Are the old folks holding us back?

Dec 27 2010 Published by under scholarly communication, STS

ResearchBlogging.orgWe've been hearing a lot about how hard it is to get a tenure track job - arguably harder even than it was during other economic recessions. We've also been hearing about how the age of NIH PIs is going up. I guess the age at first award is going up as well as the average. At the same time, there are a lot of stories about how the most disruptive ideas have come from young scientists. Also that young scientists are more productive. So if academia is aging and older scientists are not leaving, does this mean there will be an accompanying loss in productivity? That’s what this article looked into. It’s a review of the research on many different aspects of this question: “whether there is empirical support for the belief that age is negatively related to scientific achievement”

I never realized that there had been mandatory retirement at universities in the US. This of course is discrimination (violated the amended Age Discrimination in Employment Act), so was abolished, but only in 1994. What with baby boomers and all not retiring, you can see how the average age would be going up and how there would be less room to hire young faculty. Many colleges apparently try (tried?) to incentivize early retirement but maybe only with limited success.

The first thing the author looked at is if there is an association between age and scientific achievement. He proposes four factors:

  • changes in the cognitive abilities due to age
  • changes in motivation
  • availability of resources
  • legal curtailing (compulsory retirement where and when exists/existed)

As for cognitive ability, it’s interesting the author goes back to research published in the 1950s which seems to set the stage. There were cohort issues and a lot depends on the way these abilities are tested. Large longitudinal studies seem to find that abilities do decline but fairly slowly until about age 80. There are also suggestions that, after being in a specific line of work for a while, you get in a rut and you don’t think of new or creative ways to solve problems. But this is something you could fix by changing approaches or adopting a new research area.

As for motivation, there are some economic models based on anticipated life income that would indicate a drop in motivation. On the other hand, if scientists are more motivated by prestige and the rewards and recognition that come from scientific achievement, they should be more motivated to continue publishing as they go along. Likewise, if they have intrinsic motivation and enjoy their work, there’s no reason that should drop off.

As for availability of resources, the work of the Coles and of Price found that the top researchers account for the majority of the publications – it’s a standard long tailed distribution. Price’s law is: if there are k researchers in a field, \( sqrt{k}\) will be responsible for 50% of the contributions. Merton’s Matthew Effect is that the rich get richer. So if you’re successful, then you’ll get more rewards and resources which facilitates more and better science. Success breeds success, but failure leads to more failure. Aging scientists who have not been successful will most likely drop in productivity.

As for compulsory retirement – most of these studies date back to when that was the case. Also, in some countries, older scientists can’t start big new projects a few years before retirement so they actually start winding down early.

The author then goes on to review the evidence. Indicators of achievement include prizes and awards, citation count (no mention of the h-index, but that would make sense), number of publications, number of grants, etc. There are always problems with the research designs in these studies. They don’t account for the age distribution of the population of scientists when looking at the age of award winners. There are cohort effects – different groups of scientists use different methods and have different publication patterns. There are also period effects – there might be some historical situation that increased or decreased publication during a period in the author’s career.

A bunch of the older studies find a sort of downward curve effect. The awards, publications, citations all seem to peak around the 40s and then drop off.   These studies do show a decline in productivity with age, but age seems to account for only a small part of the variation. Past performance is the best predictor and the quality of the work doesn’t seem to decrease. For math, though, the story is quite different – it’s a flat line for productivity and age.

In newer studies (there are only four here), the line is basically flat with a slight rise at 11-15 years into the career and another at 26-30 years into the career.

The author didn’t find evidence to support the idea that the graying of academia will lead to a loss in productivity. Additional longitudinal studies are needed, but they will be difficult to perform for the reasons already listed as well as the changes in publication norms. Most universities are pushing very hard for their faculty to publish so the number of publications will continue to rise.

So it seems crazy to force productive scientists to retire, but it’s complicated to get unproductive but highly paid scientists to retire while keeping the productive scientists. If no one retires, the older scientists have pretty high pay and there will be little room for the pipeline.

It’s an interesting article and very readable. Recommended.

Reference: Stroebe, W. (2010). The graying of academia: Will it reduce scientific productivity? American Psychologist, 65 (7), 660-673 DOI: 10.1037/a0021086

3 responses so far

  • Mark says:

    Is it not also true that older PI's are less likely to adopt new technologies that could make science more efficient? Particularly 'online science'.

    • Christina Pikas says:

      The article discussed creativity and approaches to problems, but not new technologies specifically. I'm guessing that mid career scientists are most likely to experiment with new technologies - they are secure and can do so more than younger scientists in conservative fields.

  • Eva G. says:

    Excellent post! I definitely want to check out this article. I've heard a version of this discussion in all facets of library land as well. It's the old "Old librarians have to retire to make room for the new librarians" argument. Often, new librarians point to their productivity being higher - that's a blanket statement that I don't think is fair and it sounds like this study also counters that sentiment.

    The recession has brought out some ugly ageism. I wish younger people would try to imagine approaching retirement and realizing that they couldn't afford to. New [faculty, librarians, whatever] really need to stop thinking that "older" people aren't retiring in spite of them. Times are tough, people! People can't retire because they can't afford to, not because they're trying to stifle the newbies.