A few days ago, the Census Bureau released data on the economic characteristics of small towns for the first time since the 2000 Census. The data come from the American Community Survey which has replaced the long-form in the Decennial Census, and reports a 5-year average value from a year-round survey conducted between 2005-2009.
It reports the Mendota unemployment rate at 16%, link here. This will no doubt come as a shock to people following California water issues who have been hearing about 40% unemployment in this town for the past two years, but it is not surprising to those of us who have been explaining that 40% is a bogus number, crudely extrapolated from 2000 Census results that measured Mendota unemployment at a stunning 32%. While Census is now reporting a much smaller unemployment rate, the reported income levels in Mendota remain miserable, under $9,000 per capita in the new release with over 50% of children living below the poverty line.
Why did the Census estimate of unemployment drop from 32% in 2000 to 16% for 2005-2009? It results from an improvement in methodology to a year-round sample, not an improvement in the Mendota economy. The Census' new ACS is a year-round survey, whereas the old decennial Census figures were a snapshot of conditions on April 1. April 1 is at the tail end of the high-unemployment season in Valley farming areas, just before conditions begin improving with the season. The EDD unemployment estimates people have been citing have been inflated because they were based on an April 1 unemployment proportion, and then applying this to data year round. It therefore exagerrates unemployment estimates in Mendota and all farming communities during the rest of the year.
As I have stated repeatedly, the only valid current unemployment rates are at the County level. The official, monthly County level unemployment estimates from EDD/BLS match very well with what the Census Bureau reported from 2005-2009. For example, Fresno County unemployment averaged 10.2% from 2005 to 2009, and the new Census 5-year average is 10%. However, California EDD estimates for small farm towns are wildly off due to the crude methodology. (EDD staff have told me privately that the crude technique was created because a local area unemployment estimate was needed to distribute federal funds, the local area estimates were never intended to be used for policy purposes, and the estimates come with a warning about the untested method.) The error is now apparant since California EDD has reported average Mendota unemployment from 2005 to 2009 at 29% with their crude formula during the same period the Census measures 16% with a real local survey. As discussed above, the EDD estimate is wrong because it applies a seasonally high unemployment proportion from April 1 to all months to make a crude guestimate of unemployment rates from County data. EDD will eventually revise their city level estimates to the new Census data, and their reported unemployment rate for Mendota will drop remarkably. They should be sure to revise it backward to 2005 so that there isn't any misinterpretation that the recent rains caused unemployment to drop.
This post should not be used to minimize the problems with unemployment and poverty in Mendota. It is a large and disturbing problem, but it is also a chronic problem not something created by environmentalists.