Can you elaborate?

The Central Valley (Sacramento, Stockton, Fresno, Visalia, Bakersfield) has been one of the hardest hit areas in terms of the real estate bubble. Home prices have dropped 50% in some areas. Property taxes are based on property values, and aren't going to be that stable.

One extreme fluctuation in property values doesn't disprove the general historical observation that property tax revenue is more stable than sales or income tax revenue. There are factors that can explain this, such as the fact that sales tax is essentially assessed in real time, income tax every year, and property tax less frequently. (My understanding is that it was every few years pre-Prop 13, and now only when a home is built, sold, or substantially rebuilt.) This lag in assessment will tend to smooth out revenues.

Also due to the way assessment is triggered, a 50% drop in home prices doesn't imply a 50% drop in total property tax revenue -- only for homes that are sold. [EDIT: But apparently one can appeal to have property reassessed without selling, which presumably many people will do with prices dropping significantly.] Furthermore, if sales forced by foreclosure are bias toward the lower price range -- as is the case -- that will further limit the effect on total property tax revenue.

EDIT: See also http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2009/01/25/... -- a little short on hard data, but suggesting that the variation in property tax revenue is not going to be anywhere close to the 40-50% drop in sales and income tax reported in the OP.

Like many/most CA homeowners, my CA property tax is going up this year even though the value is going down.

How? I've owned for a while, so the post-crash zillow-value is more than value-for-tax-purposes because Prop 13 limits the year-to-year increase in the latter.

I can't find the report now, but the Merc reported that Santa Clara county's property tax revenues would drop 2%. Counties with a lot of recent construction were seeing 10% drops.

I think chancho is pointing out the effects of the crash of the housing bubble. So yeah, I should have said generally far less procyclical.