Research in Industrial Organization: The Harris-Vickers Collaboration
In the years that followed, Harris and Vickers went on to develop their ideas further. The Harris and Vickers partnership was to bear much fruit. Vickers admits modestly that while he has a good nose for an interesting problem, cracking it rigorously can call for yet greater mathematical skills than he himself might muster.
They continued to work together, collaborating in their 1993 paper with Christopher Budd, and with Philippe Aghion for a 1997 paper. Later on, Vickers and Harris built on that work with Aghion and Peter Howitt in tackling a central question in the theory of endogenous growth (Aghion et al. 2001): What is the relationship between competition between firms, and technological progress? A further Harris and Vickers paper appeared in 1995. All but three of these six papers were published in the Review of Economic Studies. The fourth, their first in print, Harris and Vickers (1985a), had appeared in the field journal, the Journal of Industrial Economics, the 1997 paper with Aghion appeared in the European Economic Review, and the last, with its focus on natural resources, would come out in the RAND Journal of Economics.Various models of endogenous growth can involve externalities, or population growth, or fossil fuel extraction, once one of these phenomena is allowed not just to influence, but genuinely to interact with, the standard dynamics of aggregate output and capital. Still more appealing in the view of many, and absolutely consistent with the phenomenon of Harrod-neutral technical progress, is the evolution of human capital and the role of training in economic growth, which was first demonstrated rigorously by Lucas (1988). But what interested Vickers was technological progress. For many years, economic growth models had treated it as a mysterious exogenous parameter, if they allowed for it at all. Meanwhile, most of the IO literature about innovation tended to focus, alas, on the narrow canvas of partial equilibrium.
So, technological progress would become the most celebrated member of the endogenous growth family of models. But in 1985, hints from Arrow's concept of learning by doing aside, that lay some years in the future.Technological progress took three stages for Vickers—first, research, then discoveries and finally, implementation and diffusion. He was convinced that inventors are typically not loners. Observation and history teach that most of them, whether successful or no, are far from solitary; rivalry is involved; there is a race. That is precisely what Vickers set out to model initially alone and soon with Harris.
The framework was deliberately kept as simple as possible. There are two inventors. They are running in the same race towards the same objective—discovering a particular new product or process. The one who will get there first will scoop all the cream, by securing the patent or the prize. What is needed in the race is not just stamina, which they both have, but effort, which is privately costly. At some early stage in the race, they might conceivably be running together, or, more likely, one may be ahead of the other, possibly by a whisker, possibly by more. One question is this: If one of them is in front, does he supply more effort than the rival behind him, or less? Another question: Does the gap between the two runners tend to widen, or to narrow, as the race proceeds?
What are the answers? If there is symmetry between the two runners, if they are patient, and if the race has two stages, the one in front always supplies more effort at stage one; and the effort gap goes up as the distance between them widens. These answers are general. In a many-stage race, the same findings hold in particular examples. If the runners are impatient (or equivalently, if the prize is discounted over time), the desire to get there sooner makes the front runner try even harder. These are the findings proved in the 1993 paper with Budd, which introduces, and focuses upon, the challenging extension to uncertainty.
The setting of the 2001 paper with Aghion and Howitt, as well as Harris, combines elements of the important Aghion and Howitt Econometrica 1992 paper—which, along with Lucas (1988) on training and Romer (1990) on invention, constitutes the core of modern endogenous growth theory—with the Harris and Vickers (1987) paper summarised above. In Aghion et al. (2001), the runners already produce products, but the extent to which those products compete with each other can vary. The runners do not just observe each other; it is possible that they can imitate each other as well. Innovation proceeds incrementally, step by step.
The key question here is this: Will the growth rate tend to increase with (a) closer competition in the product market, and (b) the probability of imitation? At the risk of oversimplification, one can state that the answer to (a) is yes, usually. To (b), the answer is yes, but for sure only if it is low enough, and definitely not if it is too high. If a policy maker seeks to maximise growth, therefore, we welcome stronger competition in the product market—ironically, because that whets the competitors’ desire to escape it—but she should turn her face strongly against anything but a low chance of imitation. Too much copyright infringement definitely needs to be stopped. But all that presumes, of course, that faster growth is good for welfare, which the Aghion and Howitt (1992) paper shows could quite possibly be false. You can have too much growth in that model, where the winner takes all, at least until she is usurped at the frontier by someone else.
Research racing was not the only field that Vickers explored while completing his doctoral thesis. There were others. Six further articles appeared in 1985 and 1986 (Vickers 1985a, b, c, d, 1986a, b), varying in length, but all single authored, and devoted mainly to aspects of theory and policy in oligopolistic markets. One was first given at a Royal Economic Society (RES) conference and chosen for publication in the Economic Journal.
Here, Vickers (1985a) looked at why the owner of a firm, call it A, in a Cournot oligopoly, whose main concern is the firm’s profits, might want to employ an aggressive manager with a quite different objective. If the manager disregarded costs, she would raise her firm’s output. If each of the rival firms, owner-managed by assumption, continued to set its output to maximise its profit, taking all other firms’ outputs as given, it would cut back, earn less profit. But A’s profits would go up, even though the industry’s total profits would drop. The moral was this: If you want to make as much profit as possible, try to do something else, so long as your rivals don’t! Vickers’ aggressive manager makes her firm behave rather like a Stackelberg leader.The other members of this quartet of early, single-authored papers included a crisply written survey of modern oligopoly theory much read by students (Vickers 1985b), as well as others on predation (Vickers 1985c) and preemptive patenting (Vickers 1985d).
5.3