Assess whether a radical reconsideration of the more traditional approaches to policy evaluation employed by economists is called for. [35]

H3 A level Economics Cambridge N2010 Q2 essay response

A controlled experiment generally compares the results obtained from an experimental sample against a control sample. 

“To determine the effectiveness of specific policy initiatives, we need to know how the target populations are likely to respond. This can only be discovered by conducting controlled experiments.” 

In the light of this statement, assess whether a radical reconsideration of the more traditional approaches to policy evaluation employed by economists is called for. [35]

There has been a rise of interest in experiments in economics. Experiments have been proven to be a useful tool for testing some economic theories but their effectiveness can be occasionally doubtful due to limits or flaws in experimental design. While some would argue that we should reconsider the approaches to policy evaluation in economics and adopt empirical methods such as in behavioural economics and experimental economics, some insist that the traditional approaches are still the best method when it comes to policy evaluation, and some kind of middle ground should be reached. 

First, the reasons why economists are interested in experimental economics and behavioral economics are quite obvious. Throughout the development of economics as a discipline, the use of mathematics and deduction usually plays an important role in constructing economic claims. Economics aims to abstract economic phenomena from the complexity of the world and model the behaviors of the economic agents using mathematical tools. During this process, economists make general assumptions of the population and ignore individual anomalies as they feel that according to the law of the large numbers, the aggregate pattern will not be affected by a few cases of anomalies. This method has allowed economists to produce many useful theories. 

However, the limitations are often also obvious. The need to make general assumptions may mean that the theories developed can be incapable of accounting for reality if the assumptions are wrong. Even though Milton Friedman argues that unrealistic assumptions can still produce good economic theories, it seems counter intuitive to claim that an unrealistic assumption can give us more confidence in producing good economic claims than an accurate one. This view is often taken by behavioural economists and experimental economists who want better and more accurate assumptions in economic modelling. 

In fact, the development of behavioral economics and experimental economics show that economists may want to seriously reconsider the methods of constructing economic claims. There are many cases where the traditional way of developing economic theories resulting in inaccurate explanation and prediction. For example, the standard model posits that when the price of an item falls, the quantity demanded will rise to the same extent that the quantity demanded will fall when the price goes up. However, what is observed is that when the price of an item falls, the quantity demanded increases more than the quantity demanded falls when the price goes up. By analysing the results of field experiments and laboratory experiments regarding this anomaly, behavioral economists, Daniel Kahneman and Amos Tversky, explain that this is due to “loss aversion”, which means that people have a stronger tendency to avoid losing than to acquire gains. Losing means paying more than what they used to pay; gaining means paying less than what they used to pay. From this case, we observe that the traditional method can sometimes fail to produce accurate theories, and behavioural economics has moved beyond the rationality assumption in order to construct a better account of this economic phenomenon. As the question states, “we need to know how the target populations are likely to respond”. When it comes to policy evaluation, it is important for the economists to consider the nature of the target audiences. The best way is to select representatives from the target populations and conduct experiments to find out their likely responses to the policy. It seems a better way as compared to the traditional approach which can be wrong or only when the conditions of the theories are satisfied in the real world. 

However, we should also recognize that controlled experiments in the new approaches can be flawed and thus produce inaccurate results. There is a reason why controlled experiments are not usually used in economics. In economic studies, the subject is people who are inconsistent, volatile and ever changing. Since the subject can be exposed to many factors that can easily change their reactions, experimental results can be wrong if the relevant factors are not controlled wisely. For example, in a dictator game where the experimenter researches on the tendency of the agents pursuing fairness in a transaction, the results are different when the experimenter is observing the actions of the volunteers. This example shows that careless experimental designs are likely to distort the experimental results. Even if the experiments are carefully planned and conducted, it may be unwise to generate the result from the volunteers to the target population because there can be self-selection into the experiments and the nature of the target population can be different from the representatives. Besides, it is also showed by research that people are likely to behave differently in experiments and real world decisions. Therefore, flaws in experiments can result in wrong predictions and finally the policy implemented may not produce expected results. 

Finally, we should recognize that both the traditional approach and the new method have their merits, and we should consider the nature of the target population before we consider which way to use in policy evaluation. When a policy is concerning a large population, the traditional way could be better since the theories produced in this way are usually more capable of accounting for behaviors of a large population. It is unwise to use controlled experiments because the population is too large and too risky for generalization of experimental results. However, new findings from behavioral economics and experimental economics can be parameterized into the standard model to better account for reality. For example, loss aversion is incorporated into many theories regarding financial markets to better explain and predict the market trend. Such finding can assist economists to give a more accurate prediction of the likely responses to the policy by the target population. On the other hand, when the target population is small in size, it can be advisable to use controlled experiments to test out the likely responses of the population as they are more likely to be accurate when the conditions of the theories are held in reality. The theory of inequality aversion is a good example to show that theories derived from controlled experiments can yield surprisingly accurate result when the nature of the population can be known. 

In conclusion, the use of controlled experiments can be beneficial in revealing the likely responses of the target population to a policy, but any flaws in experimental design can distort the experimental results. Besides, when a population is large, it can be risky to generalize experimental results to the whole population because the nature of the subjects can be different. On the other hand, the traditional approach also faces the problem of giving wrong judgment when the assumption of ceteris paribus does not hold. Besides, it can fail to give an accurate account of economic behavior when human psychology which can motivate economic actions are constantly ignored in theory construction.  Eventually, what method to choose depends on the size of the target population, the volatility of the economic environment and the availability of information about the target population, but the use of controlled experiments can be introduce beneficial refinement to the traditional approach. 

JC Economics Essays: Once again, as this is the H1/ H2 A level economics examination season, and there is also the upcoming H3 Economics examination, just like in the previous post, I have decided to include a few H3 economics essays. This particular H3 economics essay on economic methodology was contributed by WXN and written under examination conditions of around 1 hour. Special thanks to WXN for his H3 economics contribution to my economics essay site. 

Some areas of improvement for WXN's economics essay could be: More examples could have been given in this essay, and more of the real world context could have been linked to this economics essay. Often times, an essay could benefit from more contextualisation. Also, the division between positive and normative economics could also have been discussed in this essay. Normative economics often lends itself better to economic policy-making and evaluation. Statistics, econometrics, and economic methods such as cliometrics could also have been discussed. The difference between the nature of economics as a social science, and science and mathematics could also have been discussed in this essay. Is economics like a science? Can experimentation play the same role for economics that it has played for the natural and more hard sciences? There are many areas in which this economics essay could have benefited from a wider scope and range.  

However, once again WXN did also do many things right for his economics essay, making all the right moves that make an excellent essay strong. He talked about Milton Friedman, the famous economist, who once wrote a major paper on economic methodology. He also rightly talked about experimental economics and behavioural economics, which are reconsiderations of traditional economic policy-making and evaluation. Perhaps, a distinction between policy-making and economic evaluation could have been made. All in all, given the time constraints and economics examination pressure, this H3 economics essay is quite strong. 

How would you have approached this economics question, and what would you have done differently or better? Please remember to think of ways for further improvement and always question the hidden and implicit assumptions in the question. As this is a H3 economics essay, there will always be room for exploration and discussion, as well as a multitude of perspectives. Demonstrate your reading and your thinking. 

Thank you for reading and cheers!

[Postscript: WXN and his batch of economics students achieved 100% A/B for H2 Economics and 100% Distinction/Merit for H3 Economics in the 2013 economics examinations. They worked hard and performed well, so it is not all my credit. Congratulations!]