How to encourage social science academics to work on Australian policy?

In recent years, there have been many reforms to the incentive system that social science academics (those in the fields of economics, finance, psychology, management, health, marketing, etc.) live under in Australia. There was the Research Quality Framework, then the ERA, and now something based on expert panels that is perhaps closer to the ERA in Britain. There have also been fluctuations in the amount of money pumped out to academics via the ARC, the NHMRC, local state funding, and many direct research outfits (like CSIRO). I first want to make some observations on the changed incentives for academics, before going into actual recommendation as to how we could get our top academics to do more work on Australian policy matters. In short, if you want more of the top academics to write on Australian policy issues, I would recommend setting up specialised PhD institutes like the European Institutes, and I would recommend a more relaxed attitude to privacy when it comes to the use of Australian datasets by academics.

Observations:

1.       At the top end, it has been publish or perish long before the government got into the game of assessment exercises. Good young budding academics knew that in order to be seen as a top academic they had to do well in the international publication game. Like it or loathe it, that game is dominated by Europeans and Americans. This doesn’t mean that it is impossible to do work on Australian issues and get into the top journals, but it does mean that one ultimately has to work on the things that interest European and American editors and referees. Hence a cute Australian experiment with a policy that might also be introduced elsewhere (say, the baby bonus or HECS) has a shot. An analysis of the labour market dynamics in Australia has no chance. As a result, many of the top academics have told their best students in the last 30 years or so not to work with Australian data and to preferably go and study somewhere else. It is a form of cultural cringe in that one has an insufficient degree of pride in ones’ own country, but many felt this was the honest advice they owed their students. Australia is not alone in this. We interviewed job market candidates from Britain this year whose supervisor had told all his students to work with American data in order to have more chance of top publications.

2.       The research assessment exercises have made the reality above visible to everyone. In particular, it has made it visible who has not played the international game at all and who has. Even though it was only in place shortly, it has lead to major changes in the power structure inside academic schools, and now that that genie is out of the bottle it wont go back in. Those who publish a lot in leading journals were promoted much faster than they would have been otherwise, whilst those who didn’t do any visible research had to become administrators if they still wanted to have a chance of being called professor. This was both good and bad. It was good for the highly talented who felt frustrated that there was no clear way in which they could outperform others, whilst it also, at a stroke, stigmatised individuals who made different, but still worthy, investments, such as in teaching quality or local research groups oriented towards Australian policy. At a stroke they were visibly designated as second rate, whereas before they could at least with some probability get away with saying they were world-class (which, in a local sense, they were). The recent demise of the ERA changes nothing about this reality, so don’t expect any tweaked version of the ERA to make much difference. The real losers from the recent change in the ERA are those who only look good on an A/A* publication schedule, but whose work is not cited, who have few grants, who teach small classes, who do little service, and who basically play no other game than the international publication game. And they do not really lose any of their reputation, but their desirability for institutions looking to do well in ERA rankings is slightly reduced, which might cost them some of their loadings.

3.       Even though teaching is the core business of the tertiary sector, the assessment exercises have contributed to the gradual marginalisation of teaching that has happened the last 20 years. Teaching in academia is now considered a second-rate activity, one that is actively avoided by the up-and-coming. In fact, the reduction in the status  of teaching in academia has gone so far that teachers are now at the very bottom of the university hierarchy, behind administrators who can boss them around and pester them with forms; behind junior researchers who know nothing about real life but who play the journal game better; and even behind contract workers. It is no wonder that the academic teaching profession is not attracting good youngsters and has to rely heavily on migrants who will by design know less about Australian policy issues and will have less interest in it.

4.       The advent of the internet has increased the possibilities of academics to move into the market for Australian economic issues. Blogs, tweets, homepages, and virtual centers have reduced the costs of speaking to the general public and to the policy makers. This has made the returns to Australian policy work more direct and higher, though paradoxically it tends to reward fairly shallow commentary rather than real understanding (I say this as a blogger). This trend is still continuing.

5.       The government bureaucracy has had an increasing appetite for academic involvement. Hence there has been an explosion of groups loosely affiliated with universities that essentially do work for the ministries. They write reports on poverty, Aboriginal issues, health, urbanisation, transport, tourism, etc. They evaluate proposals by industry and other ministries. They set up courses for the bureaucracy and in almost every other way provide intellectual services to ministries that seem to treat academia as a substitute for in-house expert knowledge.

In short, it has been a mixed bag for the incentives to do work on Australian policy. At the top research end, all the incentives are not to bother and the best students who could be interested in it are actively dissuaded from pursuing Australian issues. At the bottom teaching end, the reduced status of teaching has lead to a large reduction in the amount of local content that will be taught to the next generation, just by virtue of who teaches it. But elsewhere, work on Australian issues has blossomed like never before. There are thus more than enough report-writers, too few teachers, and  perhaps too few top academics involved in Australian economic and social policy issues.

The problem at the teaching side is basically unsolvable. You would have to double academic teaching salaries to make a dent on their low status and such a move, whilst not actually that expensive, will be bitterly resisted by everyone else (i.e. the researchers and the administrators) because it would lower their relative standing. They would most certainly veto any change or pervert it so that it first increases their salaries. Even if one is capable of bypassing all the forces rallied against increases in teacher pay, it would take decades for the higher incentives to be an academic teacher to work through into more home-oriented teachers to enter the system. The best you can probably do in the short-run is to have additional courses on local content for the civil service, which is a market that already exists and seems to be working quite smoothly.

Presuming that one sees it as a problem that our best minds are turned away from internal issues (though of course you have the odd exception of someone so interested in local issues that they are prepared to accept a hit to their academic standing), the question is what can be done. The following possible actions come to mind:

1.       Set up a world-beating PhD education system. Australia has enough smart academics to teach at world-standards in almost any field of social science, but individual universities do not have the clout to offer their PhD students something comparable to overseas PhD places, like the European Institutes, or the Ivy league programs in the US. This means that many of our best young minds are sent overseas and don’t come back, whilst it is the home-grown young minds that are far more likely to be interested in Australian policy than the imported academics. Thus, effectively, the lack of world-class PhD programs prevents our best minds from working on Australian issues. We can set up such institutes, again at a cost that is not too great. What has worked for the Europeans is to have a dedicated PhD institute with an academic governance structure that limits the time any individual academic spends at such a place. Hence what one needs is a European Institute type scenario where one has dedicated PhD programs, located in one of Australia’s major cities.

2.       Open up the data files. It has been said many times before but it clearly needs to be said many times again: Australia actually collects great databases. Good enough so that their use could interest international researchers. However, the ABS is too afraid of the privacy laws to make it easy for academics to use them. I for instance know of a recent case where someone had to wait 18 months to get information on whether respondents to a particular survey lived in NSW or Victoria. By the time the permission came, the researcher involved had already left! The reality of international research is that one looks for low-effort, high-gain data, and the barriers put in place to use the Australian datasets are simply too high to bother. This is in a sense a ridiculous state of affairs because it means that Australian tax payers fund large datasets which are not used whilst they also fund researchers who work on American and European datasets! There is an easy solution, and it is to be much less prissy about privacy concerns. Abuses of data virtually never happen and you can punish people afterward anyway. Yet, the ABS still does not bring out data that links individual information to their medical records, their military records, their employment and tax records, etc. It has the technical know-how to do it, but thinks it is shackled by privacy concerns. A most inefficient state of affairs that is a big turn-off for academic researchers who play the international game.

3.       Do not introduce specific measurement devices for policy impact because a poor measurement instrument is worse than no measurement at all: for academics, the main reason to be interested in local policies is because one wants to make a difference. Media attention, policy impact, ego, etc. As soon as you start directly rewarding any observable impact though (which is only very loosely related to real impact), you will create a wall of noise because the market for policy advice will be flooded by academics who want to crank up their ‘policy ratings’. Institutes will specialise in targeting particular bureaucrats with their publications in the hope of being cited. Middle-range junior academics will start to nit-pick with policy reports in order to get a mention in the next one. Blogs, tweets, and websites will flood the market such that no-one will have much clue what is useful and what is not. As a result, the marginal rewards to policy work will actually reduce for the good academics who will then simply no longer bother. Australia has already had experience with this kind of perverse effect, which is really just a standard effect that occurs when you replace a market mechanism by imperfect bureaucratic measurement: the ‘DEST’ system for measuring productivity. Under the DEST system, every publication and every conference started to count as ‘output’. As a result, Australia witnessed an explosion in fake conferences, conference volumes, local academic journals, etc. The ‘publications’ in those conferences and journals all counted towards DEST and hence they were a way to prop up the CV’s. As a result though, many academics stopped coming to the local conferences and no longer sent anything to local journals, simply because it was no longer worth their while and started to be a bad signal. And one can even argue that the ‘productivity explosion’ engendered by the DEST formula was worse than worthless in that it diverted effort away from more worthwhile things to do (including leisure time with the family). Hence, let the bureaucracy and the politicians work out themselves which academics they want to contact about policy work (academics don’t get paid so much that they will really say no if offered the right price), and don’t mess that market up by inept measurement because that is worse than no measurement.