Thoughts on Normative and Positive Economics

Can economics reach definitive concrete conclusions that can be dubbed “laws” of economic phenomena in the same way that there are laws governing physical reality? In a way, there seems to be little reason as to why economics cannot reach similar conclusions to other sciences such as physics or biology. As with other sciences, economics uses the scientific method to reach its conclusions. The method which biologist T.H. Huxley claimed is: “nothing but the expression of the necessary mode of working of the human mind. It is simply the mode at which all phenomena are reasoned about, rendered precise and exact.”[1] Meaning that science is not actually anything special, and many modes of inquiry, besides economics, could qualify as science. Still, there is something that may separate economics from what people conventionally think of as the sciences. In economics beliefs and ethics are deeply intertwined in the field, that does not characterize other sciences. For example, an optimal economic policy prescription to achieve economic growth is nowhere near as clear as an optimal prescription to cure a case of the measles. An optimal economic policy will vary by what ethical assumptions an economist makes. Having good health may therefore not be as clear as having a good economy. In economics, a distinction is normally made between positive economics and normative economics. Positive economics is described by Milton Friedman as being “independent of any particular ethical position or normative judgments” or as Keynes put it, deals with “what is” not with “what ought to be.”[2]  Normative economics in contrast deals with the “ought,” as Keynes put it.

Since the financial crisis, economics has had a bad rap, and it has become popular to claim that conventional economics completely missed the crisis. This is partly true, and many economists have openly admitted the existence of problems. Still, the field was not completely blindsided and foresaw the crisis through mostly conventional economics. Nonetheless economics has been laughed off as a science because of, among other things, a lack of consensus within the field, weak predictive power, and the fact that it deals with how humans interact. Some critiques, like those of Nassim Taleb, that showcase the hubris of economists in their ability to predict the future are poignant and accurate. Yet many popular critiques are exaggerated and unqualified in disproving the viability of economics as a science. Many sciences face similar problems to economics. If consensus is a requirement then surely nutrition should be disqualified as a science, or if predictive power is a requirement, meteorology should also be disqualified, and if it is because dealing with humans makes it different then how can medicine be a science?

Historically the problems faced by economics today are very similar to those faced by early science. For centuries there was little clear agreement on many basic principles, and science was terrible at explaining things. All science, as Bertrand Russell has argued, is in a continual evolution to understand more precisely various phenomena. He claimed that, “every advance in a science takes us farther away from the crude uniformities which are first observed, into greater differentiation of antecedent and consequent, and into a continually wider circle of antecedents recognized as relevant.”[3] Science is often seen as showing or trying to explain how B was caused by A when really it is continually trying to get a more precise or highly probabilistic account of causes. Hence as Russell claimed “The principle ‘same cause, same effect,’ which philosophers imagine to be vital to science, is…utterly otiose”[4]. Newton’s physics are a very good description of how physics generally works, but no physicist today seriously practices conventional Newtonian physics as they are imprecise. Similarly Keynes view of economics as sketched in his General Theory is also a good general framework of how macroeconomics generally works, but again no economist today claims to be practicing conventional Keynesian economics as described in the General Theory as the work is highly imprecise.  Further, while there is still significant disagreement among economists, a consensus on many things exists. A symposium almost 20 years ago in the American Economic Review asked leading macroeconomists if there was a core of practical macroeconomics and the various papers reached generally similar conclusions that continue to be true today.[5]

The potentially more pressing problem with positive economics is how it’s intertwined with normative economics. In theory, there should be a clear divide between the two, but economic statements are inevitably tied to ethical claims. Even if economists just focus on the most efficient solutions in whatever they are exploring, even claiming that the most efficient outcome is best is an ethical claim. For example, economists often strive to reach Pareto optimality in their models which is the state in which it is impossible for conditions to make an individual better off without making someone worse off. This though may not actually be the best outcome depending on what the ethical beliefs of an individual or a society may be. A common critique being that distribution of resources and general equity are disregarded with Pareto optimality. Therefore a pie being distributed between three individuals would be Pareto optimal if the two halves of it were given to two people, and the third got none of it. The third person is clearly not worse off by the distribution, but many would still claim the allocation to be unfair. Within economic modeling such assumptions are ripe, and competing models exist which claim to forcefully prove opposing conclusions.

This in general points to a problem in how models are constructed, a problem that may be growing worse due to the intense use of mathematics in economics. Economist David Romer has described this phenomena as “mathiness. Where economists shroud their models in math in order to reach conclusions they want to be true[6]. Romer claims that the math in academic economics papers today is hardly taken seriously as the mathematical proofs that are supposed to underlie economic theory are carelessly constructed. He notes a recent written paper by Robert Lucas, a Nobel Prize winner and one of the most cited living economists, in the Journal of Political economy, one of the most coveted economics journals, where a simple mathematical mistake was made and overlooked. Romer explains “the fact that this oversight was not picked up at the working paper stage or in the process leading up to publication may tell us something about the new equilibrium in economics. Neither colleagues who read working papers, nor reviewers, nor journal editors, are paying attention to the math”[7]. Romer’s attack is primarily on economic methodology and specifically the construction of economic theory, but the carelessness of this “science” seems to stem from a desire to argue for positions economists just want to be true.

So economic theory may be problematic, but what of empirical economics that purely derives its conclusions from data. Surely that should be healthier as how can data lie? Yet anyone who has read any sort of empirical economic paper will quickly see how much of a game it is. Economists often spends most of their paper describing where they got their data, why the used the particular data they, why certain variables or data points were excluded or included, why certain things had to be “corrected”, why certain “instruments” had to used, and so on. As Nobel Prize winning econometrican Christopher Sims claims: “The same data generally are subject to multiple interpretations…we are always combining the objective information in the data with judgment, opinion, and/or prejudice to reach conclusions”[8]. Data by itself cannot lie, but it can be twisted to point to the conclusion one wants to be true.

An example of empirical problems can be seen in the seminal paper The Colonial Origins of Comparative Development: An Empirical Investigation, shortened to AJR for the last names of the authors, which made a strong empirical argument that institutions matter for long run economic growth something that in the past was difficult to quantify and formally prove. The basic claim of the paper is that Europeans adopted different colonization strategies and set up different institutions that were dependent on the feasibility of European settlement. So where there were low mortality rates and it was easy for Europeans to settle, like in the US or Australia, they were able to set up inclusive institutions that were conducive for long run economic growth. In contrast, where there were high mortality rates, such as in the Caribbean or sub-Saharan Africa, for Europeans and few were able to settle extractive institutions were built that were not conducive for long run growth. Inclusive institutions, being based on a strong protection of property rights, and extractive ones, on rent seeking. An initial reading of the paper is compelling as a strong narrative is backed up by convincing statistical results. The conclusions of the papers are not universally accepted though with one paper claiming that the data in AJR is faulty and the variables used, incorrectly identified.[9] Another paper directly challenges AJR’s conclusions, and claims that it is not institutions that the settlers brought and established, but the human capital that they brought which established economic growth[10]. The paper subsequently conducts its own statistical analysis “proving” its conclusions and also claims AJR misspecified their variables. AJR is still widely cited today as are its response papers, although less so. Yet how can it be that opposing conclusions are accepted in a field that claims to be a science making progress?

It seems clear that biases are an inherent problem within how economic analysis is conducted. Yet while this is clearly a problem, it is not a problem unique to economists. Problems with how models are constructed are also problematic in other sciences. Theoretical physicists also make assumptions in the construction of their models, and can have personal attachments to a theory they personally find convincing, disqualifying an impartial view. Yet no one seriously disqualifies physics from being a science that can reach convincing conclusions. The problems of empirical research in the sciences has also been highly problematic lately as data from experiments have been cherry picked to reach favorable conclusions. For example, Amgen, an American drug company, recently attempted to replicate 53 seminal cancer research studies and were only able to successfully replicate six of them[11]. Bayer Health did something similar with 67 studies and they were only able to reproduce about a quarter of the results[12]. The “mathiness” concerns Romer has discussed are also problematic in other sciences although in reverse. Instead of complicated mathematics being used to prove conclusions overly simple statistics are used to prove scientific conclusions. In scientific papers, if experimental results cross the fragile 5% statistical significance threshold they often are instantly accepted as scientifically true into academic journals. This simplicity though can easily cloud more complex phenomena. All these concerns are serious problems faced by science today, but none of them should discount the ability of science to achieve convincing conclusions just as economics should not be discounted.

Still, ethical questions remain for what role economists should actually play. Should economists only attempt to describe what is possible, and at least try to act as unbiased scientists? Or should they be allowed to argue for things they actually believe in? Currently, with economists there often is a dichotomy between the serious scientific work they conduct, and more speculative normative work where they argue what they believe should be done. For example there is an enormous difference between the academic work that Paul Krugman or Milton Friedman have conducted that is widely respected by economists as against their more controversial public personas. The danger is if the positive and normative get intertwined which, unfortunately, does often occurs. Economists frequently disingenuously claim that positive economics clearly backs up their normative claims. In a recent example economist Jeffrey Sachs attacked Paul Krugman for blatantly lying about the nature of the economic recovery in the UK relative to the US. Krugman could not accept that a conservative led government in the UK had a recovery that was comparable to the one in the US because it violated his fierce liberal ideology.[13] Yet, somewhat ironically, Sachs himself has been no stranger to overinflated hubris in his high flung beliefs about economic development. If economists hope to seriously play both roles they have to be careful in how they practice them

Thomas Piketty’s recent successful Capital In the 21st Century is a good example of where economists play both personas interchangeably. The positive economics is seen where he empirically shows that inequality is clearly on the rise and approaching levels not seen since the late 19th  and early 20th century, which should rightfully be acknowledged as factual scientific work. His normative claims though that something must be done, and his prescription of how the problem should be solved are much more contentious and speculative. Should Piketty, and economists in general, simply say that something is happening then hand over to ethicists and let them figure out the right thing to do, then to policymakers to figure out what can be done, and then to the polity to actually decide what should be done? John Broome, an MIT trained economist who now teaches ethics in the Oxford philosophy department has a good possible resolution to what should be done.  He claims that: “Economists are servants of the society, and surely they should do things the way the society wants…they should follow the preferences of the people. It would be undemocratic if economists were to insist on their own judgements of priority, and impose them in their work.”[14] Therefore economists should be allowed to argue for what they believe in as long as long as they are clear that they are arguing for what is merely their belief. Positive economics may back up or support of their claims, but it does not validate that their claim is ethically right or what really ought to be done. What ought to be done, as Broome argues, should be left to the polity at large, which economists should be a part of. As difficult as it may be, economics must strive to keep the positive, and the normative separate if it has any hope to be taken seriously as a science or to make any progress.

 

Footnotes:

[1] Huxley, TH. “Internet History Sourcebooks.” Internet History Sourcebooks. Accessed July 28, 2015

[2] Friedman, Milton. Essays in Positive Economics. Chicago, Ill.: University of Chicago Press, 1953. 4.

[3] Russell, Bertrand (1953). On the notion of cause. In Proceedings of the Aristotelian Society. Pelican Books 171-196

[4] ibid

[5] The American Economic Review, Vol. 87, No. 2, Papers and Proceedings of the Hundred and Fourth Annual Meeting of the American Economic Association (May, 1997)

[6] Romer, Paul M. 2015. “Mathiness in the Theory of Economic Growth.” American Economic Review, 105(5): 89-93.

[7] Ibid, 92

[8] Sims, Christopher A. 2010. “But Economics Is Not an Experimental Science.” Journal of Economic Perspectives, 24(2): 60.

[9] Albouy, David Y. 2012. “The Colonial Origins of Comparative Development: An Empirical Investigation: Comment.” American Economic Review, 102(6): 3059-76.

[10] Glaeser, Edward L., Rafael La Porta, Florencio Lopez-de-Silanes and Andrei Shleifer. “Do Institutions Cause Growth?,” Journal of Economic Growth, 2004, v9(3,Sep), 271-303

[11] “Trouble at the Lab; Unreliable Research.” The Economist, October 19, 2013.

[12] ibid

[13] Sachs, Jeffery. “Krugman’s Anti-Cameron Contradiction.” Accessed August 2, 2015.

[14] Broome, John (2008). Why economics needs ethical theory. In Kaushik Basu & Ravi Kanbur (eds.), _ Arguments for a Better World: Essays in Honor of Amartya Sen: Volume I: Ethics, Welfare, and Measurement an d Volume Ii: Society, Institutions, and Development