A 12-Step Program for Fuel Subsidy-aholics

(Photo: Brad.K)

(Photo: Brad.K)

Back when blog posts were composed with reed styluses on clay tablets, I put up a couple of posts (here and here) on fuel subsidies in the developing world. These are generally 1) fiscally ruinous; 2) terrible for the environment and traffic congestion;  3) highly regressive with regard to wealth distribution; and 4) market-distorting by artificially promoting fuel-guzzling industries. So I made the case that this is a pretty foolish public policy, in fact one of the worst I can think of. It’s up there with tobacco subsidies, the Concorde, pretty much everything the North Korean government has ever done, and our government’s failure in spending a paltry $615,000 taxpayer dollars for UC Santa Cruz students to digitize priceless Grateful Dead photographs, t-shirts and concert tickets.

Given the problems with fuel subsidies, I promised a third post on what to do to eliminate them. But since I have a day job, and being a professor is much more difficult than it looked when I was undergrad, I’ve procrastinated on putting this last post up. However, engineering student Kishore from India wrote asking where part three is, and customer satisfaction is a goal here at Freakonomics. Besides, no doubt governments around the world have been waiting impatiently for my post before they start dismantling their fuel subsidies, so here it is.

Given the damning case against fuel subsidies, and a rising swell of opinion that they are counterproductive on many levels, why don’t these policies go away? The IMF (see this) and I offer several reasons:

  • First is simple confusion. In many countries, particularly where the subsidies are off-budget, few people are exactly sure what the subsidies cost and who pays for them, let alone their broader effects on the economy and society.
  • Most governments with fuel subsidies do not have stellar reputations with their populaces, and thus the political body believes that any savings from removing the subsidies would disappear into a sinkhole of inefficiency and Cayman Islands bank accounts.
  • While most of the largesse from fuel subsidies flows disproportionately to the rich, the subsidies’ termination would also mean hardship for many poor people.
  • Higher and more variable fuel prices might stoke inflation as the effects ripple through the economy. 
  • They may harm the international competitiveness of domestic industries which rely on cheap oil.
  • There is the concentrated benefits/dispersed costs phenomenon, which is the case with most wasteful spending in the developed world as well as the developing. Here, a relatively small group of organized and politically connected people (in this case, wealthy drivers and/or the energy sector), who individually reap large benefits from the status quo, are far more likely to be activists and may overcome the interests of the larger populace for whom the potential gains, though in the aggregate much larger, are more abstract, uncertain, and smaller-scale on an individual basis, and for whom benefits would only materialize in the future.
  • Fuel subsidies tend to be highest in oil-producing nations. One recent inventory of gas prices worldwide showed Saudi Arabia at 45 cents a gallon and Venezuela at an astonishing 4 cents a gallon. This is good PR for these nations as it helps show the masses that they are sharing in the nation’s oil wealth, keeping people’s minds off the fact that much of the oil revenue disappears into the maw of the kleptocracy in power.
  • Being a kleptocrat is a pretty good job offering a competitive compensation package, generous benefits, considerable autonomy, challenging and rewarding work, high social status, and the chance to meet lots of interesting and diverse people before you rob them. Given this, and the fact that few of the leaders involved would prefer to spend their golden years in exile in London or Switzerland, what is to be done given that eliminating fuel subsidies may spark mass unrest?

Here are some ideas: 

1) Before fixing the problem, it would help to know what the problem is. This means a rigorous study of the current policy in all its dimensions. This should start with the direct costs, which is harder to know than it sounds given that in many cases the subsidies do not appear on government balance sheets and instead are concealed in places like the balance sheets of state-owned industries.

2) Moving further, all of the subsidies’ short- and long-term impacts on government finances, national macroeconomics, the environment, traffic congestion, and wealth distribution must be tallied up, and contrasted with projections of the potential effects of policy reform.

3) This information must be disseminated to the public transparently, making it clear to the people who is gaining and losing from the subsidy program, particularly in terms of income distribution.

4) Once the public at large understands the issues, there must be broad consultation with all affected stakeholders, not just the politically connected and highly motivated beneficiaries, about the costs of subsidies and the potential benefits of subsidies’ repeal.

5) One strategy to bring these issues home to the public might be taking the gains and losses out of the realm of abstract numbers on a piece of paper. Instead, governments and civil society could demonstrate how the size of the subsidies compare to other budgetary needs such as health and education, showing the good that the money saved could do if applied to other purposes.

6) This trade-off could be made an explicit part of the policy of eliminating subsidies, and the plan may well include measures that offset the immediate pain to losing groups. The most straightforward and perhaps effective measure might be simple means-tested direct cash transfers; an alternate strategy might be government spending on things such as health, nutrition, or education. The problem here is that many of these programs may be difficult to administer or even beyond the capacity of the governments altogether, particularly in less-developed nations. One alternative might be simply upping expenditure on easy-to-run pro-poor programs already in place, such as bread subsidies. These are also market distorting, but at least they are progressive.

7) Taking all of this into account, a set of clear long-term objectives must be formulated, and a reasonable, meaningful, comprehensive plan produced. This must tackle the problem in all its dimensions.

8) Repeal must have a hard timetable for implementation.

9) As part of this process, government should take on broader reform of the petroleum sector. This might include examining where state-owned oil companies wield disproportionate power, stifling innovation and reducing economic efficiency. It would also involve ensuring that the private fuel market functions efficiently. For example, it is crucial to limit the market power of any participant, promoting adequate competition. One step toward this might be cutting red tape to ease entry to and exit from the market.

10) Along those lines, government should also formulate ideas to improve efficiency in fuel-guzzling industries, par
ticularly state-owned ones, to cushion the blow when fuel prices rise.

11) Assuming a consensus for repeal can be achieved, the program must be designed to ensure a smooth and effective transition, with a minimum of public outcry and economic dislocation, like spikes in inflation. One step would be to reduce subsidies only gradually, for example by limiting the amount fuel prices can rise in a given time period. This would reduce sticker shock.

12) Government might deflect some potential opposition by depoliticizing the process of subsidy reduction. This could be done through the use of automatic pricing measures like formulas, or politically independent price-setting bodies. 

All of these steps are easier said than done. But the good news is that the wind is now blowing against fuel subsidies, and courageous countries like the Philippines and Turkey have successfully addressed this issue, reducing fuel subsidies or eliminating them altogether. The bad news is that many have not, and even more depressingly other countries like Ghana, Sierra Leone and Togo have tried and failed, reinstating or even increasing fuel subsidies when oil prices rose dramatically — causing public outcry. Thanks to the strategies above, more countries can take the former path rather than the latter, hastening the day this policy runs out of gas.


If you include their pot budget, $615K to run stuff through scanners and take digital photos of T-shirts really isn't unreasonable.

The Concorde was pretty cool, even if mind-bogglingly wasteful.

Enter your name...

As a first step, simply freezing the subsidy amount and allowing prices to creep up slowly over time would give people the idea that they are not guaranteed a fixed price that will never change from one year to the next.


Agreed. Or why not first alter the model to be a subsidy at a certain percentage of the market price, and choose it so that the price is unchanged at first. Let it be for two years to get people used to small market-based fluctuations. Then decrease the subsidy percentage smoothly to zero percent over the following 10 years or so. Motivate the subsidy cut each year with some nominal (they need not be real) increases of benefits or transfers in other areas.

I don't believe in making the public see the light in matters like this. Just boil the frog slowly.


Wait, what's wrong with the Concorde?


There is one more factor that helps keep subsidies in place. If you take them away, prices go up pretty much immediately. Even if you can show the voters that this price increase may result in increased government revenues that will be spent on better health care and education just as soon as the kleptocrats reach their personal savings targets, those benefits are still iffy. Many of these people paying more for gas and things that require gas (like taxis, generators, stuff transported long distances) aren't too sure they'll need the healthcare and may not value education to pay more now for benefits later. Populations tend to riot when gas prices go up, rather than send thank-you notes to their policy makers. This is unfortunate - the subsidies should go, but most countries just can't get there.