We Must Experiment to Move Economics Forward in the Middle East
From the days of antiquity, through to the middle of the twentieth century, economics was largely a deductive and narrative discipline. The leading contributions were dense treatises hundreds of pages in length, with scarcely a number or equation in sight.
Economists influenced policymakers through the general principles that they espoused, such as David Ricardo’s opposition of the protectionist English Corn Laws, rather than the precise, numerical prescriptions that a medical researcher would present to a health minister.
Two developments transformed the discipline during the latter half of the twentieth century. First, western governments started to institutionalise the process of data collection, providing economists with a large volume of professional-grade data to analyse. Second, the development of computers provided researchers with the tools required for sophisticated statistical analysis.
Thus, whereas a leading 19th century economist looked like the philosopher and polymath John Stuart Mill, a leading 20th century one looked like the Nobel laureate James Heckman, who spent several decades studying the effect of education on labour market earnings. Economists now supported government policy with fine-tuned recommendations based on rigorous data.
Today, the discipline is once again metamorphosing as experiments enter the economist’s toolkit. Regular, non-experimental data is incredibly useful, but it suffers from a key drawback – it is very difficult to isolate the causes and effects of the phenomena being studied. To see why, consider the No Child Left Behind (NCLB) legislation that the US president at the time George W Bush enacted in 2001.
The policy required standardised tests for all schoolchildren, and it prescribed improvement plans for underperforming schools. Economists and policymakers were keen to evaluate the policy’s effects on educational outcomes; however, how could they distinguish between the effects of NCLB and other changes from the early 2000s?
There was a recession in 2001 after the dot-com bubble burst; there was the September 11 terrorist attack; there was the invasion of Afghanistan and Iraq; and many other developments that make it very difficult to isolate the role of NCLB.
The Scottish physician James Lind pointed the way out of this problem in the 18th century while working for the Royal Navy. Sailors would regularly get scurvy (a terrible disease), and Lind suspected that lemon juice might constitute a cure. Had Lind given all of the afflicted sailors lemon juice and observed an improvement in their health, then how could he be sure that it was the lemon juice rather than the weather, the air pressure, or anything else that changed at the same time? Anticipating this, he randomly selected a subset sailors whom he did not give lemon juice (the control group). When the health of the control group did not improve, Lind could be confident that the lemon juice caused the improvement in the treated group’s health.
Randomised control experimentation rapidly became the standard in natural sciences, and today it is beginning to dominate economics, allowing scholars to make policy recommendations with a much greater degree of confidence. For example, a recent study by Greer Gosnell, of the London School of Economics, and John List and Robert Metcalfe, of the University of Chicago, demonstrated how giving airline pilots simple fuel-saving incentives resulted in huge carbon abatement at a far lower cost than most government environmental policies. Paul Ferraro and Mike Price of Georgia State University recently authored a study that showed how providing consumers with information on their water usage compared to that of their neighbours generated considerable savings in water consumption. Roland Fryer of Harvard University recently experimented with giving high school students financial incentives, and he found that attaching the incentives to inputs such as reading books or doing homework was more effective than outputs test scores.
Such policy insights were virtually impossible without the deployment of randomised control experiments. In July 2016, the University of Chicago hosted a workshop, the Summer Institute for Field Experiments, featuring scholars with a background in experimental methods and representatives of governments, companies, and non-profit organisations seeking to deploy experimental methods in their line of work. The aim was to create partnerships between the two sets of participants to raise the quality of decision-making by policymakers in all sectors of the economy. My role was to deliver a workshop on advanced epistemology to help the scholars design their experiments more effectively.
The UAE and the GCC are well-positioned to benefit from these intellectual developments as they embark on fundamental economic reforms. Abu Dhabi recently partnered with the NYUAD professor Yaw Nyarko and colleagues to conduct an informal experiment regarding migrant worker regulations. The UAE should be commended for its desire to ground policy decisions in cutting-edge research, and the next step would be for it to introduce formal randomised controls. Western governments, including the United States and the United Kingdom, have established units dedicated to encouraging policy experimentation, and to analysing the results in a rigorous fashion. Given the modest cost of running experiments and the potentially massive gains in terms of effective policymaking the GCC countries should strongly consider adopting such methods.
While it might look like the GCC countries have a lot of catching up to do globally when it comes to adopting cutting-edge scientific techniques, they should take heart from the rich tradition of science in the Middle East over the past 1,500 years. Scientists such as Ibn Al Nafis (1213-1288) made important contributions to human knowledge, as well as helping pave the way for the Renaissance that Europe experienced during the 15th century.
The Danish physicist Niels Bohr once described an expert as “a person who has made all the mistakes that can be made in a very narrow field”. For the GCC societies, embracing formal experimental methods and analysing data in a rigorous fashion are the quickest ways to make those mistakes and acquire the expertise that Bohr referred to. He got the 1922 Nobel Prize in physics – maybe a GCC citizen can do the same in 2022.
This article was originally published by The National.
Omar Al-Ubaydli is a non-resident fellow at the Arab Gulf States Institute in Washington.