Thursday, May 7, 2009

Productivity Residuals in this and Previous Recessions

Productivity -- output per hour worked -- has risen almost every quarter of this recession. That doesn't always happen in a recession.

Nevertheless, if there were a reduction in hours worked without any fundamental change in productivity, then we would expect output per hour to rise because of diminishing marginal product of labor. Thus, it is helpful to know what productivity would be if hours had not changed -- that's what I call the productivity residual (conceptually similar to the Solow residual used in macroeconomics).

The chart below (an update of Figure 5 in my NBER wp) shows the productivity residual in this recession and four previous ones. The productivity residual has fallen in the last two quarters -- that is productivity has not risen as much as you might expected given the drop in hours -- and remains higher than it was when the recession began. The two worst quarters of this recession are not near as bad (in terms of productivity residuals) as the two worst quarters of 1981-82.

Time will tell whether the productivity residual rises or falls 2009 Q1 - Q2.

1 comment:

Boonton said...

Looking at your graph I notice that many of the recessions do show an increase in productivity but they also show an initial dip before that increase happened.

2001 is the exception but that was a kind of strange recession. Before it you had so much nonsensical 'dot-com' companies and earnings fakery that maybe when the crash happened people started paying closer attention to producing real results rather than spouting rationalizations like 'profits don't matter'.

Another issue, productivity is output divided by hours worked. OK so if layoffs are happening firms will fire the least productive first resulting in a productivity boost. But output is only going to be produced if it can be sold. Consider the 'dead cat bounce' story....

Company A makes 100 widgets a month and keeps 20 in inventory at all times.

In December a recession hits and they only sell 80 units. Now they have 40 units in inventory.

January: They cut back dramatically, producing 50 units in the next month. They sell 80 again reducing their inventory to only 10 units.

February: Now, because they desire to have 20 in inventory at all times, they produce 90. If sales are 80 again they will have 20 units in inventory. Because they grew from 50 units to 90 units pundits start to wonder if the 'recession' is finally over. It isn't, the firm is just adjusting to a lower demand.

Between Dec. and Jan., the firm might have laid off a lot of the less productive workers. But they also cut back production so productivity might actually have fallen since those productive workers were told not to be so productive. In Feb., though, the firm took advantage of those productive workers and pushed them hard. As a result you see the steep increase in productivity.