From Grossly Exagerrated to Too High

In our report on water impacts on San Joaquin Valley unemployment, I harshly criticized the estimates from UC-Davis calling their estimate of 35,000 lost jobs (which was revised down from earlier estimates as high as 90,000) "grossly exagerrated." I am happy that this criticism prompted Dr. Howitt at UC-Davis to take another look at their earlier work. The result is this new report that pegs lost jobs from water shortages at 21,000 of which they attribute 5,000 to environmental restrictions for endangered species and 16,000 to drought.

I still think this estimate is much too high, but it is a substantial improvement over the earlier numbers. Importantly, this new report acknowledges a major error in the earlier estimates. I give Dr. Howitt and his team a lot of credit for acknowledging that and doing something about it quickly.

Without getting too mired in econo-babble, let me try to briefly explain why this number is too high. If you apply the lost revenue (or fallowed acreage or other plausible drought impact) into a basic economic impact model, you will get an estimate of 10,000 to 12,000 lost jobs from water shortages (this includes multiplier or ripple effects). Dr. Howitt and I agree on this. There are many reasons why an economist could argue that this estimate is too high or too low and adjust the model or results.

The new UC-Davis study ignores or dismisses reasons to lower the estimate, and makes 4 adjustments to the model to increase the estimates up to 21,000. I believe the evidence and arguments for downward adjustments are much stronger, and that is why our report adjusted the lost job estimate down to 6,000.

I think the best way to evaluate these estimates is simply to look at the current jobs data and determine which is most plausible. When looking at this data, it is important to recognize seasonality. Both of our studies' lost jobs estimates are an annual average. Thus, my 6,000 estimate is more likely 4,000 in winter and 8,000 during the peak harvest season; and Dr. Howitt's 21,000 would be something like 15,000 in winter and 30,000 now.

There is widespread agreement that almost all the drought impacts on agriculture are concentrated in Fresno, Kings, and Kern counties. The latest official payroll data for August show these 3 counties have lost a total of 22,000 fewer jobs than a year ago across all sectors of the economy (farm jobs are down by about 1,000).

Thus, the implication of the latest UC-Davis estimates is that these 3 counties would be showing NO job loss (or possibly a gain) if full water deliveries were made. In the midst of the "great recession" and foreclosure crisis, do you think it is plausible that these counties would be virtually the only place in America to gain jobs this year if only they had full water supply?