Publications

From July 25 to 31, 1796, Georg Wilhelm Friedrich Hegel, then working as a private tutor for the aristocracy in Bern, took a mountain hike in the neighbouring Alps. Hegel travelled from Thoune to Altdorg via the Jungfrau and the Uri, a land of glaciers. As Hegel began studying economics for good, the query would reappear: in his reading of Sir James Steuart’s Inquiry into the Principles of Political economy, Hegel would make his first step into economic theorizing. In Frankfurt, Hegel was not yet a tenured Gymnasium professor. He was again a private tutor, experiencing hardships of a salaried life – though in wealthy families. Paul Chamley selected excerpts of interest based on his first assessment of the thesis that there surely exists a solid ‘system of political economy’ by Hegel. He assumed it rather than he found it as a result of his comparative study.

Many regulations with first-order economic and environmental consequences have to be adopted under significant scientific uncertainty. Examples include tobacco regulations in the second half of the 20th century, climate change regulations and current regulations on pesticides and neonicotinoid insecticides. Firms and industries have proved adept at exploiting such scientific uncertainty to shape and delay regulation. The main strategies documented include: hiring and funding dissenting scientists, producing and publicizing favorable scientific findings, ghostwriting, funding diversion research, conducting large-scale science-denying communication campaigns, and placing experts on advisory and regulatory panels while generally concealing involvement. In many cases, special interests have thus deliberately manufactured doubt and these dishonest tactics have had large welfare consequences.
Largely and unduly neglected by economists, these doubt-manufacturing strategies should now be addressed by the field. Here, we first present a simple theoretical framework providing a useful starting point for considering these issues. The government is benevolent but populist and maximizes social welfare as perceived by citizens. The industry can produce costly reports showing that its activity is not harmful, and citizens are unaware of the industry’s miscommunication. This framework raises important new questions, such as how industry miscommunication and citizens’ beliefs are related to scientific uncertainty. It also sheds new light on old questions, such as the choice of policy instrument to regulate pollution. We subsequently outline a tentative roadmap for future research, highlighting critical issues in need of more investigation.

The bootstrap is a technique for performing statistical inference. The underlying idea is that most properties of an unknown distribution can be estimated as the same properties of an estimate of that distribution. In most cases, these properties must be estimated by a simulation experiment. The parametric bootstrap can be used when a statistical model is estimated using maximum likelihood since the parameter estimates thus obtained serve to characterise a distribution that can subsequently be used to generate simulated data sets. Simulated test statistics or estimators can then be computed for each of these data sets, and their distribution is an estimate of their distribution under the unknown distribution. The most popular sort of bootstrap is based on resampling the observations of the original data set with replacement in order to constitute simulated data sets, which typically contain some of the original observations more than once, some not at all. A special case of the bootstrap is a Monte Carlo test, whereby the test statistic has the same distribution for all data distributions allowed by the null hypothesis under test. A Monte Carlo test permits exact inference with the probability of Type I error equal to the significance level. More generally, there are two Golden Rules which, when followed, lead to inference that, although not exact, is often a striking improvement on inference based on asymptotic theory. The bootstrap also permits construction of confidence intervals of improved quality. Some techniques are discussed for data that are heteroskedastic, autocorrelated, or clustered.