State Coincident Indexes: The Sum of the Parts Greater than the Whole?

Last Wednesday the Philadelphia Federal Reserve Bank issued its estimates of state coincident activity for July. All 50 states report increases over the last 12 months; only West Virginia and Alaska showed gains under 1 percent. As has been the case for numbers of regional measures, the West (with the notable exception of Alaska) appears to be growing more rapidly than the rest of the nation, with 4 of the 6 states reporting increases of 4 percent or higher located there (California’s 4.1 percent increase is clearly to most important in this regard). The lower Mississippi Valley states (Arkansas, Mississippi, and Louisiana) appears to be the one region in which growth was substantively softer than the nation.

That said, there seems to be some anomalies between the state and national figures. The national index, produced by the same methodology as the state measures, was up 2.9 percent in the year ending in July. However, 31 states had gains of 3 percent or higher! New York was the one very large state with an increase under the national pace, but the Empire State’s gain of 2.7 percent was not qualitatively different from the nation’s. Weighting the increases of all 50 states by their 2017 real GDPs produced an increase of 3.4 percent, considerably higher than the 2.9 percent national increase! Why did this happen? It’s not altogether clear, but part of the reason probably reflects a divergence between national and state job numbers. Payroll employment growth is an important component of the coincident indexes, and the sum of states has shown job growth higher than the nation. Seasonally adjusted, the total number of jobs in the states increased 1.64 percent in the year ending in July vs. the official national gain of 1.81 percent. Not seasonally adjusted, the divergence was even larger: 2.05 percent was the sum of the states’ figures, compared to 1.65 percent for the nation as a whole. Unfortunately, until the coincident indexes are recomputed next year with benchmarked GDP (the indexes are designed to have trends comparable to real GDP; the national figure hasn’t been recalibrated to reflect the new GDP data, and the relevant state GDP numbers won’t be available until November), and the annual benchmarking of job data, we won’t know whether the national figure or the weighted sum of the state measure is “right.”

Turning back to the reported index numbers, there was substantial divergence reported for recent moves. 23 states are reported to have had strong gains of 1 percent or more over the three months ending in July, but 6 had growth of less than .1 percent, with 2 (Maine and Alabama) showing outright declines. On the whole, while the coincident indexes clearly show state growing, the increases are not uniform, and squaring the amount of growth with the figures reported for the nation seems a bit problematic.