We Need More Nuance from the CBO

Feb 20, 2014Jeff Madrick

The CBO's insistence on presenting just a single number makes its predictions misleading, and sometimes even useless.

I have long thought we need two Congressional Budget Offices, the current one and a real one. The problem with the current “bi-partisan” CBO is again apparent in the wake of its claims about the impact of the president’s proposed increase of the federal minimum wage to $10.10.

But let’s take a step back. The CBO has long had little sensitivity to the impact of how it presents its invariably uncertain findings. To the contrary, it seems to respond to a demand from Congress and the general public for a concrete number, essential ambiguities aside. It has long presented one-number conclusions – the size of the budget deficit, the rate of economic growth, the impact on jobs of the minimum wage – as if people are educated enough in the uncertainties of economics that they won’t take the numbers too seriously. 

But, in fact, they do. The CBO economists disingenuously cover themselves by burying the extensive qualifications of the data and research, as well as alternative possibilities, in dense appendices and footnotes. In the process, they feed politicians and pundits with projections the poor naïfs think are written in stone – or are, most likely, perfect for use as political cover. Is this unwitting, or does the CBO enhance its influence with these easy-to-digest but misleading pronouncements?

But the CBO is worse than merely insensitive to its public relations impact. Their economists typically treat economic hypotheses as entirely true. A couple of centuries ago, John Stuart Mill pointed out that economics is hypothetical. CBO economists should go back and read this. They base forecasts on the imbedded idea that weak economies will automatically self-adjust within three or four years, for example. This is a neoclassical assumption, not a fact of life. They acknowledge that higher deficits can stimulate a weak economy but within a few years it will undercut growth. Evidence is highly ambiguous on this point. They have little compunction about making thirty-year forecasts, no less ten-year forecasts. No one knows what will happen in ten years. 

The bi-partisan label has become comical. The CBO does not necessarily lean Republican or Democrat, but it is not truly objective. Economics does not allow that.

A real CBO would present a range of projections and forecasts and make a priority of demonstrating how tentative most of its conclusions must be. The current CBO sometimes presents a secondary forecast if certain expected changes in the laws are actually made, but it should publish a range of likely outcomes with or without legal changes.

I guess I am merely stating the obvious, but the obvious has to be addressed at some point. The minimum wage debate is another clear example. Read the appendices and footnotes and you see the CBO took its job seriously; its economists read a lot of research. For those families below the poverty line, income would rise by $5 billion; for those from one to three times the poverty line, incomes would rise by $12 billion. These are mere estimates, let’s keep in mind. Nearly one million would be lifted out of poverty. Those who earn some seven times the poverty line or better would see their incomes reduced because business profits might come down due to higher wages and prices may go up for the products they buy, If true, these estimates suggest a pretty nice redistribution of the national income to lower-income families, not bad in a time of intense inequality.

But the headlines were created by the CBO’s claim that 500,000 jobs will be lost. Here’s the actual sentence: “Once fully implemented in the second half of 2016, the $10.10 option would reduce total employment by about 500,000 workers, or 0.3 percent, CBO projects.” The sentence that follows has the tepid disclaimer that there could be a wide range of job losses. But then why make the declarative sentence above? It’s the one, of course, that anti-minimum wage politicians focused on, as did much of the media. But the 500,000 number is not a forecast, it is simply a midpoint on this wide distribution from essentially zero jobs lost to one million. Oh, yes, the CBO eventually says that, but as a writer myself, I have to ask why the CBO doesn’t present the uncertainty immediately.

The CBO, as is now widely reported, did no original research. It looked at existing studies. A recent one in particular, which showed substantial losses, used a highly dubious methodology. The study showed the biggest future job losses were in manufacturing, which has relatively few minimum-wage jobs. By contrast, it showed few prospective losses in retail and similar industries, which have many such minimum wage jobs. This is highly implausible. Economic critics concluded the methodology was flawed. Why did the CBO pay attention to it?

It’s high time to rethink the purpose and practical capabilities of the CBO. It should be forthright about the ambiguities of economic science, it should avoid single-note forecasts, and it should make sure policymakers understand the risks and sensitivities of what they are doing. In sum, it should not produce simple answers to complex questions.

Does any of this really need saying? Judging by the minimum wage brouhahah, it does. Ideally, we need a “shadow” CBO to challenge its findings and explain their many assumptions on a regular basis. That would be costly, I fear. But some review of the purpose of the CBO and what they emphasize would be very useful. 

Jeff Madrick is a Roosevelt Institute Senior Fellow and Director of the Bernard L. Schwartz Rediscovering Government Initiative.

Share This