« Puzzled | Main | Income Inequality, International Payments Imbalances, and Crises »

Tuesday, July 27, 2010

"Government as Deux Ex Machina"?

Dave Henderson has responded to my post earlier today (responding to a post of his), and I probably did overreact to the title of his initial post -- the title I chose for my post was clearly motivated by his title choice. The title was based upon what seemed to me to be a mischaracterization of my views, and that was a large part of what prompted my response (so in his description below, if annoyed=hurt feelings, he has it right). Also, the feeling that he was trying to paint me into an ideological corner led me to return the favor, but I was overly stark in my characterization.

So I appreciate this:

Government as Deux Ex Machina, by David Henderson: The above title should have been the title of my previous post. The title I gave it, "Mark Thoma Doesn't Get It," was unnecessarily provocative, as one of my co-bloggers has pointed out. I know the myth of male power, and part of it is that we men are not supposed to have feelings. We do have feelings. I think, based on his reaction in a subsequent post, that I hurt Mark Thoma's feelings. I didn't mean to. It was thoughtless of me to think that with that title, I wouldn't upset him. What I really meant to do is talk about the following:
When people advocate government intervention, they rarely, maybe never, tell us how the incentives will be set up so that government will do the right thing. Think about how asymmetric the argument is. Incentives in the private sector are such that someone will do something in his interest that hurts others in society, but he doesn't take account of that hurt in his decision. Or, someone could take action that would benefit others a great deal but it isn't in his interest to take the action. Notice the use of reasoning about incentives to show why the market fails. Therefore, continues the argument, we should have government intervene.
Did you catch the non sequitur? The argument proceeds at first using standard economic tools. We show that the incentives are such that the private actors make the decision that leads to sub-optimal results. Then we (not really we, but many of us) conclude that government should step in. But there's no analysis of government incentives. Why would government do the right thing? That's the unjoined debate. The late George Stigler once said it's like a judge at a beauty contest seeing just the first contestant and then awarding the prize to the second contestant.
I wasn't naive enough, as Mark Thoma suggests in his response, to think that he has "unqualified support for regulation." Also, he has been critical of specific regulations. One of the things that makes his blog interesting is his eclectism. I was just saying that although I read his blog a fair bit, I've never seen him, when he advocates a regulation or a government program, explain why he thinks the incentives will be set up so that it works. I had hoped to get him to address this. It would still be nice if he would. ...

I don't have time to deal with this directly at the moment by writing something new, but the implicit suggestion that I have ignored these issues is also part of what prompted my initial reaction. Here's one example that comes to mind showing that the suggestion that these issues are "never" discussed doesn't hold up. This relates to his call to discuss how the incentives associated with a government program, and how the incentives will be set up so that it works:

To look at the economics of this proposal, I decided to examine a fairly standard textbook treatment of the topic where a tax on each gallon of gas consumed is imposed along with a lump-sum tax rebate to consumers on an equal per capita basis. (I hope the microeconomists won't mind a macro guy stumbling around in their territory. This proposal is discussed in Pindyck and Rubinfeld 5th ed., pgs. 114-115.)
Here's a graph of what happens before and after the tax, and after both the tax and the rebate.
Click on figure to enlarge
The consumer starts out at point A consuming QA gallons of gasoline and has a utility level of U2. After the tax, which rotates the budget line downward as shown by the dashed budget constraint, the consumer moves to point B which is on a lower indifference curve U1, and consumption falls to QB. Finally, after the rebate which shifts the budget line outward, the consumer moves to point C and consumption increases to QC (the tangent indifference curve at point C is omitted for clarity).
Overall, the consumption of gasoline has fallen, as intended, and the consumer is worse off because the level of utility attainable at point C is below the level U2 at point A. Even though the money comes back to consumers in the form of a rebate, the reason consumption falls from A to C is because the income elasticity of demand for gasoline is relatively low (around .3 by some estimates) so that the substitution effect dominates the income effect.
In this example, a low-income household would be made worse off by the tax and rebate proposal (because indifference curve U2 is no longer attainable), but it's still possible for some low-income individuals, those who consume less gas than the value of the lump-sum rebate, to benefit. However, the substitution effect induced by the tax makes the average household worse off. To aid low income individuals, other proposals such as linking the size of the rebate to income could be examined as well.
Finally, this highlights the costs to households, but there are also potential benefits. To assess the proposal, the costs must be compared to the benefits from reduced dependence on foreign oil and the additional security that brings about, and the environmental and other benefits from lower consumption of gasoline.

That sure seems like a "textbook" analysis of the incentives and outcomes surrounding a potential government policy to me, including why it works to reduce consumption. And, though that was the first example to come to mind and it's a bit dated at this point, there's plenty more where that came from (e.g., today's post called "Puzzled" brings up the topic of market-based regulation which gets at the heart of the incentive problem -- I've explained in the past why I favor a market-based regulatory approach -- and this version of the original post from Schmalensee and Robert Stavins has an extensive set of links to research on the topic). Should I ask to see his blog entries going into this much detail when he objects to government programs? Nah, I won't do that. But I will keep the suggestion to discuss the incentives in mind as I write about these issues in the future. So thanks for that.

Update: A comment suggests:

mark the knock is on regulators motives and incentives
not on the impact on citizens

I thought the challenge included government policies as well, and since this is getting tiresome -- I didn't start all this and I'd like to do something else today -- I thought I'd stop there. But ok, I'll briefly address this. One of the points of market-based regulation mentioned above is to help with the regulator incentive problem, and I've talked about the incentives of regulators many times in the past. This may not be the best example, but a very quick search brings up:

regulators of these markets were captured by powerful forces that wanted the game to continue. The power of regulators, and the will to enforce the regulations, must match - in fact exceed - the will and power of those being regulated to resist having constraints placed on their behavior. I've talked about why ideology may have eroded the will of regulators, but their will is partly a function of their power. So long as we allow huge, clearly over-sized financial institutions to exist, this problem will potentially be present. Therefore, if the current anti-trust legislation is adequate to the task, then yes, let's give regulators the power to enforce it, and ensure we have people in place with the will to do so. ...

It's perhaps not expressed as clearly as I'd like, and may not talk enough about how to solve the incentive problem, but the term "will" is meant to refer to the incentives regulators face along with their underlying regulatory philosophy. And it does suggest that breaking up politically powerful banks under existing or, if necessary, new legislation will help to prevent regulatory capture and improve incentives (by removing bad ones). So a solution to the incentive problem isn't altogether ignored either. Anyway, that's enough time on this. Once again, even though I'm convinced I've addressed these issues in the past, I will keep this in mind as these issues come up in the future.

    Posted by on Tuesday, July 27, 2010 at 04:50 PM in Economics, Market Failure, Regulation | Permalink  Comments (26)


    Feed You can follow this conversation by subscribing to the comment feed for this post.