27/12/2014: Are Graduate Students Rational?

A fascinating study into expectations formation mechanism for career outlooks by entering doctoral students in the US. Authored by Blume-Kohout, Margaret and Clack, John, titled "Are Graduate Students Rational? Evidence from the Market for Biomedical Scientists" (PLoS ONE 8(12): e82759, December 2013, http://ssrn.com/abstract=2506810) the paper looks into whether entering graduate students make rational choices in selecting specific fields of study, given information available in the jobs markets.

The authors use increases in the U.S. National Institutes of Health (NIH) budget from 1998 through 2003 that in turn also "increased demand for biomedical research, raising relative wages and total employment in the market for biomedical scientists." This should send a signal to incoming students that biomedical research careers prospects have been expanding. This signal is not the same as a signal of what one can expect of the biomedical careers prospects in the future for a number of reasons. Crucially, if at early stages of funding expansion wages and career promotions for those already in biomedical profession were rising fast, any future increase in supply of biomedical professions will reduce earnings and career prospects for future entrants into profession. In other words, current conditions are not identical to future conditions.

This is especially salient in professional fields where studying and research required for qualifying into profession requires a long period of time, such as biomedical field. where "research doctorates in biomedical sciences can often take six years or more to complete."

Hence, in biomedical field, "the full labor supply response to such changes in market conditions is not immediate, but rather is observed over a period of several years."

If prospective students considering entering doctoral studies were rational (in economic sense), they should not tai current conditions in the filed for granted and should instead "anticipate these future changes, and also that students take into account the opportunity costs of their pursuing graduate training."

As authors note, "prior empirical research on student enrollment and degree completions in science and engineering (S&E) fields indicates that “cobweb” expectations prevail: that is, at least in theory, prospective graduate students respond to contemporaneous changes in market wages and employment, but do not forecast further changes that will arise by the time they complete their degrees and enter the labor market."

The Blume-Kohout and Clak analysed "time-series data on wages and employment of biomedical scientists versus alternative careers, on completions of S&E bachelor's degrees and biomedical sciences PhDs, and on research expenditures funded both by NIH and by biopharmaceutical firms, to examine the responsiveness of the biomedical sciences labor supply to changes in market conditions."

They find evidence rejecting rational expectations model of students' decision making: "Consistent with previous studies, we find that enrollments and completions in biomedical sciences PhD programs are responsive to market conditions at the time of students' enrollment. More striking, however, is the close correspondence between graduate student enrollments and completions, and changes in availability of NIH-funded traineeships, fellowships, and research assistantships."

In other words, state-funded research can contribute to over-production of future doctoral graduates later in the period of increased funding, exacerbating future wages downturns for later stage doctoral graduates, and at the same time fuel increased inflows of new entrants into profession.

29/6/2013: Research Funding Does Not Seem to Match Research Performance

Very interesting:

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0065263

From the abstract: "Agencies that fund scientific research must choose: is it more effective to give large grants to a few elite researchers, or small grants to many researchers? Large grants would be more effective only if scientific impact increases as an accelerating function of grant size."

So, the paper examines "the scientific impact of individual university-based researchers in three disciplines funded by the Natural Sciences and Engineering Research Council of Canada (NSERC)", based on "four indices of scientific impact:

• numbers of articles published, 
• numbers of citations to those articles, 
• the most cited article, and 
• the number of highly cited articles."

All of the above parameters are "measured over a four-year period" and referenced against "the amount of NSERC funding received".

Core findings:

• "Impact is positively, but only weakly, related to funding" - which is disturbing, as it suggests that funding allocation is not academically efficient. 
• "Researchers who received additional funds from a second federal granting council, the Canadian Institutes for Health Research, were not more productive than those who received only NSERC funding." Which suggests that the most important agency has trouble identifying and signalling by its grants allocation the academic 'winners'.
• "Impact was generally a decelerating function of funding." Which is really bad. In basic terms, the more funds were pumped the less was the positive marginal impact. So that "impact per dollar was therefore lower for large grant-holders". Which "is inconsistent with the hypothesis that larger grants lead to larger discoveries". 
• "Further, the impact of researchers who received increases in funding did not predictably increase." So obtaining a larger grant did not lead to subsequent improvement of the researcher output! 
• "We conclude that scientific impact (as reflected by publications) is only weakly limited by funding."
• And a big Boom! "We suggest that funding strategies that target diversity, rather than “excellence”, are likely to prove to be more productive." 

Now, the last point is what all funding agencies around the world are trying to avoid. Everywhere, the policy in funding research is on more concentration and on more winner-picking. Not on funding broader research and research groups.

Incidentally, if you are interested in this topic - what and how should be funded and prioritised in research and education - read my Sunday Times article tomorrow.