Goldman-Sachs’ Old Keynesian report on budget: garbage in, garbage out

Washington is starting to absorb Goldman-Sachs latest missive on the projected economic effects of the budget cuts proposed by House Republicans. At first glance, it’s not good (ABC News):

A confidential new report prepared by Goldman Sachs for its clients says spending cuts passed by the House of Representatives last week would be a drag on the economy, cutting economic growth by about two percent of GDP. “Under the House passed spending bill [which cut spending by $61 billion],” says the report, which was obtained by ABC News, “the drag on GDP growth from federal fiscal policy would increase by 1.5pp to 2pp in Q2 and Q3 compared with current law.”

Naturally, the Democrats are all touting it, while Republican Speaker John Boehner’s office is dismissing it (same link). I decided to do a little digging into the report first. Mainly, I was looking for G-S’s multiplier.

For those who are not aware: the multiplier is what economist use to measure the effect of government spending on the economy. As I described here, a multiplier of 3, for example, would turn $1 of government spending into $3 of economic growth (or, as more appropriate here, a cut of $1 in government spending cuts growth by $3).

For “Old Keynesians,” the multiplier is a straightforward and simple way of measuring how government spending improves the economy. For the rest of us in the field, it’s not so simple. Old Keynesians tend to ignore the effect of government spending on available funds for the private sector entreprenuers and business. Financial investors can choose between private companies and U.S. Treasury notes; the more the government borrows, the less these folks will invest in the private sector. This “crowding out effect” (i.e., private business are crowded out of the funds available from investors) can reduce the impact of government spending (as shown in the multiplier).

Recent analyses, in fact, have shown a dramatic crowding out effect: the International Monetary Fund determined the multiplier to be 0.7 (John Taylor); the European Central Bank had an even lower figure: 0.5. Both of these mean that the “multiplier” is in fact a divider (i.e., $1 of government spending leads to only 50 or 70 cents in growth). Harvard economist Robert Barro actually find a multiplier of zero (Taylor).

So, what does Alec Phillips (the G-S economist) use as his multiplier? He doesn’t come out and say it, but this paragraph lets us pull it out (ABC again):

(Shutting the government down for the whole month of March) would equate to $32bn in annualized terms, or around 0.2% of GDP for each week of shutdown. Pulling this spending out of Q2 would reduce the contribution to quarterly GDP growth from federal activity by a little over 0.8pp (RWL note: short for percentage points) at an annualized rate for each week the shutdown lasted . . .

So a 0.2% of GDP cut in government spending would lead to of reduction in growth by 0.8 percentage points – that’s a massive multiplier of 4. Even some Old Keynesians don’t use multipliers that high; New Keynesians and non-Keynesians would consider it laughable. More to the point, it is completely outside the trend of recent analysis.

It reminds me of something I had drilled into my head in my undergraduate and graduate students days from economics professors: “garbage in, garbage out;” i.e., bad data, bad variables, or a bad model lead to bad analysis no matter how much you dress it up.

In the end, that’s what Goldman-Sachs did here. They plugged numbers into a terrible model and came up with an “analysis” that isn’t worth the bandwith on which it was carried.

Or, as my old professors would say: garbage in, garbage out.

Cross-posted to RWL

Advertisement