Offsetting changes leave the forecast unchanged
Look for a big decline in Q1 and a bounceback in Q2
Data over the past week left the Nowcast GDP forecast unchanged. Consumer spending is up slightly in the forecast, as the CPI came in a little lower in February than the model had predicted. However, PPI came in a bit higher, pushing the inventory forecast down a bit, and offsetting the improvement in consumer spending.
I should stress that the inventory number still looks unrealistic given the surge in imports. It’s quite likely that inventories will be much higher than the model forecasts, which would reduce the decline in GDP. The time series approach may be misleading in this particular case, since it’s unlikely that all those imports went immediately to final demand. A positive print for Q1 GDP is still not completely out of the question!
For those interested in economic momentum: the real story is in consumer spending (which is slowing, but still positive) and business investment (which may well show a jump in Q1 as equipment investment comes roaring back).
A word about business investment: a lot of commentary is assuming that business investment will slow because of policy uncertainty. That’s certainly true. However, Q1 is likely to show strong invesment spending because of the monthly numbers already in place. Shipments of nondefense capital goods grew over 3% in December and January. The simple arithmetic of this indicator suggests a strong Q1 even if shipments were weak in February and March. Also, some of that import surge was captial goods, (up over 5% in January) which will count as investment if they are put into operation. To be clear: policy uncertainty is a problem, and we may begin to see it in the data in coming months. But Q1 GDP is going to show a big jump in investment spending.
The Fed forecasts remain split. The Atlanta Fed’s GDPNow, based on components (like The Nowcast) is showing a big decline. The New York Fed and St. Louis Fed forecasts, based on indicators are showing reasonably strong growth. I suspect that the inventory story is the difference, and that the forecasts will become closer once we know more about inventories.
A further note about Government spending in GDP
Paul Kupic at the American Enterprise Institute published a note that explains some of the issues with measuring government in GDP. It’s worth reading if you are interested in a more complete explanation of sone of the problems of measuring GDP than I provided last week.
It’s also very misleading in one important sense. Kupic explains that the marginal value of government services is unlikely to be the same as the marginal value of nongovernment services, which makes comparing them impossible. That’s correct. But he then deftly inserts the assumption that the marginal value of government services must necessarily be less than the marginal value of other goods and services.
There is absolutely no reason to think that this is the case! Government produced services include National Defense, education, and basic research, all of which are likelly to have very high marginal values. It’s just as likely that the marginal value of government services is greater than the marginal value of private goods and services, and that redirecting resources towards government would increase welfare. The slight of hand involved here is revealing, and disappointing in an otherwise accurate discussion of some of the problems of estimating GDP.
