28 May 2009

Simulations vs. Reality

If you haven't been paying attention to the ag-water-jobs debate, here's an update.

In January, Howitt et al. published an academic study -- a simulation -- that estimated 60,000-80,000 job losses would result from the reduction in water exports from the Delta. (I went to Firebaugh to find out how it was looking on the ground. Their unemployment was at 40%, but that was one data point.)

Howitt, btw, is my adviser; I'm having a beer with him later today. (I hope he doesn't mind this post :)

Importantly, the simulation took a reduction in water supply, ran it through a crop optimization model, and then ran that estimate through an economic model that imputed lost jobs from the reduction in gross crop value.

Put differently, this back-to-back simulation is like hitting the cue ball (water) to the red ball (crops) to the striped ball (economic output) to the green ball (jobs). Can a simulation tell you where the green ball is going? Sure -- as long as each hit is perfectly on target, the consistent-friction surface is flat, and the spin off the pool cue is neutral.

Of course, things can go wrong. Imagine that you were playing pool with balls that changed shape when you hit them, on a table that was lumpy and vibrating, with a warped cue stick that changed length on every shot. Oh, and you are blind in one eye...

Nevertheless, simulations are a good way of thinking of how things fit together, and they give you an idea of the relative magnitudes of responses. (Note, of course, that those magnitudes are often specified in advance, when the parameters are given.)

So, taking these shortcomings for granted, how useful are simulations in the real world? Well (as the saying goes), it depends on who's paying...

Simulation results are often used in arguments over policies, laws, regulations, etc. when people are trying to understand the implications of actions. The trouble begins when they put too much weight on the results, and that trouble often begins because people want to believe the results.

So what are the results so far?

According to California's Employment Development Department [XLS], year-on-year farm employment (April 2008 to April 2009) is up by 4,900 (or 8.7%) in the eight south of Delta counties.* For the entire State, it's up by 9,400 (2.5%). Against two and three years ago, the numbers are also up.

Put differently, there is no sign of 60,000-80,000 lost jobs. (Non-farm employment is MUCH worse, down 31,000 in the eight counties and 687,000 for the entire State.)

I have no trouble concluding that the south-Central Valley is better off and that -- inside this area -- farm workers are better off relative to non-farm workers.**

But the political wheels have not been waiting around for real data. In this post, I note the grandstanding of politicians claiming that their districts are going to dry up and blow away if the water is not turned on. In this post, I discuss how the "farmworkers march for water" was more about land owners.

The mayors of San Joaquin and Mendota have resorted to ad hominem attacks against those who do not support their view that water will make their economy recover.
Claims that unemployment is not connected to drought are, in our opinion, a challenge to our honesty. That is why we have to stand up and describe the reality of the situation in defiance of those who would prefer that we just go away. We will not let someone unfairly label us because we are Latinos and are not from some privileged class.
Now the mayors are referring to this op/ed by Jeffrey Michael, who looked at the numbers:
Since the drought began three years ago, Fresno County farm payrolls have increased by 12 percent, while nonfarm employment has crashed, led by a loss of more than 7,000 construction jobs.

In light of these statistics, how can water exporters, politicians and others claim that rising unemployment in the Valley is a result of water shortages for farms rather than the broader recession? The foreclosure crisis is at the heart of the recession, and the Central Valley has the highest foreclosure rates in the United States. Homebuilding has shut down, and service sectors have cratered, costing many former farmworkers their higher paying, nonseasonal jobs.

Water contractors point to 40 percent unemployment in Mendota as evidence of the water crisis. These unemployment estimates for towns aren't a current survey, but are crude extrapolations from the 2000 Census, the last time any real data were compiled for these areas.

The 2000 census gives a good picture of the prosperity that increased water pumping would bring to Mendota's hard-working residents. Delta water exports were above average in 2000, and local farm employment was at a nine-year peak. Despite this, the 2000 census found unemployment in Mendota exceeded 32 percent, highest of the state's 494 towns.
I agree with this assessment -- especially as it is the conclusion that Michael draws from real data. (And, yes, I am a micro economist talking about macro issues, but I surely understand markets!)

In a development that is probably NOT coincidence, Judge Wanger ruled last week that
The human residents of the Central Valley must be taken into consideration when the federal government comes up with water allocation rulings to protect an endangered fish.
I wonder if the judge had the employment statistics at hand when he made his decision -- or was he listening to the people telling stories of how they were suffering?

(Interestingly, Senator Feinstein reintroduced a bill relaxing restrictions on seasonal farm workers. Either she agrees that there are plenty of jobs for new workers OR she wants to make it easier to bring NEW workers to different places -- leaving the current "unemployed" workers stranded, complaining about the lack of water! Either way, the bill will help farmers -- the same farmers who want the water to flow to their farms, and the same farmers who have given DiFi millions in campaign contributions.)

Bottom Line: Political and judicial water allocation is fraught with difficulty. When it's THAT easy to take water from one person and give it to another, you can be sure that mistakes in allocation will be made. That's why I favor markets. They allocate based on willingness to pay, not willingness to contribute to a re-election fund or willingness to grandstand on camera.
* Stockton country is more in the Delta than south of the Delta, but let's let that slide.

** As a rule, farm workers are not absolutely better off, but here, they are relatively better off.


Kevin Dick said...

I really like your analogy about simulation models. I hope you don't mind if I shamelessly steal it. It captures perfectly how I feel about the climate models :-)

Anonymous said...

Your coverage of the central valley water crisis makes one key mistake. It assumes that growing crops are the major economic activity in the valley these days. Its not. Paving over the world's most valuable ag land and growing more sprawl housing subdivisions is. Scratch a factory farm owner and you'll find a wanabe real estate developer.

David Zetland said...

@Kevin -- feel free but remember that climate sims have about 50 balls!

Post a Comment

Note: only a member of this blog may post a comment.