Thursday, November 19, 2009

Guest Post About County Fiscal Policy And Jobs

I'm happy to post the work of Ph.D candidate John Kovari, who has worked in City Hall and at the Public Policy Forum.


We need more polished, young thinkers like this.

-----------------------------------------------------------------------------------------------------------------

Lately I’ve been using economic data for some research on economic development, and curiosity led me to test some political assumptions that have popped up recently.

I started with looking at the relationship between taxes and jobs. Debate over the 2010 Milwaukee County Budget has highlighted competing philosophies about how the county’s property tax rate can impact the economy. Shaping the debate is County Executive Scott Walker’s insistence on lowering the property tax rate. It’s an interesting approach, and I wanted to see how the economic data supported it.

First, from Walker’s campaign website, his basic approach:

The fastest, most effective way to create new jobs is to cut taxes [italics added] and implement regulatory and fiscal policies that encourage job growth and economic investment.”

So what does the historical evidence about property tax rates and jobs size up with Walker’s claim? The statistics I’m using here are the Milwaukee County property tax rate and jobs per 1,000 residents, as identified the federal QCEW reports. In the analysis, taxes are lagged by a year so we can find how taxes in 2000, for example, translated into the number of jobs in 2001.

When we look at the Milwaukee County data between 1990 and 2006, a surprising dynamic emerges. Jobs and tax rate have a positive statistical relationship. Go figure. For you stats geeks, the Pearson’s correlation coefficient is 0.536 and statistically significant (p<0.05).

While the small sample size raises concerns, the econometric evidence doesn’t seem to produce strong evidence for Walker’s claim about stimulating job creation with lower taxes. Key Walker years (2003-2006) have the lowest observed tax rates but also the lowest number of jobs per thousand residents. Years with higher taxes were actuallyassociated with more jobs. (Email me at jpkovari@uwm.edu if you want a graph and the data.)

So, should we use this as evidence for raising the county’s property tax rate and hope for job growth? Of course not.

What the evidence suggests is that the economic situation in counties is more complex than simply focusing on the tax rate. Many other factors are in play when it comes to job creation, including proximity to markets, proximity to public services, quality of infrastructure, educated workforce, etc. Economic development depends more on developing these amenities than just lowering taxes and tax rates.

From a quick look at this economic evidence, Walker’s claim about lowering taxes as the fastest way to stimulate job creation seems less valid. Milwaukee County Supervisors might want to keep this in mind when considering Walker's upcoming vetoes on the 2010 Budget.

John Kovari is a Ph.D. student in political science at the University of Wisconsin-Milwaukee and the 2008-2009 Norman N. Gill Fellow at the Public Policy Forum.

2 comments:

Anon Jim said...

Always heartwarming to see a Poli-Sci Egghead acting like they understands economics.

And using questionable & inappropriate statistics to support their conclusions is just a standard part of the bureaucratic playbook.

Nathaniel Holton said...

The county has a fiscal policy? Ha.

Nice post. Correlation doesn't equal causation. I think the other
factors that he mentioned, and especially the state of the larger
economy, are going to overwhelm any positive or negative effect that
relatively minor changes in local property tax may have on local jobs.

You'd have to try to cancel out larger factors - national GDP and state taxes/incentives for example - to even have a chance to look at the relevance of local taxes. For all we know, it could be that the
national economy took Milwaukee down in those years, but county tax
cuts made the downward spiral less significant than it would have otherwise been. Doubt it, but you can't tell.