If you’ve been following the debates over the current federal budget*, you might have caught a little maneuver of the Republican members of the Senate: introducing another Balanced Budget constitutional amendment. The full text is available here. Bruce Bartlett calls it the “Dopiest Constitutional Amendment of All Time” and for good reason. The proposed amendment has two main forms of limitation on federal expenditures:

SECTION 1. Total outlays for any fiscal year shall not exceed total receipts for that fiscal year, unless two-thirds of the duly chosen and sworn Members of each House of Congress shall provide by law for a specific excess of outlays over receipts by a roll call vote.

SECTION 2. Total outlays for any fiscal year shall not exceed 18 percent of the gross domestic product of the United States for the calendar year ending before the beginning of such fiscal year, unless two-thirds of the duly chosen and sworn Members of each House of Congress shall provide by law for a specific amount in excess of such 18 percent by a roll call vote.

The first part is classic balanced budget – limit expenditures to revenues. It’s silly for a lot of reasons, mainly the fact that government ought to be really good at spending countercyclically, since a lot of things that make revenue go down also make expenditures go up, e.g. when unemployment rises, tax revenues fall and benefits paid out go up. These sorts of increases in expenditures and decreases in revenue when the economy suffers are known as “automatic stabilizers” and are generally believed to be a good thing – when most people are not spending money, the government spends more, without anyone having to see it coming and pass a special bill. Automatic stabilizers were much smaller or non-existent 100 years ago, and there are arguments that our economy is much more stable because of them. Beyond that, a lot of these programs have the flavor of insurance, and in general I’m a big fan of the government providing insurance for the poor (a role that is under direct assault, and has been for years, in what Hacker calls “the great risk shift“.) So, a balanced budget, if actually balanced by year-to-year revenue, would have to somehow stop this automatic stabilizer business.

That’s all old news. What’s fascinating about this proposal, to me and to Bartlett, is the attempt to limit government’s size by capping expenditures at 18% of GDP of the previous year. Bartlett’s take is interesting and coincides with mine, though he is much more shrill:

Right at the moment, we think GDP was $14,657.8 billion in 2010, so that means that outays in FY2012 could not be more than $2,638.4 billion. But GDP generally rises over time. The CBO expects that calendar year 2012 GDP will be $15,858 billion. Assuming that Congress was somehow or other able to meet its spending target of $2,638.4 billion, that would equal just 16.6 percent of calendar 2012 GDP. The same thing would happen year after year, forcing down spending as a share of GDP under the guise of balancing the budget.

I won’t repeat all of my previous criticisms of the balanced budget amendment that can be found in the links. But let me discuss one other thing. The gross domestic product is not a concept defined in law and is revised constantly; from time to time, the Bureau of Economic Analysis revises the GDP data all the way back to 1929. And of course, the 18 percent figure is totally arbitrary; the proposal effectively assumes that all federal outlays consist of funds that are appropriated annually, rather than consisting primarily of mandatory programs such as Social Security, Medicare and interest on the debt. Even if Congress was willing to cut mandatory spending, it is practically impossible to do so quickly unless it is willing to reduce the monthly checks going to current retirees and other actions difficult to contemplate.

In short, this is quite possibly the stupidest constitutional amendment I think I have ever seen. It looks like it was drafted by a couple of interns on the back of a napkin. Every senator cosponsoring this POS should be ashamed of themselves.

The key part for me is, of course, that the GDP is not a legal concept, but rather a constantly shimmering** data series generated by the Bureau of Economic Analysis. Near the end, the amendment suggests that: “The Congress shall have power to enforce and implement this article by appropriate legislation, which may rely on estimates of outlays, receipts, and gross domestic product.” That’s a nice addition, because all measures of gross domestic product are estimates. We have reason to think they are really good estimates, given the definitions we currently choose to employ and the data we have available, but they are still estimates. And there are thorny conceptual debates still only partially resolved at the heart of GDP calculations, about the nature of income and wealth (e.g. what counts as final production vs. an intermediate product), about the timing of production, about the relevant boundaries (which should be counted, e.g. unpaid housework), and about how to deal with environmental damages. Almost undoubtedly, economists and statisticians will push for changes to the current system, and that will probably be for the best. 90 years ago, we didn’t think much about the environment impacts of economic growth, now (especially with the looming climate crisis) we do, and hopefully someday our measures of development will factor those costs in. By keying something as huge as the Federal budget to these statistical estimates, the Republicans would ensure the overt politicization of these data as politicians suddenly became legally bound to a concept whose borders they could attempt to influence.

Of course, this amendment will never be ratified, so perhaps the details are not entirely important. What this proposal signifies to me, though, is that we take very much for granted that we have something called an economy which is measured by something called gross domestic product, and that both of those things are uncontroversial enough that you could imagine fixing important decisions to them. How that came to be is what I’m studying right now. I’ll let you know what I find.

* Hard to avoid here on my trip to DC, since tons of people will be out of work if there’s a shutdown, and since everyone is just a bit more into talking politics.
** Shimmering data is a concept borrowed from Paul Edwards, who uses it to refer to global climate data, which are constantly being updated as we incorporate new bits of information about past periods or make adjustments to our data models. See Edwards’ A Vast Machine for more. In this case, not only are the data shimmering, the definitions are quite open for debate – I believe this is considerably less true in the case of global temperature data, for example. So, some of the shimmering involves redefining the concept being measured, not just mapping from the historical record to the (stable) concept differently.