CEA Weighs in on Government Shutdown

Private firms have given their estimates on the effects of the government shutdown on GDP. Macroeconomic Advisers says 0.2 percentage point. JP Morgan says 0.2 percentage point (no link). S&P says 0.6 percentage point. Goldman Sachs says 0.5 percentage point. Now, the President's Council of Economic Advisers is having their say and their whack at the pinata yielded a 0.25 percentage point loss in fourth quarter GDP (at an annualized rate, of course). This is more in line with the estimates of Macroeconomic Advisers and Goldman Sachs. However, the CEA goes one further than the private estimates and tacks a jobs number to their estimate: 120,000 less jobs created in the private sector for the first two weeks of October than otherwise would have been. Average the estimates out and you get a nice round 0.35 percentage point of lost annualized economic output in the fourth quarter.

But Jason Furman and the rest of the CEA are displeased with the private sector estimates (okay, they didn't actually say that). The CEA says that many of the private estimates of the shutdown's impact only took into account the direct loss of government spending and didn't take into account the broader effects of the economy - something that the Council of Economic Advisers did. How, you ask?

Well, every week the CEA releases their Weekly Economic Index. The WEI is created by the CEA by taking eight economic indicators and indexes, and performing what's called principal component analysis on them. This analysis captures a signal from the data that indicates which direction the economy is moving. The basic theory behind this is that the economy's movements have a predictable effect on certain indicators and that - since we can't directly observe the economy every week - we can estimate the weekly state of the economy from our observations of the indicators. After analyzing the indicators, the CEA normalizes that signal into one easy-to-digest number that corresponds with the annualized GDP growth rate. For example, if you have a WEI of 3.2, then if the economy grew at that week's pace for a whole quarter then you'd have a GDP growth rate of 3.2% for that quarter.   

So which 8 indicators does CEA use? The CEA follows two consumer spending indicators: ICSC's Same-Store Retail Sales and Redbook's Same-Store Retail Sales; two consumer confidence indicators: Gallup's Economic Confidence poll and Rasmussen's Consumer Index; two labor market indicators: Gallup's Job Creation Index and the Department of Labor's Initial Claims numbers; one industrial production indicator: the American Iron and Steel Institute's Raw Steel Production numbers; and one housing market indicator: the Mortgage Bankers Association's Mortgage Purchase Applications numbers.

Now, of course, these data are noisy. And the CEA warns against putting too much stock into one Weekly Economic Index number. All the WEI can do is give you an estimate of the state of the economy that week. One week may be good because it's the week of a holiday. Another week may be bad because the Heat won the championship and everybody is in a foul mood. But, when a large event occurs in the economy (a government shutdown in tandem with the threat of default, say), then you can make a reasonable assumption that the change in growth rates between two weeks is largely due to that event. And the CEA did just that. They saw that the week before the shutdown the economy was growing at a 3.6% clip. At the end of the first half of October, they saw that the economy was only growing at 2.0% clip - a loss of 1.6 percentage points of growth. Since the numbers are reported in a format of quarterly annualized GDP growth, they multiplied the weekly number by 2 (the shutdown lasted two weeks) and divided it by 13 (the number of weeks in the quarter). And that's how they priced the GOP's derp at 0.25 points of GDP growth.