Why Good Policy is Not Good Politics


What is the most important thing in politics that young people should know, but usually do not? I strongly believe that it the simple idea that was best formulated in Bruce Buena de Mesquita et al.’s 2003 book “Logic of Political Survival”: good policy is not necessarily good politics.


This is a spectacularly brief restatement of a very straightforward argument. Politicians are humans who by and large want to stay in power. Therefore they do whatever it takes for political survival. This implies that if there is a good policy (i.e. a strategy that improves social welfare) which is bad politics (i.e. reduces the changes to stay in power), politicians are not likely to choose it. This simple idea is key to modern political economy, explaining why (a) governments keep choosing obviously inefficient policies, and (b) what characteristics of political and economic institutions affect the chances that good policies are also good politics.


In my academic paper with Georgy Egorov (Kellogg) and Konstantin Sonin (New Economic School) “Why Resource-poor Dictators Allow Freer Media: A Theory and Evidence from Panel Data,” we consider one of the crucial examples of such “inefficient” behavior by the governments: media censorship. It is clear that censorship is very detrimental to society and the economy. If the government does not have any feedback on the performance of its own officials, it is unlikely to assess their performance and therefore will be unable to implement any policies. In democratic countries, it is well understood that the government cannot afford censorship – as well as it cannot afford electoral fraud, breaking rule of law, or expropriating ordinary citizens.


The situation is different in dictatorships where the rulers have a choice whether to allow at least partial media freedom or censor media completely. On the one hand, benevolent dictators are interested in media providing honest feedback on their bureaucrats. Indeed, if bureaucrats are not monitored, the dictator knows that this may result in growth of corruption and reduced living standards, which may eventually cost the dictator his job. On the other hand, media freedom presents a direct threat to political survival. Citizens may use it to make inferences about the abilities and performance of the ruler himself. Even worse (for the dictator), free media serve as a coordination device by making negative information, if any, common knowledge, which allows citizens to coordinate in revolting and replacing the ruler. It is therefore natural to expect that media will be less free whenever the ruler values job security over provision of incentives.


In our paper, we directly test this implication using the differences in natural resource abundance. Different countries (including different dictatorships) have different amounts of mineral resource wealth. Moreover, mineral reserves change over time. This allows us to conduct a very powerful empirical test. We check whether, in the very same country with the same history, culture, religion etc., an increase in resource abundance results in lower media freedom. Our theory implies that in dictatorships, this should be the case. Indeed, more resources imply more rents to expropriate; on the other hand, resource abundance implies less need for high-powered incentives to bureaucrats (resource rents allow making up for the inefficient bureaucracy).


This is exactly what we find. In non-democratic countries, an increase in oil reserves over time is correlated with lower media freedom (controlling for all other determinants of media freedom including those that change with time). Interestingly – and this is also in line with our theory – there is no such relationship in democratic countries.


Our econometric results are reinforced by many examples – the growth of oil price and oil reserves in 2000s was accompanied by substantial growth of censorship in Iran, Russia and Venezuela – but not in such large oil producers as Canada, Norway, or the U.S. The inefficient state-owned oil producer in Mexico, PEMEX, has depleted Mexico’s oil reserves – and this happened alongside growing media freedom. When the Soviet Union discovered West Siberian oil, it shelved its reform plans and entered twenty years of 1970-80s stagnation. And vice versa, once the drop of oil prices by a factor of three destroyed the value of Soviet mineral wealth, Gorbachev decided to introduce (initially partial) media freedom. As the famous New York Times journalist Tom Friedman writes in his “First Law in Petropolitics,” “Give me $18 per barrel, and I will give you democracy in …”. Our paper certainly proves him right, at least with respect to media freedom if not democracy.


Photo Credit: Wikimedia Commons