Category: Data Source

Does alleviating poverty increase cognitive performance?

by  Tyler Cowen in

From an RCT by Barnabas Szaszi et.al, due to travel I have not yet had the chance to look at this one:

In this Registered Report, we investigated the impact of a cash transfer based poverty alleviation program on cognitive performance. We analyzed data from a randomized controlled trial conducted on low-income, high-risk individuals in Liberia where a random half of the participants (n=251) received a $200 lump-sum unconditional cash transfer – equivalent approximately to 300% of their monthly income – while the other half (n= 222) did not. We tested both the short-term (2-5 weeks) and the long-term (12-13 months) impact of the treatment via several executive function measures. The observed effect sizes of cash transfers on cognitive performance (b = 0.13 for the short- and b = 0.08 for the long-term) were roughly three and four times smaller than suggested by prior non-randomized research. Bayesian analyses revealed that the overall evidence supporting the existence of these effects is inconclusive. A multiverse analysis showed that neither alternative analytical specifications nor alternative processing of the dataset changed the results consistently. However cognitive performance varied between the executive function measures, suggesting that cash transfers may affect the subcomponents of executive function differently.

Via Michelle Dawson.

Further positive results on school choice

by  Tyler Cowen in

Does a school district that expands school choice provide better outcomes for students than a neighborhood-based assignment system? This paper studies the Zones of Choice (ZOC) program, a school choice initiative of the Los Angeles Unified School District (LAUSD) that created small high school markets in some neighborhoods but left attendance zone boundaries in place throughout the rest of the district. We study market-level impacts of choice on student achievement and college enrollment using a differences-in-differences design. Student outcomes in ZOC markets increased markedly, narrowing achievement and college enrollment gaps between ZOC neighborhoods and the rest of the district. The effects of ZOC are larger for schools exposed to more competition, supporting the notion that competition is a key channel. Demand estimates suggest families place substantial weight on schools’ academic quality, providing schools with competition-induced incentives to improve their effectiveness. The evidence demonstrates that public school choice programs have the potential to improve school quality and reduce neighborhood-based disparities in educational opportunity.

That is from a new paper by Christopher and Caitlin Kearns in the QJE.

Building networks: Investigating the quid pro quo between local politicians & developers

by  Alex Tabarrok in

My latest paper, Building Networks, is with Vaidehi Tandel and Sahil Gandhi in the Journal of Development Economics. We look at the connection between politics and land markets in Mumbai, India. The subject is inherently difficult to analyze because most such connections are illicit and under the table. What we find, however, is that when the local politician loses power development projects slow down which suggests that it takes time to build the connections that are necessary to speed approval through the bureaucratic process. Much more in the paper.

Abstract: Mutually beneficial arrangements between politicians and real estate developers are common in many developing countries. We document what happens when the politician-developer nexus is disrupted by an election. We construct a novel dataset of real estate projects and electoral constituencies in Mumbai’s municipal government. We find that an incumbent party losing the election increases real estate project completion times by 5%. We find no effect of quasi-random redistricting or changes in voter preferences on project delays. We investigate two mechanisms for the slowdown associated with party turnover — delays in construction approvals around the time of the election and increase in litigation against projects after the election. While we see no rise in litigation, we find that delayed approvals near an election explain 23% of the increased total delays due to party change.

The “Deaths of Despair” narrative is somewhat wrong

by  Alex Tabarrok in

Matt Yglesias does an excellent job laying out the case against the “deaths of despair” narrative and putting it bluntly. 

Over the past few years, Anne Case and Angus Deaton have unleashed upon the world a powerful meme that seems to link together America’s troublingly bad life expectancy outcomes with a number of salient social and political trends like the unexpected rise of Donald Trump.

Their “deaths of despair” narrative linking declining life expectancy to populist-right politics and to profound social and economic decay has proven to be extremely powerful. But their analysis suffers from fundamental statistical flaws that critics have been pointing out for years and that Case and Deaton just keep blustering through as if the objections don’t matter. Beyond that, they are operating within the confines of a construct — “despair” — that has little evidentiary basis.

…Novosad, Rafkin, and Asher have provided a compelling analysis of a very concentrated problem of worsening health outcomes for the worst-off Americans. Case and Deaton, by contrast, have delivered a very misleading portrait of worsening health outcomes for the majority of Americans that (because they mistakenly think it’s a majority) they attribute to broad economic forces that exist internationally but which for some reason only cause “despair” in the United States.

…The point is that we face a set of discrete public health challenges that we need to think about both as policy matters and in terms of politics and public opinion. But there is no “despair” construct driving any of this, and the linkage to big picture political trends is simply that Republicans are more hostile to regulation. Case and Deaton, meanwhile, have sent us on the equivalent of a years-long wild goose chase away from well-known ideas like “smoking is unhealthy” or “it would be good to find a way to get fewer people to use heroin.”

I tend to agree with Matt but I would offer a few cautions. Case and Deaton have been too broad in identifying the at-risk population. Identifying more carefully the at-risk group(s) is important so that we can target different problems with different solutions. Indeed, part of what makes the very important opioid crisis so bedeviling is precisely that it is not limited to “despairing” populations but cuts across many groups.

I wouldn’t, however, throw out despair as an organizing principle. The evidence on “despair” goes beyond death to include a host of co-morbidities such as mental stress, marriage rates, labor force participation rates and other measures of well being. Regardless of the precise population to which these problems attach they are co-morbidities and I suspect not by accident. Education is a proxy for the underlying problem but likely not causal. Matt’s cheeky suggestion to promote ideas like “smoking is unhealthy” illustrates part of the issue. Education and information will not solve that problem. Smokers know that smoking is unhealthy but they do it anyway–perhaps because it’s one of the few easily available pleasures if you are unmarried, out of work and stressed.

Nevertheless, do read the whole thing. 

Campus censorship and Israel-Palestine

by  Tyler Cowen in

Do looks matter for an academic career in economics?

by  Tyler Cowen in

We show that physical appearance plays a role in the success of economics PhD graduates and investigate the underlying mechanisms driving this relationship. Leveraging a unique dataset of career and research productivity trajectories of PhD graduates from leading economics departments in the United States, we provide robust evidence that appearance is a predictive factor for both research productivity and job placement. Our analysis goes beyond establishing the association between attractiveness and success within the profession. By jointly examining appearance, job outcome, and research productivity, as well as the longitudinal development of the latter two over time, we show that the effect of appearance can be partially, but not fully, attributed to its role as a predictor of research productivity, with the remainder of the effect reflecting an intrinsic demand for attractiveness.

That is forthcoming in the Journal of Economic Behavior and Organization, by Galina Hale, Tali Regev, and Yona Rubinstein.  Via Rolf Degen.  And here is a recent piece on lookism and fund managers.

On U.S. life expectancy disparities

by  Tyler Cowen in

From the excellent Dylan Matthews:

Case and Deaton are highlighting a real problem, confirmed by other researchers: Americans with different levels of education die at different rates, and the least-educated Americans have seen their death rates surge in a way that more-educated Americans have not.

But the relevant divide does not seem to be between people who earned a bachelor’s degree — who remain a minority among American adults — and people who didn’t. Other research suggests that the problem is concentrated in specific areas of the US, and between the very least-educated Americans (particularly high school dropouts) and the rest of the country, rather than between college grads and non-grads.

Moreover, the cause of the divergence between high school dropouts and the rest of the country does not seem to be caused by “deaths of despair.” There is no doubt that the opioid epidemic in particular has wrought spectacular damage in the US. But some researchers are finding that stagnating progress against cardiovascular disease is an even bigger contributor to US life expectancy stalling out, and to mortality divides between the most- and least-educated Americans.

A lot of what you read about “deaths of despair” is in fact wrong or misguided.

The Labor Market Returns of Being An Artist

by  Tyler Cowen in

The labor market penalty to choosing the arts seems to be rising:

Using individual-level data from the Census Bureau’s American Community Survey (ACS) between 2006 and 2021, I study the labor market experiences of artists. First, I find a decline in the relative earnings of artists to non-artists from zero to a 15% disadvantage. After controlling for demographic differences, the decline is sharper, declining from a 15% earnings disadvantage to 30%. That the inclusion of demographic controls raises the earnings gap suggests there is positive selection into the arts. Second, these differences decline in magnitude to 4.4%, but remain statistically significant, after exploiting variation among artists and non-artists in the same industry-year and major occupation. Third, when restricting the set of individuals to those with at least a college degree, those with a fine arts degree also incur an earnings and employment penalty even if they work in the arts. These results highlight the increasing financial precariousness of artists over the past decade.

That is a new paper by Christos Makridis, via the excellent Kevin Lewis.  Overall this result makes sense to me.  Success in the arts requires extreme talent of some kind in most (not all) cases.  Those individuals can earn increasingly more in other endeavors.  But if the arts are trapped in a “Malthusian equilibrium,” with intense entry competing down returns because it is fun, artistic earnings may not keep pace.

Work From Home Works

by  Alex Tabarrok in

It took firms decades to adjust to electricity by redesigning factories, products, and workflows to take full advantage of the new possibilities. Similarly, the benefits of work from home start to come most profoundly when expensive offices can be shrunk, employers can draw from a much larger pool of workers and workers can adjust when and where they work, including the location of their homes. It’s not surprising, therefore, that with little time for either the workers or the firms to adjust and with few options to choose how much to work from home, productivity fell when COVID sent workers home.  But, with more time to plan and more options for hybrid but extensive work from home (e.g. work from home Mondays and Fridays), work from home has large benefits. We are also seeing management redesign to take advantage of work from home in the same way we saw factory redesign to take advantage of electricity. Management, for example, is shifting from input metrics–do you show up?–to output metrics–did the work get done? Designing and validating new metrics takes time, but these changes are helping to increase the benefits of work from home.

Nick Bloom reviews the evidence (slides here).

Looking at micro economic studies on working from home productivity, the classic is “The Stanford Study” I helped oversee in 2010-2012. We randomized 250 employees in a large multinational firm into those who would work from home and those who would report to the office. The expectation, of course, was that home-based employees would goof-off, sleeping or watching TV rather than working.

So, we were shocked to find a massive 13 percent increase in productivity.

The productivity boost came from two sources. First, remote employees worked 9 percent more in minutes per day. They were rarely late to work, spent less time gossiping and chatting with colleagues, and took shorter lunch breaks and fewer sick days. Remote employees also had 4 percent more output per minute. They told us it’s quieter at home. The office was so noisy many of them struggled to concentrate.

The macro evidence also suggests or at least is consistent with work from home working:

In the five years before the pandemic, U.S. labor productivity growth was 1.2 percent; since 2020, this picked up to 1.5 percent. Given the state of the world, that acceleration was miraculous.

What could have caused this? Perhaps rising government expenditure and easy monetary policy? Possibly, but greater government activity traditionally is associated with lower, not higher, productivity growth. Perhaps an acceleration in technology and computerization? Possibly, but the pandemic did not witness any pickup in technological progress. Perhaps the five-fold surge in working from home post-pandemic. Maybe cutting billions of commuting hours, replacing millions of business trips with Zoom meetings, increasing the labor supply of Americans with disabilities or child-care commitments, and saving millions of square feet of office space increased productivity? It is honestly hard to say.

Revealed preference is the most telling piece of evidence. Workers value the option to work from home and many firms now advertise the options for hybrid work as a benefit:

…millions of firms around the world are adopting hybrid and remote work, there has to be something there. I have spoken to many hundreds of managers and firms over the last three years and I repeatedly hear they use home working as a key part of their recruitment and retention strategy. Indeed, another recent experiment on 1600 employees found hybrid reduced employee quit rates by 35 percent. \

Despite a rocky start, work from home appears to have stabilized at around 25% of work days overall and stunningly, nearly 40% of work days for college educated workers! Work from home thus appears to be a permanent and beneficial change in how work is structured.

It is the poor who are lonely (on average)

by  Tyler Cowen in

Lower-income people are more lonely

Jiska Cohen-Mansfield did a literature review with Haim Hazan, Yaffa Lerman, and Vera Shalom of the statistical correlates of loneliness in older adults and found that being low-income is a strong correlate of loneliness. You see the same thing in surveys of middle-aged and elderly Portuguese people, in the Nova Scotia Quality of Life Survey, and in Eastern Europe.

Michelle Lim, Robert Eres, Shradha Vasan have the interesting finding that low income predicts loneliness not only on the individual level but also that “living in poorer neighborhoods” is associated with loneliness.

Sometimes scholarly literatures feature big disputes, or at least nuanced disputes, but in this case there seems to be no dispute at all: loneliness is associated with lower income and thus probably not caused by big houses or lack of huts. I also think it’s notable that at least among rich countries, loneliness seems higher in the poorer (or perhaps “less rich”) ones like Greece and Italy than in the United States and Switzerland.

The low rates of loneliness in egalitarian Sweden and Denmark, in particular, suggest that having more money pretty literally leads to less loneliness. Note as well that while the United States has a somewhat threadbare welfare state, this is data for senior citizens who do enjoy universal health care in the United States and a basic income via Social Security.

It may be, in other words, that being able to afford to do more leisure activities is a significant protector against loneliness. You go do more stuff and you make more friends. Or you have more opportunity to maintain your relationship with friends because you can afford to hang out and do stuff. I don’t think the exact nature of the causal relationship is clear from the studies that I’ve seen, but it bears more examination, especially because a lot of people seem to intuitively spin out to “paradoxical” accounts of loneliness that don’t seem well-supported.

That is from Matt Yglesias ($).

Does GPT represent a triumph of WEIRD values?

by  Tyler Cowen in

Yes, in short: