In Recessions, Rising US Productivity Numbers Do Not Represent Real Gains

US productivity statistics are an average over the whole economy. In recessionary times, many more firms with lower productivity are closing. These are the firms whose cost of producing a product or providing a service is more expensive than their competitors.

Even without any economic growth, just by the mathematics of averaging, removing the least productive firms will raise productivity. For example, suppose there are four firms that for each unit of input, output, 2, 3, 3, and 4 units. The average is 3 units per firm per unit input (12/4). If the least productive firm cannot survive and goes out of business, then for 3 units of input, there are 3, 3 and 4 units of output from the three remaining firms. That averages to 3.33 units of output per unit of input (10/3) for an 11 percent gain in average productivity. Also, in my example, change in output is (-2) over change in input (-1), for a 200 percent gain in marginal productivity.

This measure is non-sustainable. It is just the mathematics of weak firms closing. The firms that do not close and that remain in business are not becoming more productive. The US reported productivity number does not solely represent the marginal productivity of new production. As long as the least productive firms that are closing have lower productivity than the average of the firms that remain, whether or not the remaining firms increase productivity, the economy wide productivity numbers will show an increase. The change in the economy wide number will represent the additive effect of both, the closing firms and the remaining firms. It will be higher than the true measure of the change in productivity of the firms remaining.

The change in US productivity number is exceptionally high because it represents additive factors: the effect of low productivity firms closing, the effect of remaining firms that do not increase productivity but have a higher productivity than the closing firms, and the effect of firms that can increase the productivity of their output.

The result is that firms that are continuing to produce have a productivity change number that is lower than the economy wide statistic. It is 'juiced' by removing the low productive, closing firms from the average productivity.