yes, it was significantly larger than average, but I’ve seen numerous moves that were far larger.  As I write this, the S&P500 is up over 2%.)   So why do I say “yes and no” at the beginning of this post?  I think it depends on exactly what one means by “placate the markets.”  Imagine there were a NGDP futures market.  In that case, I would strongly support having the Fed adopt a monetary policy that placated the NGDP futures market.    Of course we do not have an NGDP futures market.  But we do have many markets that indirectly provide information as to market expectations of NGDP growth.  Start with the fact that NGDP growth is the sum of inflation and real GDP growth.  And then note that we have (admittedly imperfect) market indicators of expected inflation.  In addition, there are many market indicators that are substantially correlated with expected real and nominal growth.  Yesterday I recall a market commentator mentioning that risk spreads in the bond market had increased.  Risk spreads are certainly correlated with NGDP growth, as borrowers have more trouble servicing debt when NGDP growth slows sharply.   Suppose the Fed constructed a model to estimate market expectations of NGDP growth, which used a weighted average of all sort of relevant market prices.  It might make sense to try to stabilize that index, without trying to stabilize any single individual component of that index.  Would that be “placating the markets”?  I think that’s sort of a question of terminology.  The Fed would not have market stability as a primary goal; rather they would merely be trying to stabilize markets to the extent that doing so would stabilize NGDP growth.  As a practical matter, they might occasionally respond to severe stock or bond market movements, but not because they cared about the plight of investors. (1 COMMENTS)

In the last 5 weeks, I’ve written two articles on the draft for the Hoover Institution’s on-line publication Defining Ideas. The first made the case against the military draft; the second made the case against universal national service. In responses on the Defining Ideas site, some commenters argued that one advantage of the draft is that it causes people who benefit from defense to have “skin in the game.” In response to my first article, one commenter wrote: Our freedom is not for free. David Henderson wants those who are prepared to risk their lives for our freedom to do that for the benefit of those who want their freedom for free. In response to my second article, one commenter wrote: When American men do not serve their country, they put no skin in the game and, as a result, do not feel that they are obliged to fight and defend. Actually, though, if the goal is for beneficiaries of defense to have skin in the game, an all-volunteer force does a better job than the draft. Why? The reason is that the draft puts a disproportionate burden on draftees. An all-volunteer force, on the other hand, spreads the burden to beneficiaries of defense whether or not they are in the military. In the late 1970s, there was a serious push, spearheaded by Senator Sam Nunn (D-GA), to restore the draft. I got copies of all of the bills to do that. Every single one of them—and there were many—explicitly cut first-term pay, often by a large percent. Why pay when you can threaten potential draftees with prison sentences for not complying? So the burden would have been placed disproportionately on those who were drafted. Consider, by contrast, an all-volunteer force. The reason the military had problems recruiting high-quality personnel in the late 1970s was that we had an economic boom combined with high inflation. It was a double whammy. The boom gave potential recruits good alternatives to military service; failure to raise pay in line with the Consumer Price Index made military service even less attractive than otherwise. President Jimmy Carter got wise to the situation relatively late in his 4-year stint in the White House and, with Congress, raised first-term pay. Then Ronald Reagan became president and raised it again. That’s how we got out of the late 1970s recruiting doldrums. So note what happened. Because we had a volunteer military, the burden of defense couldn’t be shifted onto the shoulders of young military personnel. Instead it was shared by all taxpayers. We saw something similar in the middle of the 2000s, during the second war against Iraq. Here’s what I wrote in September 2015, drawing on a scholarly article co-authored with then Marine Major Chad W. Seagren: Henderson and Seagren note that, as the number of troops in Vietnam increased from 1964 on, real military personnel outlays per military member barely budged. By contrast, real military personnel outlays per member rose substantially as the U.S. government got in wars in Afghanistan and Iraq. From an average of $73,887 per member between 1996 and 2001, real outlays rose to an average of $103,772 from 2004 to 2010, an increase of 40 percent. The reason: the government had to increase pay to meet its manpower targets. Henderson and Seagren point out that this higher cost per military member resulted in about an extra $45 billion per year in U.S. government spending. That higher cost was, admittedly, financed mainly with deficits rather than with current taxes. But deficits now, unless the government later defaults or cuts spending, lead to higher taxes in the future. And if, as seems likely, the future tax system even roughly resembles the present tax system in forcing higher income people to pay a much higher percent of their income in taxes, the rich and powerful will pay more for war. The bottom line is that if you want all people who benefit from defense to have skin in the game and not just focus on a small group, you should oppose the draft and favor an all-volunteer military. Postscript: In researching this piece, I came across this Econlib article by Chad Seagren, “Service in a Free Society,” May 2, 2011. I had lined it up and edited it during my time as editor of the Econlib articles. I had forgotten about it. It speaks to many of the issues with the draft, and does so well. (0 COMMENTS)

Our new model incorporates several innovative features: For example, rather than using a representative household, it features a demographically accurate synthetic population with millions of households (matching age, education, race, and consumption habits). Instead of using a representative firm, we model the behavior of tens of thousands of the largest firms, in one-to-one correspondence with real firms…. J. Doyne Farmer, Making Sense of Chaos: A Better Economics for a Better World,1 p. 258 Mainstream economics is willing to build a model of an economy in terms of a “representative individual.” One hypothetical consumer represents every household. One hypothetical firm represents every business. Many different types of workers are aggregated as “labor.” Many different types of machines and other productivity-enhancing factors (such as business reputation or process knowledge) are aggregated as “capital.” I have long questioned this way of doing economics, which I refer to as the “GDP factory” method of analysis. For decades, J. Doyne Farmer and his relatively small cohort of like-minded researchers have advocated for and implemented a different approach. Borrowing from the field of ecology, they wish to build models incorporating agents that employ different strategies within the overall system. The representative-individual approach involves carefully choosing a set of assumptions in the economist’s head about human behavior, representing those as equations, and solving the equations for a single equilibrium. It predates the age of the computer. Farmer’s approach, illustrated in his new book Making Sense of Chaos, requires a very different modeling strategy, called “agent-based modeling.” It starts with observations about how different individuals choose strategies for earning a living, consuming, and investing. The goal is to see how these strategies interact with one another over time. This requires computer simulation. For example, consider the stock market. The “representative individual” approach assumes a single investor with full information and one strategy for maximizing returns relative to risk. Farmer’s approach instead starts by looking at the types of strategies different investors actually use. Some focus on fundamentals. Others try to spot trends. Everyone has different information and uses different heuristics. For more on these topics, see “The Stock Market,” by Robert P. Murphy. Library of Economics and Liberty, Oct. 4, 2010. John Bogle on Investing. EconTalk. Luca Dellanna on Risk, Ruin, and Ergodicity. EconTalk. The representative-individual models of the stock market tend to have dynamic properties that are uninteresting and unrealistic. They predict minimal market movement, much less trading than we observe, and nothing like the pattern of run-ups and crashes that seems to characterize existing markets. The models with heterogeneous investors are able to replicate the patterns we actually observe in the stock market. One of the most interesting findings from representative-agent models is that as the influence of players using one strategy increases, the dynamics of the financial market change. Strategies that dampen volatility for a while can suddenly cause instability. For example, Farmer points out that in the late 1990s major investment banks adopted “value at risk” (VaR) as a strategy for controlling market exposure. VaR measures the loss from, say, an adverse price movement of two-standard deviations. Using such a metric, a risk manager would say that you can increase risk exposure as market volatility declines, and you have to decrease it when volatility goes up. In good times, you get a self-reinforcing feedback loop that raises asset prices as banks expand their portfolios. But then a little adversity leads everyone using VaR to try to sell at once, causing a really severe self-reinforcing loop on the downside. Farmer says that this describes what happened in financial markets before and during the financial crisis of 2008. Farmer and colleagues also have used computer simulations of heterogeneous-agent strategies to analyze the energy market, with a particular focus on trying to assess the feasibility of an “energy transition” to forestall climate change. According to their analysis, the main cost from shifting toward renewable energy sources is upgrading the electric grid. But actually producing energy will be cheaper, so that overall a faster energy transition is a positive for the economy. In 2050, for example, our estimated global annual expenditure on the electricity network for the Fast Transition is about $670 billion per year, compared with $530 billion per year for the No Transition. However, the expected total system cost in 2050 is about $5.9 trillion for the Fast Transition and $6.3 trillion per year for the No Transition. Thus, although the additional $140 billion of grid costs might seem expensive, it is significantly less than the savings that come from cheaper energy. p. 253 The mainstream approach to doing economic theory will always have the advantage of being easy to communicate and to replicate. When someone shows the results of a mainstream model, you can solve the equations yourself and get a feel for what is driving the results. For empirical work, replication is not so reliable. Farmer reports that when he was with a company that was interested in exploiting stock market inefficiency, his team looked at published papers on market anomalies. For around half of the papers, we couldn’t reproduce the results, even when we tested the postulated deviation from efficiency using the same data. p. 146 “If economists are going to adopt agent-based modeling, they are going to have to develop ways to articulate, explain, and justify the choices they make in constructing the models.” Simulations are more opaque to those of us who are not on the team that built the model. We cannot reproduce the results for ourselves. If economists are going to adopt agent-based modeling, they are going to have to develop ways to articulate, explain, and justify the choices they make in constructing the models. I think of economic models as being like maps. With an old-fashioned triptych, if the map said to take the George Washington Bridge to get from where I live to Boston, I would have been stuck with that. With the map on my smart phone, I can consider alternatives, and even make adjustments in real time based on traffic conditions. For economists, vast amounts of data are becoming available. Computer power has gone up by orders of magnitude. This presumably makes the trends favor agent-based modeling relative to the representative-individual standard. But as maps for policy makers, agent-based models are still far from reliable. I would be careful not to presume that they make centralized decision-making a good way to operate an economy. One should not bet the farm on Farmer. Footnotes [1] J. Doyne Farmer, Making Sense of Chaos: A Better Economics for a Better World. Yale University Press, 2024. *Arnold Kling has a Ph.D. in economics from the Massachusetts Institute of Technology. He is the author of several books, including Crisis of Abundance: Rethinking How We Pay for Health Care; Invisible Wealth: The Hidden Story of How Markets Work; Unchecked and Unbalanced: How the Discrepancy Between Knowledge and Power Caused the Financial Crisis and Threatens Democracy; and Specialization and Trade: A Re-introduction to Economics. He contributed to EconLog from January 2003 through August 2012. Read more of what Arnold Kling’s been reading. For more book reviews and articles by Arnold Kling, see the Archive. As an Amazon Associate, Econlib earns from qualifying purchases. (0 COMMENTS)

A Book Review of The Anxious Generation: How the Great Rewiring of Childhood Is Causing an Epidemic of Mental Illness, by Jonathan Haidt.1 When an academic writes a book for a popular audience, one of their main goals is to have an impact on the world. Jonathan Haidt’s new book The Anxious Generation is clearly already having an impact. The book has been sitting near the top of the New York Times bestseller list since it was released. But it’s already having an impact where academics care the most: public policy. Consider the following set of recommendations that Arkansas Governor Sarah Huckabee Sanders recently gave in her state-of-the-state address:2 Experts suggest goals like no smartphones before high school; no social media before 16; phone-free schools; and more outdoor play and childhood independence. I listened to that speech at the same time I started reading Haidt’s book. Here is how Haidt summarizes the suggestions in the conclusion of his book (page 290): 1. No smartphones before high school 2. No social media before 16 3. Phone-free schools 4. Far more unsupervised play and childhood independence It seems quite clear that, either directly or indirectly, the expert that Sanders is getting her recommendations from Haidt. It’s almost word-for-word. She prefaced her list with this context: Study after study shows that too much social media exposure leaves our kids anxious and depressed. Suicide rates for young teens have tripled since 2007. Depression among teenagers is up 150%. 30% of teenage girls now seriously consider suicide. This paragraph could easily have been a blurb for Haidt’s book. Clearly, Haidt is having an impact, and a very immediate impact on national conversations, that most academics could only dream of from a lifetime of doing research on a topic and trying to inject it into the public debate. Given Haidt’s clear impact, it is useful to examine his book in detail, and the research that he (and others, like Governor Sanders) cites to support the conclusion that the mental health issues teenagers are experiencing in America (and elsewhere) are the direct result of social media use on smartphones. And states are not waiting for more research to take action. Even before Sanders’ speech this year, Arkansas has already been at the forefront of restricting social media use for teenagers, with the 2023 “Social Media Safety Act” which requires age verification to use social media, and requires parental consent for anyone under age 18. While I have not seen Haidt comment on the Arkansas law, he praised a similar Utah law and said that every state should do the same. From Free Play to Instagram Addiction Haidt’s story of generational decline is neatly summed up by his four recommendations from above, but in reverse order. It starts with number 4 on his list, the decline of “free play.” For most of human history, kids had a lot of unsupervised play with kids their own age. Haidt tells us that this is an important part of childhood development, both in terms of learning how to interact with others (and resolve disputes) and to correctly wire young brains. Starting in the 1980s, parents began to give kids in the United States and other countries less time to engage in free play, in response to real and perceived threats to children in the world, driven by media stories about child abductions and murders. That’s part one of the story. But the real problem doesn’t arise until decades later in Haidt’s telling. The decline in free play left a void in child development, but also a void in how children spent their time. One way that time was occupied was with more intensive parenting, as parents started spending a lot more time actively involved with their kids (rather than their kids playing with other kids). Structured “play” time was also increasingly introduced, such as organized team sports, music and dance lessons, and all variety of clubs for kids to engage their interests. There were, of course, screens to occupy the time of kids. Lots of screens: TVs, VCRs and DVDs, video games, computers, computers with the internet. Then in the mid-to-late 2000s, a new and dangerous form of screen entered the scene: the smartphone. Around the same time, social media began to grow in influence, among both young and old. 2010 is a crucial year: the newest iPhone adds a front-facing camera and Instagram is introduced. After 30 years of mulling around with the decline of free play, kids finally found something to really occupy their time: spend 5 hours a day perfecting their online persona. The lack of actual social connections with other young people, and the replacement of it with pseudo-, often toxic, social connections online has led to a crisis of mental health. Haidt documents in chapter 1 the “tidal wave” of suffering in the United States, especially among young people, and particularly among girls. This suffering includes not just self-reported or diagnosed mental health issues (e.g., 30 percent of teen girls with major depression, more than doubling since 2010), but obviously objectively bad outcomes, such as suicide, rising 167 percent in a decade for girls ages 10-14 in the United States. A crisis, no doubt. The cause? Haidt is convinced and spends the rest of the book trying to convince the reader, that the double-whammy of declining free play in the 1980s and the rise of smartphones and social media in the 2000s is the culprit. Screens Are Annoying Any parent will tell you that screens are a double-edged sword. On the one hand, screens provide additional learning and entertainment opportunities, as well as temporary distractions for kids when adults need a break or need the kids quiet. This was true from TVs all the way up through the iPad (also introduced in 2010). One of the biggest downsides of screens is that they are annoying. Most of the entertainment is pretty mindless, whether it was the Honeymooners in the 1950s or Cocomelon today. What’s worse is kids become dependent on the screens. They whine for them when they don’t have them, and they seem unable to sit still in normal social settings. Kids never misbehaved in the past, of course (I’m only half kidding). Screens are not just disruptive for young people: they often disrupt adult social interactions too. The annoyance of screens, though, is not what Haidt’s book is about. Haidt is concerned very narrowly on smartphones with access to social media; this is clear from his first three recommendations. And he’s primarily concerned with mental health, not annoyance per se. Because Haidt is a good social scientist, he demands the best evidence, for he knows that two things can often follow each in time trends but either be unrelated, or both be caused by some third factor. Is It Causal? “Haidt argues very strongly that the relationship between social-media-equipped smartphones and teen mental health is causal, not a mere correlation. What is the evidence?” Haidt argues very strongly that the relationship between social-media-equipped smartphones and teen mental health is causal, not a mere correlation. What is the evidence? For a book that asserts this so heavily, you might be surprised to learn that there is just one paragraph focusing on randomized-control trials that address this question. It can be found on page 148. In that paragraph he summarizes just two studies that attempt to measure the effect of social media on mental health. In a footnote, he tells us further that there are 14 RCTs showing harm, and another 6 that found no harm (but he regards these 6 as low quality studies), and then points us to an online Google Document that he put together with his collaborators. There are multiple documents that he has put together on his website that relate to the research behind this book, but the one on Social Media and Mental Health runs 356 pages, longer than the text of the book itself! Haidt is to be applauded for putting this all online transparently, but as a social-science nerd, I would have liked to see this take up more than a single paragraph of the book. A few chapters perhaps? But that probably wouldn’t have landed the book on the New York Times bestseller list. Anyway, back to the 356-page Google Document. The discussion of RCTs begins on page 168 (of the current version as I write—this is a living document) and runs for 20 pages. There are now 23 studies showing causal negative effects, and another 8 studies showing no effect, 11 more studies than when the book went to print just a few weeks ago. I won’t dive into all 31 of these studies, but given that Haidt is laser-focused for both the trends and policy recommendations on teenage girls, how many of the RCT studies would you guess are about teenage girls? The answer: just one. The other papers study college undergraduates, adults, or young adults. It’s not that there is nothing we can learn about teenage mental health by studying people older than them. But what’s so striking about this fact—just one study of teenage girls!—is that Haidt is so confident in his overall hypothesis despite the evidence being so razor thin. And what of this one single study? Also interesting: it wasn’t even teens in the United States, but rather in the Netherlands. Again, there is nothing wrong about studies outside of the United States, but as I will argue below, the worst of the mental health problems seem confined to teenage girls in the United States, yet we have no studies of teenage girls in the United States. The paper in question is well done. It randomly assigns girls to two groups, and one group shows manipulated Instagram photos that make the subjects more attractive. The girls that received the treatment reported lower body satisfaction, about 0.4 points on a 9-point scale. This result is statistically significant, but… is it enough to worry us? Is it good proof that social media is causing a mental health crisis, when you have a small change on the scale of body image from one single study of 144 teenage girls in the Netherlands, with no follow-up for long-term effects? This seems, to me, to be a very weak reed to build an entire apparatus of restricting phone use for teenagers around the world. Is It Really Happening Everywhere? While Haidt spends much of the book discussing evidence from the United States, he suggests that this is a global phenomenon. In Chapter 1, he spends three pages expanding his charts on teen mental health to other English-speaking countries (Canada, the United Kingdom, Australia). He has another two and half pages on “the rest of the world,” but the evidence here is pretty thin: a chart on psychological stress in Nordic nations, and a chart on alienation in school by broad regions (Asia, Europe, and English-speaking Latin America). As with the summary of studies, there are other Google Docs to consult from Haidt and collaborators. For example, while the Nordic nations get just one paragraph and one chart in the book, the Nordic adolescent mood disorders online document runs over 100 pages. And much of that document is less certain and ambiguous than the text of the book. While there is plenty of evidence of rising mental health diagnoses and self-reports, evidence on self-harm and suicide doesn’t show increases. In some cases, it shows decreases. In Denmark, self-harm was reduced by almost 50 percent from 2007 to 2016 among teenage girls and boys—there had been a rise in the decade before 2007, but it came back down after that. Teenage suicides in Sweden exhibited a similar pattern, with a rise from 2000 to about 2008, then coming back down. To go beyond the Nordic Nations, another Google Doc (Haidt is very thorough and transparent) on Adolescent mood disorders is useful. But the studies summarized in Section 3.6, looking outside of the Anglosphere (United States, Canada, United Kingdom, Australia, New Zealand) are not very convincing. The first study they summarize looks at suicides rates for ages 15-24 from 2006-2017 in several high-income countries. The only clear increases are in the Anglosphere, and even if we limit the analysis to girls, only Spain is added to the unfortunate group of rising youth suicide. France and Italy are declining, Germany and Poland are flat. There is no evidence that social media and smartphones have proliferated less in those countries than the Anglosphere. They also look at OECD data, but it is no more promising for their hypothesis: “Teenage suicides rates have, on average, declined slightly over the past two decades or so.” There does seem to be something particularly bad happening in the United States and other large English-speaking countries, but our non-English-speaking peer nations aren’t seeing the same trends (though some are seeing the rise in reported mental health issues). What Is to Be Done? We can think of the question “What is to be done?” in two ways. Haidt is convinced that the evidence is overwhelming on the connection between social media, smart phones, and teen mental health. If you are also convinced, the thing to be done is find policy solutions or suggest changes in social behavior. But the second way to think of “What is to be done?” is to think about what further research needs to be done to better understand the relationship between social media and teen mental health. Perhaps the two questions can be merged: targeted policy interventions could also produce good research results, to be used for future potential interventions. On page 263-64, Haidt suggests just that merging of the questions, when he proposes that state educational authorities set up random-assignment of schools into one of four groups, such as phone-free, free-play, both, and a control group (status quo policy). What’s really important here is to note that no such studies exist or Haidt would have cited them. He is really making recommendations without much good evidence yet. But as Haidt notes earlier in the book (page 249), 77 percent of schools in the United States already say that they ban phones—they just aren’t enforcing the bans. So clearly schools have this power (as they do have the power to limit all sorts of student behaviors and activities), they just aren’t using it. Haidt says: use that power but use it in a way that we can learn from it. If the governors of, say, Utah, Florida, and Arkansas (three states that have passed some restrictions on youth social media use) took this opportunity to conduct randomized experiments on schools, other states could learn from their experiments. It may seem cruel to treat school children as test subjects, but that’s actually we are already doing, we’re just doing it poorly and in a way that is hard to study the actual effects. Unfortunately, it does not seem that any well-done studies have tried this randomized approach yet (you may have heard about a new paper supposedly on phone bans in Norway, but Haidt acknowledged on social media that the paper “does not really tell us much”). What’s a Parent to Do? Most of Haidt’s recommendations aren’t clearly directed at anyone, but rather are directed at everyone. The “no social media before 16” recommendation could be a call for laws. But the book can also be read as advice to parents. Indeed, Haidt’s final chapter before the conclusion of the book is pitched as advice to parents. Haidt correctly identifies the nature of the issue: this is a collective action problem. Deciding on their own, most parents would probably follow Haidt’s recommendations about delaying the use of smartphones and social media. But one family alone making this decision is in a difficult spot, given that most other families are allowing their kids to use smartphones and social media. How to back out of this social dilemma? For more on these topics, see Emily Oster on the Family Firm. EconTalk. Emily Oster on Cribsheet. EconTalk. Emily Oster on Pregnancy, Causation, and Expecting Better. EconTalk. Bryan Caplan on Parenting. EconTalk. “Parenting Tips (and Other Helpful Advice) for Economists,” by Michael L. Davis. Library of Economics and Liberty, Apr. 5, 2021. Haidt has a lot of advice for parents! Not all of it is backed by randomized controlled trials, but most of what we do as parents doesn’t have strong evidence (but thank you to Emily Oster for trying to write several books which do bring together the evidence that exists). He breaks down the evidence by the age of your children, and certainly some it is involves less screen time (or more productive uses of screens), but much of the advice is just good parenting advice. For example, for kids ages 6-13, Haidt recommends things such as encouraging sleepovers, walking to school, free play after school (instead of “enrichment activities), camping, sleepaway camps, and forming child-friendly neighborhoods. These suggestions have nothing to do with screens (though some are to avoid the temptation of screens), and are all good advice regardless of whether social media is causing a mental health crisis. Parents are always making choices under uncertainty. That’s the nature of parenting. But hopefully public policy requires better evidence for imposing rules on the entire population, especially when public policy is precisely the area that has the power to conduct controlled experiments, so that we may possibly find out what is going on with youth mental health in the English-speaking world. Footnotes [1] The Anxious Generation: How the Great Rewiring of Childhood Is Causing an Epidemic of Mental Illness, by Jonathan Haidt. Penguin Press, March 26, 2024. [2] “Governor Sanders Delivers State of the State Address”. Online at governor.arkansas.gov. *Jeremy Horpedahl is Associate Professor of Economics at the University of Central Arkansas. He blogs at Economist Writing Every Day. For more articles by Jeremy Horpedahl, see the Archive. As an Amazon Associate, Econlib earns from qualifying purchases. (0 COMMENTS)