Global labour markets of PhDs

In early January 2019, economists and social scientists will jostle at the annual American Economic and Social Science Association meetings in Atlanta as academic departments globally will try to recruit freshly minted doctoral students for tenure track and tenured positions as well. They will however have new competition from big data firms like Facebook, Amazon and Uber who are also going to be there to recruit PhDs. And the tension that academia may lose well-trained young minds from their university garages is already palpable.

In doing so, big data may not just be the new oil but also the new robber barons, stealing PhDs from academia in the global labour market. Or so it seems is the feeling of several North American university academicians in recent conversations of this author with them. On econ-twitter, an economist colleague tweeted recently that “PhDs who end up working outside the academy are not failures, enough already with professional elitism”, but such a view seems a minority. The rise of technology firms, generating click-stream data in platforms by the minute has created new markets and ways of acquiring information, where digital economists now are increasingly playing an important role in design, strategy, pricing, and dealing with regulation.

In a 2018 paper, Stanford and Harvard economists Susan Athey and Michael Luca point exactly to this. As they document, North American job postings from technology firms for economists with a PhD increased to 21 in the 2017-2018 cycle from 15 just four years earlier, while economics departments saw during the same period, a fall in postings from 232 to 194. While the numbers are still not threatening, the skills that economists bring to technology firms (including assessing causal empirical relationships besides analysing equilibrium market structure) offer value that are difficult to ignore. To add to it, digital economists in their daily jobs can also offer econometric advice and troubleshoot on design of online advertising, provide statistical intuition behind ranking, reviews and returns from advertising apart from helping out policy teams of firms in understanding market structure, thereby anticipating and dealing with regulatory effects on firm competitive advantage. As a senior academic quipped, universities train PhDs in economics and social sciences for 5 to 6 years spending millions of dollars (factoring in tuition waiver, health insurance, a stipend in graduate school along with faculty time and salaries) and then these new ‘robber barons are stealing them away with 6 figure salaries’.

To be sure, such a quibble may be incomplete. If positions in economics are declining and faculty getting tenure upstream is also on the downtrend (besides witnessing the rise of new positions in North American universities of associate professors without tenure), one can only expect rational agents (in these cases, PhD students) planning their way out apriori from the academy into more lucrative industry options. The job can be fun too, some firms allowing data scientists to write papers building on their daily troubleshooting work. 

All that said, one can expect some course corrections too going forward in the formal PhD labour market. First, universities can be expected to raise their graduate student stipends and fellowships to attract more students in the doctoral life, especially if the effect of the robber barons is going to spill over downstream with lesser number of students entering the field per se. Second, universities can also be expected to enforce contractual obligations in the doctoral programmes ensuring that graduate students after finishing their PhD need necessarily continue in academia for couple more years, to make sure that some of the investments on graduate doctoral training indeed witness private as well as social returns. In addition, technology firms in responding competitively, may end up creating their own internal universities where bachelors and masters students can be offered training on the job in the same skills that a university trained PhD economist may come with, this in preparation for a supply gap that may emerge. Also, one can also expect a more concerted shift in the academic curriculum of what traditional economics and public policy PhD programmes offer in universities globally. Finally, one can also expect welfare losses from PhDs in other disciplines (like say in mechanical or civil engineering or mathematics and physics) turning data scientists for the new age technology firms. Many of these are already happening but equilibrium outcomes for each would be interesting to keep an eye out for in the future. 

This apart, one can also expect international ramifications with doctoral students from emerging economies providing an alternate labour pool for technology firms for the requisite skills, though, if that happens, what will be the implication of that for basic research, postdoctoral work and doctoral studies in emerging economy universities remains to be seen. In addition, there could also be creation of new immigration visa categories by countries not just in North America but also in other central locations of big-data work (like China and India) to accommodate these digital social scientists. Finally, the geographic location of such work could change if digital social scientists are cheaper to purchase in international locations like in Pune or Bangalore, India or Shenzhen, China.

Overall, these are interesting times coming from unintended consequences of the maturing of big-data firms in the global labour markets of PhDs. Early economists who toggled their lives between universities and technology firms (like Steve Tadelis of EBay and Amazon or Google’s Hal Varian – both noted UC Berkeley professors, or Susan Athey of Stanford University who has worked with Microsoft), may have shown the way, but the digital industrial wave in the market for PhDs in economics, social and physical sciences both in the global north and south is far from over.

Author is with IIM, Ahmedabad