Northern Trust’s Carl Tannenbaum recently penned an article in which he concludes US productivity has risen past couple of years:

Pushing for Productivity

This conclusion is based on GDP having increased by a greater amount than hours worked, so that the amount of GDP each hour worked produces has risen, the textbook definition of productivity.

These data nevertheless do not establish an increase in productivity.  The problem is in measuring the value of GDP.  If we measure it in terms of dollar value, measured GDP can increase even if real GDP doesn’t change at all, merely by virtue of a depreciating unit of account.  Since the period in question has notoriously seen high inflation, it’s possible that a shrinking dollar accounts for the apparent increase in GDP.

One way around this problem is to adjust GDP for inflation to find real GDP.  But what if our inflation data understate the full degree of dollar depreciation?  In that case, the adjustment will be insufficient and our “real” GDP will still be exaggerated.

What if our inflation data lag actual inflation?  Possibly changes in consumer prices lag inflation.  In this case productivity will appear to rise in the acceleration phase of inflation, because our consumer based inflation data aren’t keeping up.  Given the historical pattern seen in the charts, this certainly can’t be ruled out.  And in light of the known lag between recent stunning increases in the CPI and the less recent but even more stunning increases in the size of the Federal Reserve’s balance sheet and in government deficit spending, it is more than plausible.

If this is the case, my hypothesis predicts that productivity by such a flawed measurement will flag over the coming quarters.  Readers should check me up on this.

Or possibly the inflation data systematically understate inflation.  This also cannot be ruled out, given the exclusion of galloping home prices from official consumer price series and the liberal use of “hedonics” to reduce the amount by which actual price increases make it into the official results.

In order to show an increase in productivity, then, dollar depreciation must be excluded and adjustments for inflation must be shown to be accurate and complete.