Why economics should be in the HEFCE part-laboratory cost band

The classification of economics as a ‘low-cost’ library-based discipline has long-been a source of grievance to English academic economists who see subjects like business and management, with lower technical resuirements, receive more generous funding by the Highr Education Funding Council for England. In autumn of last year John Beath, then chair of the Society’s Conference of Heads of University Departments of Economics (CHUDE) and Stephen Nickell, the Society’s President, put a strongly-argued case for change to HEFCE. Although HEFCE’s response was not overly encouraging it does hold out hope for some improvement. The case and the response are set out here.

Economics and Econometrics in England is currently classified as a library-based subject in band C (low cost). This is the lowest possible cost group, on a par with Arts subjects, for example. By comparison Business and Management, though also in band C, is classed as part-laboratory; Geography is part-laboratory and in band B; Statistics and Operational Research is part-laboratory, band A; Pure mathematics is part-laboratory, band B.

Economics and Econometrics plays the principal role in the quantitative analysis of economic, social and environmental data. Such data sets are large and complex and considerable resources, both human and technical, are involved in their effective analysis. This has required the development and use over the last twenty years of advanced statistical and mathematical techniques as well as the use of experimental methods to study and analyse business and government applications such as auctions, behaviour subject to regulation, and the like. This has been the major direction of development in the subject and is something that is not captured in the historical cost data reflected in the current band in which Economics is placed. The similarity with Statistics and Operations Research is striking. On a simple comparison of the subject matter and methods of research, Economics and Econometrics should be at least treated as part-laboratory. This has been the case in Scotland and is also partly the case in Wales where quantitative modules in Economics are funded at a higher rate than non-quantitative ones.

Further particular arguments follow. First, we present some direct data on research costs by subject, where we show that economics is comparable with other part-laboratory subjects. Second, we provide some information on the use of econometric and experimental economics laboratories in UK universities. Then, because we are essentially arguing for a transfer of resources to economics, we present information which suggests that such a transfer is a sensible policy. Thus, in the third section, we discuss the shortage of economists and finally we provide some evidence on the relative outputs generated by teaching expenditures in a variety of subjects.

Research expenditure on technical equipment and staff
Where does economics stand, relative to other subjects, in the percentage of research expenditures devoted to equipment and technical staff? The following data have been obtained from the ESRC and cover all Large Grants, Small Grants and Fellowships active on 10 January 2003.

The data reveal that the equipment/technical staff expenses in economics research is comparable to that in part-laboratory subjects such as geography and statistics/computing and far exceeds such costs in the other ‘library-based’ subjects. However, these ESRC figures may give a false impression because they do not necessarily reflect the technology costs required to do serious empirical research in economics. Two major centres of such research in the UK are the LSE Research Laboratory (combines the Centre for Economic Performance, the Suntory/Toyota/Centre for Economics and Related Disciplines and the Financial Markets Group) and the Institute for Fiscal Studies. The former spends some 15 per cent of its £3 million total cost on technical staff and equipment, the latter around 5 per cent of its £3.2 million total cost. These figures tend to be far higher than the ESRC numbers in part because of composition effects and in part because these major centres of economics research are less constrained in the allocation of their expenditures than purely ESRC-funded research.

Overall, these numbers suggest that research costs in economics are more comparable to Part-Laboratory based than to Library based subjects.

Econometric and experimental economics laboratories
Economics departments in the UK make use of two distinct types of laboratories. The first consists simply of a room full of networked PCs which is used for teaching statistics, econometrics and analytical economics. The second is a room, or set of rooms, with a large number of workstations which can be flexibly configured. These are used to undertake research in experimental economics. This research uses human subjects in an experimental setting to analyse aspects of economic behaviour such as individual decision making, the workings of markets and the outcomes of differently structured auctions. The classic practical auction study was the long period of experimental research carried out at UCL to provide the optimal auction structure for the Government’s sale of 3G licences.

From a sample of sixteen of the large UK economics departments, we ascertained that all made extensive use of statistics/econometrics reaching laboratories, to the tune of around 15 to 30 percent of graduate and undergraduate teaching time. Exactly half of our surveyed economics departments had dedicated econometrics laboratories, the other half using centrally provided facilities.

Turning to experimental laboratories, around 40 percent of the departments had an experimental economics laboratory with one department sharing with the psychology department. Experimental economics is a rapidly expanding branch of economics, so this number is bound to grow. Overall, therefore, the words ‘part-laboratory’ seem to be a perfect description of this scale of activity.

Two types of shortage are relevant. The first arises because of the high earnings available to economists outside the academic sector. The resulting difficulty in recruiting high quality academic economists to universities has been well documented (see Machin and Oswald, 2000). Recent data on Ph.Ds suggest that this is unlikely to be rectified in the short-term. In 1994-95, 164 PhDs were gained in economics in British HEIs and of these only 62 were UK nationals (37.8 per cent). By 2000-01, this had grown to 279 Ph.Ds in economics but the number of UK nationals in this figure was a mere 79 (28.3 per cent) (Table 10.11 in Commission on the Social Sciences, 2003). The implication of this is that academic economists are, and will continue to be, much more expensive than academics in most other subjects. The second type of shortage is that of academics with the quantitative skills needed to work on the large and complex datasets now available for social scientists. Economists are particularly well placed to be involved in their analysis because their subject is on the frontier when it comes to using quantitative methods in the social sciences, including social policy analysis. However, the shortage of economics graduates to work on these problems is desperate and this has now spread into the public sector, particularly in the Government Economic Service and the Bank of England. While one response to this is to hire from abroad, this cannot fully compensate for the underlying shortage.

The arguments in the first two sections are based on input costs. We are arguing here for a shift in resources towards economics and it might be helpful, therefore, to provide some evidence on the relative outputs generated by expenditures on various subjects.

An analysis of earnings data reported in Table 2 reveals that the earnings gap between those in the workforce with an economics degree and those with only two or more A levels is greater than the equivalent gap for those with any other degree, except accountancy. If one puts to one side the strictly vocational subjects (law and accountancy), economics is top. This gap is important because it indicates the value added by a degree in economics.

Looking at the numbers in more detail, we see that, in terms of pay subsequent on graduation, Economics degrees generate a gap (value added) which is around 40 per cent larger than for degrees in Business and Management or Engineering, despite the fact that the resource costs of generating these latter degrees are currently considerably higher. These data provide a prima facie argument for shifting resources towards the production of Economics degrees. Perversely, the incentives in the funding system have been encouraging the opposite. The trend has been for Economics groups in universities to be absorbed into Business and Management. This has been encouraged by the belief (and there is evidence for this) that economists of a given quality typically score more highly in the RAE if they are located within Business and Management and submitted to that panel. Associated with the move of Economics departments into Business and Management has been a de-emphasis of the single honours degree in Economics and this has potentially significant implications for the supply of quantitatively trained graduates for which there is a recognised national need, as we have already noted. Students trained in Business and Management rather than Economics are inadequately prepared to understand and use effectively the quantitative methods needed for the analysis of economic, financial or business issues.

The purpose of this note has been to argue that Economics and Econometrics should be upgraded to the Part-Laboratory cost band from the Library-Based cost band, where it currently languishes. We have noted the following points.

• Teaching and research in economics is, at least in part, undertaken in IT and experimental laboratories which mean that average costs are comparable to the typical Part-Laboratory subject.

• There is a significant shortage of economists in both academia and in the wider community.

•The value-added generated by an economics degree (over and above an individual with 2+A levels) is significantly higher than for any other degree subject with the exception of the specifically vocational subjects, law and accountancy.

These facts suggest that upgrading Economics and Econometrics to the Part-Laboratory cost band is consistent with both equity across subjects and overall economic efficiency.

The response so far...

Unfortunately, in its announcement of 23 December 2003, HEFCE made no mention of any decision to raise the cost band of economics. However, there is some good news in the HEFCE announcement.

• HEFCE proposes to move to an alternative way of determining subject weightings based on full economic costs. However, this will take several years.

• The base price for the lowest cost band (D), where economics currently resides, is expected to rise by 20 per cent in 2004-5. Furthermore, the weightings for the four cost bands are going to be A = 4, B = 1.7, C = 1.3, D = 1. This contrasts with the previous weighting, which is A = 4.5, B = 2, C = 1.5, D = 1. This change clearly benefits all subjects in D, including economics, relative to those subjects in all the other cost bands.

Commission on the Social Sciences (2003), Great Expectations: The Social Sciences in Britain, March.

Machin, S. and Oswald, A. (2000), ‘UK Economics and the Future Supply of Academic Economists’, Economic Journal (Features), 110, pp.F334-49

Page Options