Every Monday, the National Bureau of Economic Research, a nonprofit organization made up of some of North America’s most respected economists, releases its latest batch of working papers. The papers aren’t peer-reviewed, so their conclusions are preliminary (and occasionally flat-out wrong). But they offer an early peek into some of the research that will shape economic thinking in the years ahead. Here are a few of this week’s most interesting papers.
Author: Peter Cappelli
What he found: Many trade groups in the U.S. claim that people increasingly lack the necessary skills to warrant hiring, citing a systemic lack of high-quality workers (a “skills gap”) or few workers in sought after occupations (a “skills shortage”). But there is little evidence supporting these claims; it’s more likely a problem of “skills mismatch,” with many workers “overskilled” for their jobs.
Why it matters: In recent years, trade groups have published a lot of reports with recommendations for addressing these supposed problems, such as increasing quotas for skilled immigrants or encouraging students to enter STEM fields, such as computer science and engineering. But the author of this paper criticizes these claims, saying they lack merit, employ dubious methodology or promote arguments not backed up by the data. The real problem seems to be a skills mismatch. Thus the opposite problem, being overskilled, seems to be evident: individuals with more education than required for the job they do. The real aim, then, of these employer-biased recommendations, the author claims, is to shift the responsibility onto the government and workers themselves to acquire the necessary skills, rather than encourage apprenticeship and on-the-job training.
Key quote: “Overall, the available evidence does not support the idea that there are serious skill gaps or skill shortages in the US labor force. The prevailing situation in the US labor market, as in most developed economies, continues to be skill mismatches where the average worker and job candidate has more education than their current job requires.”
Data he used: Various data sources covering labor markets and education, but notably the Organization of Economic Cooperation and Development’s Program for International Assessment of Adult Competencies
Authors: Judd B. Kessler, Alvin E. Roth
What they found: A yes/no question for organ donation registration (as opposed to the traditional opt-in question) does not increase, and may decrease, the number of donors.
Why it matters: In the U.S., more than 10,000 people die every year waiting for an organ donation, yet only 48 percent of adults are registered as donors. Popularized in the book “Nudge,” organ donation registration using the “active-choice” frame — when the question is posed explicitly as having a “yes” or “no” answer — has gained adoption by 41 of the 50 states. The motivation behind this idea, rooted in behavioral economics, is that a policy of simply changing the question on a form can increase the number of donors (relative to the traditional “opt-in” frame, where explicit agreement to become an organ donor is required).
However, this new research casts doubt on such hopes. Using donor registration data from California, which covers a period when the phrasing was changed in 2011, the authors find the switch from an opt-in to active-choice phrasing lowered the number of organ donors anywhere from 2.2 to 2.9 percent from what it would have been otherwise. Furthermore, the active-choice frame may lower organ donation even among those not registered as donors. This is because an unregistered deceased person’s next-of-kin may be less willing to donate his organs if she believes the deceased “had actively chosen not to donate” rather than just not opting-in. In a separate experiment, the researchers found that by re-asking some respondents who had previously declined, they got a significant number of new registrants; and so the authors recommend an opt-in frame that does not have an explicit “no” option — hence, “don’t take ‘no’ for an answer.”
Key quote: “Our results suggest that not only does active choice not increase the registration rate, it may decrease the transplantation rate through suggesting to next of kin that unregistered donors had actively chosen not to donate.”
Data they used: Organ donor registrations following a 2011 California policy change; and experimental data from individuals asked again about their registration as organ donors with the Massachusetts Department of Transportation.
Authors: Byungkyu Lee, Dalton Conley
What they found: The sex of the first child has no effect on a parents’ political affiliation. This null finding is in contrast with two recent pieces of research that found having a girl instigated a change of political affiliation — to hold more liberal views in the U.K. and more conservative ones in the U.S.
Why it matters: Null findings — where research demonstrates no such relationship — often get the short end of the stick in academic research. But this paper comes on the heels of two other conflicting studies claiming the gender of parents’ first-born affects their political affiliation. Instead, this paper found no such effect. Using two large, multiyear surveys covering 36 countries, the authors find that previous studies’ conclusions vanish when controlling for the heterogeneity of the countries covered. Furthermore, the data is noisy, and the results reported elsewhere were probably due to chance. The authors of this paper do not claim that the previous findings should be ignored; they leave open the possibility that such gender-specific effects could exist and warrant further study. But the paper is a reminder of the dangerous likelihood a statistical finding is spurious.
Key quote: “We re-examined these results using the GSS and ESS and found no statistical discernible effect of the first child’s sex on party identification in either the U.S. or the U.K. — where significant effects on party identification were reported — nor on parental political ideology in 36 countries.”
Data they used: General Social Survey (GSS) data from 1972 to 2012 and European Social Survey (ESS) from 2002 to 2012.