First, the major categories of US government spending have been shifting. Consider this graph showing federal spending divided into mandatory, discretionary, and net interest. The vertical axis measures spending as a share of GDP. As you can see, over the last half-century, the category of “mandatory” is way up, the category of “discretionary” is way down, and the category of “net interest,” while small, is at its highest level since the start of the CBO data in 1940.
What these categories? “Mandatory” spending arises when the level of spending was predetermined by earlier legislation. A little more than one-third of spending in this category is related to health care: Medicare, Medicaid, and subsidies to make health insurance more affordable. A little less than one-third is Social Security. Other large categories are income support programs like the earned income tax credit, the child tax credit, food stamps, and others. Other items in this category include retirement programs for government workers and supports for military veterans.
In contrast, CBO notes: “Discretionary spending includes most defense spending; spending for many nondefense activities, such as elementary and secondary education, housing assistance, international affairs, and the administration of justice; and outlays for certain transportation programs.”
A second long-term change involves a shift between “primary” and overall budget deficits. The “primary” budget deficit is just the overall deficit, with interest payments subtracted out. As the figure from CBO shows, the projected “primary” deficits for the US government don’t look especially high by historical standards. But notice two shifts. First, in the last 50 years the primary deficit would rise and fall, and occasionally dip into becoming a primary budget surplus. Looking ahead in the next decade, it’s all primary deficits. Second, the higher interest payments mean that the gap between the primary and overall deficits has widened. The danger here is what I’ve called the “interest payments treadmill,” where the annual deficits stay large because of the high interest payments on past borrowing, which raises total debt in a way that guarantees even larger interest payments in the future.
For awhile back around 2010, there was a lament among Democratic-leaning politicians and economists that they didn’t “go big” when raising spending in response to the Great Recession. When the pandemic recession hit, some of them were determined to “go big” this time. Personally, I think the government fiscal boost at the time of the Great Recession was about right, but the fiscal boost in response to the pandemic recession was too much. But whatever our judgements about the past, we now face the bills in the form of high interest payments.
A third substantial change is the projection that federal tax revenues as a share of GDP are going to remain fairly close to their historical levels during the last half-century, while federal spending is moving to a higher level.
The main reasons for the shift in spending have already been mentioned: the rise in spending on the elderly as the Baby Boomer generation retires, a steady rise in health care costs, and the higher interest bills from previous government borrowing. There are fundamental choices to be made here. One option is for federal spending overall to rise in response to the growing share of the elderly in the US population. If we don’t want federal spending as a share of GDP to rise, then either we need to cut back on benefits to the elderly, or cut back on other federal spending, or raise taxes. Given that cutting back on interest payments is not a good idea, the “other spending” that can be cut has already been a shrinking share of the US budget for a few decades. Raising taxes isn’t any fun, nor is an ever-growing federal debt.
When it comes to the federal budget, we are headed in the next decade or so for a situation where either we either make choices at times of our own choosing, or we have choices forced upon us quite likely at times not of our own choosing.
I have been the Managing Editor of the Journal of Economic Perspectives since the first issue in Summer 1987. The JEP is published by the American Economic Association, which decided more than a decade ago–to my delight–that the journal would be freely available online, from the current issue all the way back to the first issue. You can download individual articles or entire issues, and it is available in various e-reader formats, too. Here, I’ll start with the Table of Contents for the just-released Winter 2024 issue, which in the Taylor household is known as issue #147. Below that are abstracts and direct links for all of the papers. I will probably blog more specifically about some of the papers in the few weeks, as well.
______________
Symposium on the Global Market for Talent
“Skilled Immigrants, Firms, and the Global Geography of Innovation,” by Britta Glennon
This article begins with an overview of the policy environment in the United States and abroad for skilled immigration, with a particular focus on “supply-driven” versus “demand-driven” systems. The overview emphasizes that firms play a central role in the skilled immigration process in most countries. I then survey the ample evidence that skilled immigrants have a strong positive effect on firm outcomes, followed by a discussion of the many margins of adjustment that firms have when their access to skilled immigrants is affected by national immigration policy. Finally, given such margins of adjustment and the importance of skilled immigrants to firms, I consider how the policies that affect skilled migration shape the global geography and quality of innovation. I conclude by discussing policy implications and open questions. In particular, I emphasize that evaluations of the impact of skilled immigration should not be constrained within borders: immigration flows and national immigration policies affect the global geography of innovation and investment.
“Migration and Innovation: Learning from Patent and Inventor Data,” by Francesco Lissoni and Ernest Miguelez
Research on international migration and innovation relies heavily on inventor and patent data, with “migrant inventors” attracting a great deal of attention, especially for what concerns their role in easing the international transfer of knowledge. This hides the fact that many of them move to their host country before starting their inventive career or even before completing their education. We discuss the conceptual and practical difficulties that stand in the way of investigating other likely channels of influence of inventor’s migration on innovation, namely the easing of skill shortages and the increase of variety in inventive teams, firms, and location.
“Tax Equity in Low- and Middle-Income Countries,” by Pierre Bachas ⓡ Anders Jensen ⓡ Lucie Gadenne
Income inequality is high and persistent in developing countries. In this paper, we ask what role taxation can or might play in reducing inequality in low and middle-income countries. Drawing on the recent literature, three findings emerge. Due to both structural factors and limited enforcement capacity, the effective distributional impacts of taxes often deviate from their ‘statutory’ objectives, in ways that are hard to predict based on evidence from high-income countries. Moreover, administrative reforms which are meant to be distributionally neutral end up having significant equity impacts because of the practical realities of implementation. Finally, the global challenges which tax authorities face to tax the very top of the income distribution appear to be even more pronounced in developing countries. We conclude by offering thoughts on future research and emphasize the need to carefully study equity characteristics of taxes at each stage of a country’s development path.
“How Can Lower-Income Countries Collect More Taxes? The Role of Technology, Tax Agents, and Politics,” by Oyebola Okunogbe and Gabriel Tourek
Increasing tax revenues is a major policy goal in many low- and lower-middle-income countries. While economic growth is an important determinant of taxation, available evidence indicates that it does not automatically increase taxation. Rather, countries must make targeted investments in their tax capacity. In this paper, we examine the rapidly growing body of evidence on different interventions to improve tax capacity and increase tax revenues in lower income countries, with a focus on two key inputs: information technology and tax officials. We examine the role and limitations of digitization for identifying taxable entities, verifying tax liabilities, and ensuring collection of tax owed. We also consider how the deployment and incentives of tax officials shape their performance, and the interplay between them and technology tools. Lastly, we emphasize the importance of political incentives and consider the conditions under which governments choose to invest in tax capacity and expand tax collection.
“Does the Value-Added Tax Add Value? Lessons Using Administrative Data from a Diverse Set of Countries,” by Anne Brockmeyer ⓡ Giulia Mascagni ⓡ Vedanth Nair ⓡ Mazhar Waseem ⓡ Miguel Almunia
The value-added tax (VAT) is a cornerstone of the modern tax system. It has many desirable properties in theory: it does not distort firms’ production decisions, it is difficult to evade, and it generates a substantial amount of revenue. Yet, in many countries there are discrepancies between the textbook model of the VAT and its practical implementation. Where the VAT implementation diverges from its textbook model, the tax may lose its desirable properties. We draw on firm-level administrative VAT records from 11 countries at different income levels to examine the functioning of real-world VAT systems. We document four stylized facts that capture departures from the textbook VAT model which are particularly pronounced in lower-income countries. We discuss the effects on VAT performance and simulate a counterfactual retail sales tax and a turnover tax. Despite its shortcomings, we conclude that the real-world VAT is superior to the alternatives.
“The Failure of Silicon Valley Bank and the Panic of 2023,” by Andrew Metrick
The failure of Silicon Valley Bank on March 10, 2023 brought attention to significant weaknesses across the banking system, leading to a panic that spread to other vulnerable banks. With subsequent failures of Signature Bank and First Republic Bank, the United States had three of the four largest bank failures in its history occur over a two-month period. Several features of the Silicon Valley Bank failure make it an ideal teaching case for explaining the underlying economics of banking (in general) and banking crises (specifically). This paper tries to do that.
Countries around the world are enacting pay transparency policies to combat pay discrimination. Since 2000, 71 percent of OECD countries have done so. Most are enacting transparency horizontally, revealing pay between coworkers doing similar work within a firm. While these policies have narrowed coworker wage gaps, they have also led to counterproductive peer comparisons and caused employers to bargain more aggressively, lowering average wages. Other pay transparency policies, without directly targeting discrimination, have benefited workers by addressing broader information frictions in the labor market. Vertical pay transparency policies reveal to workers pay differences across different levels of seniority. Empirical evidence suggests these policies can lead to more accurate and more optimistic beliefs about earnings potential, increasing employee motivation and productivity. Cross-firm pay transparency policies reveal wage differences across employers. These policies have encouraged workers to seek jobs at higher paying firms, negotiate higher pay, and sharpened wage competition between employers. We discuss the evidence on effects of pay transparency, and open questions.
“Immigration and Crime: An International Perspective,” by Olivier Marie and Paolo Pinotti
The association between immigration and crime has long been a subject of debate, and only recently have we encountered systematic empirical evidence on this issue. Data shows that immigrants, often younger, male, and less educated compared to natives, are disproportionately represented among offenders in numerous host countries. However, existing research, inclusive of our analysis of new international data, consistently indicates that immigration does not significantly impact local crime rates in these countries. Furthermore, recent studies underscore that obtaining legal status diminishes immigrants’ involvement in criminal activities. Finally, we discuss potential explanations for the apparent incongruity between immigrants’ overrepresentation among offenders and the null effect of immigration on crime rates.
“Care Provision and the Boundaries of Production,” by Nancy Folbre
Whether or not they provide subjective satisfaction to providers, unpaid services and non-market transfers typically contribute positively to total output, living standards, and the social climate. This essay describes some quantitative dimensions of care provision and reviews their implications for the measurement of economic growth and the explanation of relative earnings, including the gender wage differential. It also calls attention to under-explored aspects of collective conflict over legal rules and public policies that shape the distribution of the net costs of care provision.
“Job Training and Job Search Assistance Policies in Developing Countries,” by Eliana Carranza and David McKenzie
Governments around the developing world face pressure to intervene actively to help jobseekers find employment. Two of the most common policies used are job training, based on the idea that many of those seeking jobs lack the skills employers want, and job search assistance, based on the possibility that even if workers have the skills demanded, search and matching frictions make it difficult for workers to be hired in the jobs that need these skills. However, reviews of the first generation of evaluations of these programs found typical impacts to be small, casting doubt on the usefulness and cost-effectiveness of these programs. This paper re-examines the arguments for whether, when, and how, developing country governments should undertake job training and job search assistance policies. We use our experience with policy implementation, and evidence from recent impact evaluations, to argue that there is still a role for governments in using these programs. However, success depends critically on program design and delivery elements that can be difficult to scale effectively, and in many cases the binding constraint may be a lack of firms with job openings, rather than a lack of workers with the skills to fill these openings.
There’s a stereotypical hero in any number of movies and books, who is standing alone against the big project that is going to ruin the environment, ruin the community, or both. (In a slightly alternative version, the big project has already started to ruin the environment or the neighborhood.) If the hero is not already a lawyer or a journalist, they often ally with a lawyer or journalist in exposing the truth, before it’s too late. My point here is that the hero is blocking something from happening, and the US legal and regulatory system is decentralized in a way that creates multiple potential blocking points: multiple regulatory agencies, multiple levels of courts, multiple options for media communication.
But what happens if the storyline requires a hero who can get something built? For example, someone who can build a dramatic expansion of solar and wind-power, or electrical transmission lines, or housing that is more dense and affordable, or a mass transit system? A US-based hero of this story will now be pinned down and tormented by multiple blocking points.
In 1921, the U.S. Department of Commerce, under its then-Secretary Herbert Hoover, supported the formation of an Advisory Committee on Zoning. The Advisory Committee’s charge included aiding communities interested in the “promotion of the public welfare and the protection of property values” … The Committee published two documents in 1922: A Zoning Primer (Primer) and A Standard State Zoning Enabling Act, Under Which Municipalities May Adopt Zoning Regulations (Enabling Act; U.S. Dept of Commerce, Advisory Committee on Zoning, 1922). … Fischel states, “Before 1910, there was not a single zoning ordinance in the United States. By 1930, it had spread to all sections of the country” (2015: 170). Zoning ordinances had been adopted in 8 cities by the end of 1916, another 68 cities by 1926, and an additional 1,246 municipalities by 1936, constituting 70 percent of the U.S. population (Fischel, 2015: 171).
The phrase that Hoover’s commission wanted to act in support of “promotion of the public welfare and the protection of property values” is of course thought-provoking, because it does not seem to acknowledge that there may be cases where these goals could conflict. Most of the symposium involves papers with a US focus about housing regulation. But I found myself especially interested in the final essay by Paul Cheshire, “An International Perspective on the U.S. Zoning System.” Here’s the overall perspective:
Zoning (or planning) has important functions. Markets play a fundamental role in efficiently allocating urban land (Bertaud, 2018), but there are endemic problems of market failure. There are also conflicts of interest in land use—between owners of undeveloped and developed land and between local interests and the wider society. If ‘rule-based,’ planning can also reduce uncertainty and development risks. In planning systems, the level to which decisions are rule-based, discretionary, or reflect local or wider societal interests varies globally. Internationally, the U.S. system is among the most locally controlled but significantly rule-based because of the use of zoning. In contrast, in the United Kingdom and a range of other countries, local politicians largely decide on development on a case-by-case basis. More local control and discretionary decisions increase the power of the “not in my backyard,” or NIMBY, interest because development costs are highly localized, but benefits range over a wide area, even a whole country. This process tends to end with generally restricted development, resulting in higher housing and land costs. This problem is increasingly visible on both U.S. coasts. Local control also enables zoning systems to protect the interests of insiders and exclude those below the poverty line, for example, by applying extravagant minimum lot sizes or zoning for single-family housing. More recently, attempts have been made to use planning to reduce carbon emissions or force mixed communities. The evidence suggests that zoning is unsuited for achieving either objective. …
The planning system common to Continental Europe, the Master Planning system, is more clearly rule-based, prescriptive, and detailed than the U.S. zoning system. Uses for every parcel are planned, and permission to develop is virtually automatic if the plan and any other relevant regulations are followed. In countries such as Germany, France, or the Netherlands, plan formulation and decision control has an important element, which is national, or at least regional. The U.S. and U.K. systems are at the local end of the spectrum—the U.S. system by design and legal foundation and the U.K. system because an elected committee of the lowest tier of government, the Local Authority Planning Committee (LAPC), is the primary decision-making body. A national policy framework and often local plans exist, but the reality is that enforcement is weak to absent, so any local decision not flagrantly in breach of national policies is likely to stick.
These differences in zoning regulations, and the degree to which they are rule-based or discretionary, and local or national, will be reflected in the cities that emerge. Cheshire offered this interesting comparison of skyscraper heights:
Lake Michigan may constrain Chicago on one side, but a rigid growth boundary and height restrictions have constrained London wholly since at least 1955. Per head of population, however, there were nearly seven times as many skyscrapers—buildings more than 100 meters—in Chicago than in London. Even Paris has significantly more skyscrapers per capita than London. The only tall-building league London tops is the proportion of its skyscrapers designed by Trophy Architects (TA), architects who have won one of the internationally recognized lifetime achievement awards in architecture. Of London’s skyscrapers, 25 percent were TA-designed compared with 3 percent in Chicago and zero in flexibly regulated and rule-based Brussels.
Careful analysis demonstrates that although Chicago may have been the birthplace of great modern architecture, any competent architect can get permission to build a skyscraper there if it meets the zoning regulations and building standards (Cheshire and Dericks, 2020). With London’s discretionary planning, employing a TA seems to help developers generate a powerful signal of design quality, providing a passport to political approval and a bigger building. In London, TA-designed buildings are 17 stories taller than non-TA-designed buildings, increasing a representative site value by 144 percent. Also, buildings designed by an architect after winning a lifetime achievement award increased between 13 to 17 floors (depending on model specification) compared with those the same architect had designed before receiving the award. In Chicago, an architect gaining TA status did not affect the height of their buildings.
This analysis might not seem to be important, but it represents a serious, albeit difficult to observe, deadweight economic cost—estimated as £59 million ($75 million) for a representative site in the City of London—an extra cost symptomatic of an unpredictable planning system that injects opportunities for gaming the system (rent-seeking) and additional risk into the development process.
Of course, what we would all prefer is a perfect process by which any and all building projects can be clearly divided into the desirable and the undesirable, and our zoning rules will sort them out accurately. But back here on planet Earth, the practical questions involve a generalized willingness to block or to allow changes, when in many cases there is unlikely ever to be a consensus about desirability. In the US context, the question is whether we are ready–at least in some cases–to give up the narrative that the blockers of projects are necessarily also the heros.
I see a lot more fire engines than I do fires, but as an economist, I trust statistics more than my own two eyes. The National Fire Protection Association collects the relevant data. Shelby Hall authored the most recent version of “Fire Loss in the United States” (November 1, 2023). The author notes: “On average, a fire department responded to a fire somewhere in the US every 21 seconds in 2022. A civilian was fatally injured in a fire every two hours and 19 minutes. Every 40 minutes, a civilian suffered a non-fatal fire injury.”
However, longer-term trend of fires (based on fires reported to local fire departments) has generally been downward–although with a definite upward bump in 2022. Here are some basics:
Despite the fall in fires over time, the number of firefighters has climbed along with the general rise in population, so that the rate of firefighters per 1,000 people has remained the same. This data is from a different report from the National Fire Protection Association by Rita Fahy, Ben Evarts and Gary P. Stein, called “US Fire Department Profile 2020” (September 2022).
With fires and fire deaths down by half since 1980, what are the firefighters doing? Many fire departments have for a long time included health-related events, not just fires, in their basic mission. But as the number of fire-related responses by fire departments has declined, the number of health-related responses has risen substantially. Here’s the data from Fahy, Everts, and Stein (the provide annual data, but here, I’m just comparing 1980 and 2020):
As you can see from the above figure, the number of career fire-fighters has expanded by about 50% over the last 40 years (from about 240,000 to 360,000). In that time, the number of fire department calls has more than tripled (from about 11 million to over 36 million). However, the number of calls related to fires has dropped by more than half, reflecting the decline in number of fires. Medical aid or rescue was already a more common fire department activity back in 1980, but it has expanded quite rapidly. Fire department also end up involved in a number of other situations like downed power lines, disaster relief, or even bomb threats. What it means to “be a fire-fighter” has been evolving over time.
The euro technically started in 1999, when the 11 founding European members of the currency agreed to keep their exchange rates fixed and to hand over monetary policy to the European Central Bank. The euro then became the actual currency that people and firms used in 2002. I confess that, back in the early 1990s, I did not expect the euro ever to happen. My logic was simple: I reasoned that the euro would not take off without Germany, and Germany would not surrender the deutschmark. I was wrong.
As the authors note, “the euro effectively sailed on with an incomplete constitution.” That is, when the euro began it had the European Central Bank, a promise from the member countries that they wouldn’t run overly large budget deficits, and a “no bailout” pledge if they did. But the consequences of violating these promises and pledges were unclear. It wasn’t clear to what extent the new monetary order would be enforced from above, or would bubble up from below. It wasn’t clear what would happen in a debt or financial crisis. It was not clear if there would be a “safe asset,” similar to US Treasury bonds, with the full backing of the euro area, or just separate bonds from different countries. There was no centralized European budget.
But there was a sense that if the European Union was to be an economic success, with free movement of workers, goods and services, and capital across national borders, the euro was part of the solution. Indeed, Europeans had for some decades had various agreements to limit or block movements in their exchange rates, so to at least some, the euro just seemed to formalize earlier arrangements and make them permanent. And indeed, for the first 10 years or so, the euro worked remarkably well. It was “the 2% decade,” with the economies of the euro-zone countries on average grew about 2% per year, with annual inflation staying low at about 2%, and average government budget deficits across the euro area at about 2%.
Then it went sideways. The European Union was first hit by the Great Recession of 2008-9, with many EU countries having their own versions of credit and housing bubbles and financial crisis. But for a time, global credit markets were pricing debt from all EU countries at very similar levels: that is, countries that seemed to have worse problems with credit bubbles, bank failures, and government debt were paying pretty much the same fairly low euro-based interest rates as everyone else. As a result, these higher risk countries (Greece, Portugal, Spain, Italy, Ireland, others) just kept dramatically over-borrowing.
Around 2010, the EU powers-that-be made clear that the neither the EU nor the European Central Bank was standing behind such loans. The interest rates for the most debt-ridden EU economies spiked. A decade followed of debt rescheduling, bankruptcies, emergency loan packages, and uncertainty. For the decade from 2009 to 2019, the annual GDP growth rate for the euro-area countries was only 0.8%–which implies that a number of countries had growth rates of zero or less during this period.
Corsetti and Buti go over the many, many summits and announcements and policy proposals through this difficult decade for the euro. Looking back on it now, I would emphasize that the problem wasn’t just slow growth and a sense of slow-motion crisis in the euro area, but a sense that the countries within the euro area were diverging. The author provide this useful figure, comparing the “interquartile range” of unemployment rates across US state and the EU-15 countries: that is, it’s the range of unemployment rates from the state or country at the 25th percentile up to the state or country at the 75th percentile.
Notice that the interquartile range for US states is comparatively small. The range across the EU countries looks as if it’s getting smaller for the first decade of the euro, but then appears to be getting much bigger from about 2012-2015. The gap then declines to a smaller, but comparatively still large, levels.
But by 2020, as the EU was having some success in gradually building institutional structures to deal with sovereign debt issues and to support the euro and the European Central Market, the pandemic hit. From the standpoint of the euro, the pandemic had two big effects: one clearly positive and one potentially negative.
The positive effect was that the pandemic offered a crystal-clear case for economic coordination and support, along with additional institution-building, across the countries of the EU. The negative effect was that fiscal deficits across the euro-area countries spiked as they sought to reduce the economic shock of the pandemic, and together with supply-chain problems, the reality of too much spending power chasing too few goods led to the euro’s first experience with widespread inflation, averaging 7% in 2002-2003. As Corsetti and Buti point out, the euro-area seems to innovate only in times of crisis:
[A]gainst all odds, EMU has proven to be resilient … The political drive underlying its creation, which seemingly withers away in normal times, resurfaces powerfully, especially when crises threaten the survival of the common currency. Indeed, historical records confirm the leitmotiv in the EU narrative: the ‘true reaction function’ of Europeans emerges only in conditions of extreme distress. But the same records also show that steps forward only come at higher-than-necessary and social costs. Looking forward, to keep counting on the idea that the right decisions are (eventually) made only under distress is risky. At 25, the key challenge for the euro area is to learn to design and implement the necessary reforms in ‘normal times’.
Some steps are happening. For example, during the pandemic the EU issued bonds backed by the European Union as a whole, not by individual countries, with the proceeds used to support economies and labor markets. Thus, countries could be less tempted to run huge budget deficits on their own. There is also discussion of European-wide funding of European public goods, like certain kinds of infrastructure or reductions in carbon emissions. The EU is working on a “banking union,” where there will be a common set of rules and supervision across all EU banks. (And yes, the euro was launched without a common set of bank rules or euro-wide banks supervisors.) A more general “capital markets union” is under discussion. It’s now clear that the European Central Bank will play a role in addressing financial crises. (And no, that wasn’t clear when the ECB was created.)
The euro was an incomplete work-in-progress when it started, which is part of why skeptics like me could barely believe it. But while the push toward further European integration has its pauses and jolts, the forward momentum continues, which means that the institutions surrounding the euro continue to evolve, as well.
For those who would like to dig more deeply into these issues, farther than the Corsetti-Buti discussion will take you, I can recommend a couple of symposium from the Journal of Economic Perspectives, where I work as Managing Editor. The Spring 2021 issue included a four-paper Symposium on the European Union:
Megan T. Stevenson is an active researcher in the criminal-justice-and-economics literature. She has also noted a disconcerting fact: When you look at the published studies that use randomized control trial methods to evaluate ways of reducing crime, most of the studies don’t show a meaningful effect, and of those that do show a meaningful effect, the effect often isn’t replicated in follow-up studies. She mulls over this finding in “Cause, Effect, and the Structure of the Social World” (forthcoming in the Boston University Law Review when they get around to finalizing the later issues of 2023, pp. 2001-2027, but already available at the Review’s website).
(For those not familiar with the idea of a “randomized control trial,” the basic idea is that a group of people are randomly divided. Some get access to the program or the intervention or are treated in a certain way, while others do not. Because the group was randomly divided–and you can check in various ways whether it appears to be random–a researcher can then compare the outcomes between the treated and untreated group. This method is of course similar to drug trials, when you randomly divide up a group and some get the medication while others get a placebo. This approach is sometimes called a “gold standard” methodology, because it’s straightforward and persuasive. But of course, no method is infallible. One can always ask questions like: “Was it really random?” “Was some charismatic person involved in the treatment in a way that won’t carry over to future projects?” “Was the sample size big enough to draw a reliable result?” “Did the researcher study a bunch of treatments, on a number of groups, but then only publish the few results that looked statistically significant?”)
As one example of the evidence on interventions to reduce crime, Stevenson writes (footnotes omitted):
In 2006, two criminologists published a survey article of every RCT over the previous fifty years in which: (1) there were at least 100 participants, (2) the study included a measure of offending as an outcome, and (3) the study was written in English. The authors uncovered 122 studies, evaluating interventions such as:
Counseling/therapy programs;
Criminal legal supervision, including intensive probation;
Scared-straight programs;
Work/job-training programs;
Drug testing, substance abuse counseling, and drug court;
Juvenile diversion;
Policing “hot spots”; and
Boot camps.
Note that these interventions include those associated with a tough-on-crime framework (e.g., scared-straight programs and boot camps) as well as those that provide support and resources (e.g., work/job training programs and counseling). Note further that inclusion in this analysis required that the study was written up and disseminated so it could be discovered by the survey authors—a filter that is likely to have eliminated many of the nonstatistically significant results already. Nonetheless, only 29 of the 122 studies (24%) found statistically significant impacts in the desired direction.
Stevenson reviews a number of more recent studies as well. But the likelihood of successful results remains low, and worse, the chances that a successful result is not replicated by a future study seems high.
As Stevenson points out, this finding is reminiscent of what Peter Rossi several decades ago called: “The Iron Law of Evaluation: The expected value of any net impact assessment of any large scale social program is zero.” Here, I don’t want to quarrel over whether their might be a few strong counterexamples to Stevenson’s pessimistic evaluation. Instead, what does Stevenson suggest should be learned from this discouraging pattern of findings? I’d paraphrase her arguments this way.
While it’s an attractive idea that a relatively small treatment will fundamentally alter an unpleasant outcome like crime (say, a job-training program or “hot-spot” policing), there are often underlying reasons why people make the decisions they do. Stevenson writes: “That doesn’t mean that human actions never have an impact, but rather that the type of discrete, limited scope interventions that are the primary domain of empirical causal inference research generally have limited or nonreplicable impact.”
The positive effects of some policies may be so obvious that they don’t get studied by a randomized trial. For example, feeding the hungry accomplishes a goal of feeding the hungry. One might study other possible effects of such a policy on crime or labor force participation or family dynamics, and that’s where the randomized control trial doesn’t reliably find positive effects. But the hungry did get fed. Stevenson writes:
There is an old cliché that if you give a man a fish, he will eat for a day; if you teach him how to fish, he will eat for a lifetime. Such sentiments form the basis of many of the interventions discussed in this study. These interventions, designed to give people the resources to thrive on their own, rarely have large or lasting impact. The cliché is wrong, at least when it comes to the limited-scope, systems-conserving interventions. However, there remains a straightforward and obvious way to ameliorate harm: simply give people what they need. If they are hungry, give them food. If they need shelter, give them a home. If they need work, give them a job.
The effects of certain policy choices may never get studied by a randomized control trial, because the policies are so sweeping. Perhaps changing people’s lives requires a group of policies sustained over a long period of time, and then evaluated after an even longer period. When people call for “systemic” change, they presumably have in mind a set of changes that can’t be captured by dividing up a group at random and treating one part of the group in a specific but limited way. But of course, systemic change can be very hard to evaluate in advance, and can have either good or bad outcomes.
Finally, Stevenson is asking the social science research community about whether it is overemphasizing the “gold standard” method of randomized control trials, rather than perhaps seeking out evidence from real-world experience. Her sense is that researchers may tend to follow the randomized control trial methodology because they think it is more likely to result in published papers, rather than because it’s the best way to get a persuasive answer. To put it another way, persuasive evidence for a policy can come from a variety of methods, and randomized control trials are only one of those methods.
Stevenson’s paper made me think of a recent wave of research on some of the social programs implemented several decades ago. For example, the food stamp program was rolled out, county-by-county, over the period from 1961 to 1974. The order in which counties were selected was determined by practical and political considerations, and for practical purposes can be viewed as largely random (that is, no particular group was systematically overrepresented in being covered earlier by the foot stamp program). This is sometimes called a “quasi-experiment,” referring to the idea that some families were randomly eligible for food stamps and others were not, but that pattern wasn’t designed by anyone. However, a researcher can come along later and take advantage of the randomization. In this case, it turns out that children under the age of five who were in counties that got food stamps earlier had positive long-term effects in adult health, earnings, and lower crime rates, among other factors.
The short answer is that the possibilities are both real and limited. The authors write:
Not every office property, however, is suitable for conversion [to apartments]. Three conditions must be met for a conversion to take place: (i) the building has to be physically suitable for conversion, (ii) the zoning and building codes have to permit and facilitate such a conversion, and (iii) the financial return of the conversion has to properly compensate the developer for the risk they are taking.
The authors have data on commercial office markets in core urban areas across 105 cities. (They acknowledge that that some office buildings outside the urban core might also be suitable for conversion, but it’s not their focus here.) Using that data, they consider what kinds of office buildings are most suitable for conversions. They write:
We believe that buildings built before 1990 are the most viable conversion candidates. Many historic buildings tend to be less expensive, have smaller floor plates, and have more character, all of which increases their conversion appeal. … The size of the building cannot be too big or too small, so we exclude buildings with a total size less than 25,000 square feet as well as large buildings with deep floor plates. Smaller buildings could be convertible, but they are less likely to attract institutional capital and federal grants. Deep floor plates have existing floor plans that start the building at a disadvantage: Too little interior light and air, too little plumbing, and too many elevators. Structural changes to remedy these buildings for residential use are likely cost prohibitive. … We narrow our sample of candidates further by selecting buildings with no (or few) major long-term leases left.
After carrying out this exercise, what’s left? They identify 2,431 properties out of a total of 22,215 office buildings in these 105 cities–with about one-fifth of all the possible properties being in New York City. They estimate:
At 875 square feet per apartment unit, and after incorporating a 30 percent loss factor, these conversions could create 158,654 additional housing units. Scaling up for incomplete data coverage results in 367,750 apartment units. For comparison, about 260,000 apartment units were created in the U.S. in a typical year between 2001 and 2022.
Of course, not all of these potential conversions will be financially viable, either–a determination that would vary across cities and properties. But in a very broad sense, it seems reasonable to say that office-to-residential conversions might be equal to about a year’s worth of standard apartment-building for the nation as a whole. Thus, on one side, every city should be looking over its inventory of office buildings and figuring out which ones might be suitable for conversions. On the other side, such conversions are likely to make only modest progress on the goal of additional housing.
The Bureau of Labor Statistics reported an increase in union membership, with 139,000 more union members in 2023 than in 2022, meaning this country has 400,000 more union workers than we had in 2021. The gains under the Biden-Harris administration underscore President Biden’s commitment to being the most pro-worker, pro-union president in history. We have seen large private sector increases in unionization among health care workers, transportation and warehousing workers, and in educational services. These are workers who recognize that they have power and are organizing to use that power. Workers in health care, auto manufacturing, transportation, entertainment and more have delivered big wins at the bargaining table in the past year.
2023 was a banner year for labor actions and unions. “Hot Strike Summer” morphed into “The Year of the Union” as a strong job market, more than a decade of intensive labor organizing kicked off by the “Fight for 15” and a growing recognition of the need for worker protections through the ongoing pandemic helped drive major wins for workers. Striking United Auto Workers, writers, actors, UPS workers and Kaiser Permanente health care workers secured strong contracts that include benefits like increased wages, health care access and job protections, while U.S. rail workers secured paid sick leave for a large segment of their workforce. Workers at Starbucks continued to gain momentum towards their first contract in defiance of blatant union-busting tactics, building on years of organizing work and hundreds of successful store votes. In part because of these high-profile strikes, unions saw near record highs of public support in 2023, with two-thirds of people approving of unions, more than 6 in 10 saying unions help the U.S. economy and one-third of people predicting unions will be stronger in the future. Workers’ efforts were coupled with the Biden administration’s success in reinvigorating the National Labor Relations Board, the federal agency dedicated to protecting employees’ rights to organize and addressing unfair labor practices.
Remember that during this “banner year for labor actions and unions,” the share of US workers who actually belong to a union was shrinking–and has been shrinking for decades, including last year and in fact during the entire presidency. It’s true that in general, public attitudes seem more supportive of unions. But some of the union successes in the last few years, like the first success in unionizing an Amazon warehouse (on Staten Island), have since become mired in controversy and seem in danger of failing.
Maybe we’ll all look back on 2023 as the year union membership in the US bottomed out, and the beginning of a great union resurgence, but I doubt it. A couple of years ago, Suresh Naidu wrote “Is There Any Future for a US Labor Movement?” in the Fall 2022 issue of Journal of Economic Perspectives, where I work as Managing Editor. Naidu is sympathetic toward unions, but also clear-eyed. For example, he points out that old-style union organizing outside a large physical facility is not going to work well in an economy where many people are working from home, or doing gig jobs. He points out that a decline in unionization may also reflect a broader decline in “social capital” of people acting together in a variety of contexts. He discussed a range of organizations that try to speak for worker interests in a systematic way (like the movement to raise the minimum wage to $15 per hour) without actually being unions.
But Naidu also points to an even more fundamental issue that US unions face: they need to organize one company at a time. In a dynamic US economy, where some companies are always shrinking or going out of business, this means that unions are running on a treadmill: they need to keep organizing new unionized companies just to offset the typical year-to-year loss of previously organized companies. Naidu writes:
In the United States and other establishment-level bargaining systems, a basic constraint on union density is that it is hard to organize new firms fast enough to keep pace with the exit of already unionized firms. Even if unionization were an order of magnitude easier, the costly trench warfare of establishment-by establishment organizing in the face of structural change and natural business dynamism makes keeping union density constant, let alone expanding it, an uphill battle.
Naidu cites a study from a couple of decades ago, when about 13% of American workers belonged to a union, which calculated that just to keep the rate of unionized workers stable, it “would require that the unions organize each year new members equal to 7.5 percent of their current membership.” That would require increasing the then-existing union organizing successes by six-fold, just to keep the unionization rate stable. The calculation would be a little different today, but the basic lesson remains: Americans say nice things about unions in surveys, but when it comes to organizing and supporting a union in their own workplace, most of them aren’t interested.
Before your writing can persuade those who actually read it of the merits of your thinking, the writing needs to persuade the reader–from the start–that it’s worth reading in the first place. I work as an editor, so I am predisposed to believe that editing matters. But Jan Feld, Corinna Lines, and Libby Ross provide some evidence on the point in “Writing matters” (Journal of Economic Behavior and Organization, January 2024, pp. 2378-397). Their study is of interest both for the methodology and findings, and also for how they define “good writing.” I’ll say a bit about both.
The study started with the authors contacting PhD students in economics and asking if they would send a paper, in exchange for free editing help. The authors got 30 papers. All 30 of the papers were then edited by professional editors, albeit professionals who didn’t usually work on academic writing. Thus, the 30 papers all had an original and an edited version. The editing took about six hours per paper.
The authors then recruited a set of 30 professors of economics and a different group of 18 writing/editing experts who worked in jobs like copywriter, technical writer, and communications manager. Each of these evaluators was sent a group of 10 papers. The economists evaluated the prospects for publication; the writing expert evaluated the quality of the writing. However, although the evaluator did not know it, they were each receiving a different mixture of original and already-edited articles.
In addition, both groups were asked to do a quick evaluation, spending less than 8 minutes per paper. This may seem harsh. But consider the situation of an academic who is evaluating a large batch of papers that might be included in a conference, or a journal editor looking at a large batch of papers and considering which ones to desk-reject and which ones to send to referees. Quick evaluations of papers are a reality of academic life. Again, before you can impress a reader with the details of thinking, you need to get over that hump of that first five-minute read.
Feld, Lines, and Ross can then compare the evaluations of the original and the edited papers. For the writing experts, the edited papers were scored 1.22 points higher on an 11-point scale for being “better written overall” (0.6 of a standard deviation). For the economics readers, “Economists judge the overall paper quality [of the edited papers] 0.20 SD better (0.4 points on the 11-point scale). They are also 8.4 percentage points more likely to accept edited papers for a conference, and are 4.1 percentage points more likely to believe that edited papers will get published in an economics journal that is classified as A* or A on the ABDC [Australian Business Deans Council] journal ranking.”
In short, six hours worth of work on the writing–done by an outside nontechnical editor who probably didn’t fully understand the technical economics–led to economist-readers believing that the paper was of higher quality. To put it another way, if you can’t or don’t do a good high-level edit of your own research papers, your chances of impressing readers is lower than it would otherwise be.
What is involved in taking a day to do a good high-level edit? In Appendix C, the authors spell out in a few pages what they asked the editors to do. In addition, the editors seemed to have used the software program StyleWriter to help the process along. I won’t reproduce the instructions to the outside editors in full, but here’s a short version:
Appendix C. Language Editing Guidelines for Experiment
An outline of the general approach the language editors will take
The goal is to edit the paper so that an expert who has 10 min to evaluate the paper will understand it more easily. We’ll focus on improving the title, abstract, and introduction using these guidelines. For the rest of the paper, we’ll focus on making the paper easier to skim read. …
Heavy edit of the title, abstract, and introduction only …
We’ll start by making sure the structure is clear.
The title should explain what the paper is about
We’ll make sure the title is clear. …
We’ll check the abstract is one paragraph that contains:
•the research question
•an explanation of how this question is answered
•the main findings. …
We’ll edit the introduction so that it has one paragraph for each of the following parts:
•the motivation for the research
•what the paper does (this paragraph often starts with “In this paper,…”)
•results
•the related literature (unless there is a separate literature section)
•contribution to the literature.
Avoid roadmap paragraphs
A good structure and informative section titles will do the trick in most cases …
Signposting for the reader
We’ll make sure the information flows well and is clear for the reader.
Remove roadmap phrases used to connect paragraphs
Make sure paragraphs are focused and only discuss one idea. For example, have separate paragraphs for describing the results and for discussing the related literature. …
The secret to a clear and readable style is in the first five or six words of every sentence. At the beginning of every sentence, locate the reader in familiar territory. The writing needs to have a clear flow of logic that is easy for the reader to follow — don’t frame information in a way that breaks the flow. …
Find the actor of the sentence and the actions they perform. If the actors are not the subjects and the actions are not verbs, we’ll revise so that they are.
Keep a short distance between nouns and their accompanying verb …
Use simple, familiar words
Use “use” instead of “utilize”. Use “people” instead of “individuals”.
Delete unnecessary words or clauses …
For example, in “We are the first to introduce a novel method”, there is no need to mention both “first” and “novel”.
In Section 2, we explain to the reader how our results are estimated.
Many introductory clauses that end with “that” can be deleted. Everything before the “that” should be deleted from a sentence.
It is usually the case that most good writers find that…
It should be noted that writing is an art and a science. …
Avoid abbreviations and acronyms
Use them only if they help the reader and choose the ones that sound good. OECD is fine, but use Facebook instead of FB. Write New Zealand, not NZ. …
We’ll remove any hedging statements that seem unnecessary. Writers don’t need to always say “all else equal”, “fairly”, “I would argue”. However, sometimes they need to qualify the statements to avoid people getting it wrong.
Avoid naked this or that in the beginning of the sentence
We’ll add more information where needed, such as “this regression shows” instead of “this shows”. …
We’ll add information to section titles and keep them short and concise. For example, “The credit market in New Zealand” is better than “Background”. …
Use self-explanatory titles of tables and figures
“The effect of peer gender on educational outcomes” is better than “Main Results”.
Most of these changes are already covered in detail in the above sections.
•Fix problems with long sentences (with StyleWriter)
•Fix problems with passive voice (with StyleWriter)
•Fix problems with nominalizations (with StyleWriter)
•Untangle noun strings
•Delete unnecessary words and clauses
•Use more personal pronouns (like “we” and “our”) where possible
A lot of this advice, along with the comments about appropriate fonts, spacing, and formatting, may seem obvious. So why not take a few hours of extra time to do it?
The 2022, total US exports of goods and services were $3 trillion, while imports were $3.9 trillion. This overall trade provides benefits to the economy: sellers who have access to global markets (ask a farmer!), buyers who have greater access to products with the price/quality mix they prefer, and heightened competitive pressures on domestic producers to do better. However, the gains are not distributed evenly, and in some cases–say, workers in an industry facing especially strong competition from imports–people can experience outright losses.
How big are the overall gains? How should we think about the losses?
To summarize economic gains from trade, Hufbauer and Hogan first point to pre-2017 research on more than a dozen studies of international trade, which “calculated an average `dollar ratio’ of 0.24.” They write:
Simply put, the dollar ratio is the dollar increase in GDP divided by the dollar increase in two-way trade. In language familiar to economists, the dollar ratio is the elasticity of income (GDP) with respect to trade. Expressed another way, the calculation indicates that a 1 percent increase in trade yields a 0.24 percent increase in GDP—i.e., a $1 billion increase in two-way trade increases GDP by $240 million.
They redo and update these calculations based on a newer wave of studies and suggest an updated “dollar ratio” of 0.30–that is, a $1 billion increase in two-way trade increases GDP by $300 million.
What about those who experience losses from trade? Hufbauer and Hogan offer essentially two responses. One response is to put the number of those who suffer from trade in the context of job churn in the larger US economy. After all, there are also workers who lose jobs because their employer is losing market share to domestic competitors, perhaps because of a shift in consumer preferences, or shifts in the skill level of workers that US employers are looking for, or because of below-average management. Other US workers may lose jobs because of automation and new technology. Many more workers switch jobs, looking for higher pay or better opportunities, especially if they sense that their current employer may be in troubles. The second response is to argue that all US workers who are laid off from a job deserve government support through unemployment insurance and other programs, while they move to a new job. They write:
Compared with our estimate of US workers who lost or changed jobs because of increased imports (242,000 annually between 2019 and 2022), roughly 50 million American workers change their jobs each year. A small fraction of these workers is “displaced,” meaning laid off, including a smaller fraction displaced by imports. Displacement reduces earnings over the long term. Throughout the vast US economy over the past two decades, annual displacement ranged from under 1 percent to over 3.5 percent of the labor force (1 million to 5 million workers). All displaced workers, including those displaced by trade, deserve better public safety nets.
Of course, if you find these estimates of the gains from trade to be implausible, and instead you would prefer to see a substantial decline in international trade–take heart! This is your time! US trade in goods and services as a share of GDP have been declining since the Trump administration. However, international flows of data, information, and foreign investment have continued to rise. The forms of globalization have shifted, but the technologies driving greater global connectedness continue to develop.
