Another day, another salty article decrying Texas’ business success. The author of that piece, Louise Story, attempts to paint Texas’ corporate incentives as coming at the expense of public spending.
The problem with that analysis, though, is that spending and tax incentives have the same effect on cash flow. Whether you spend $100M on a new school or cut corporate taxes by $100M, you still wind up with $100M less cash on hand at the end of the day. From the article:
For Texas to give up more than $250 million in tax revenues in exchange for 2,500 jobs amounts to about $100,000 per job. Most distribution workers are paid $20,000 to $30,000 a year.
Or, you could interpret the deal as a $250M state funded jobs program. Because, by cash flow, that’s exactly what it is. As I’ve said repeatedly, the political spectrum isn’t a spectrum as much as it is a circle. The further to either side you find yourself, the closer you get to the other side. Texas’ traditionally conservative business incentives are monetarily the equivalent of a liberal government program.
Oh, and another thing: the $100K/job figure quoted above is actually spread over 7 years, since the article says Amazon opened the facility in 2005 and enjoyed the tax exemption until July 2012. $100K/7years = $14 286/year spent per job that pays $25K/year on average. And that’s per job that wouldn’t be located in Texas without the incentives in the first place. The kicker here is that even if Texas had taken the far left route and simply funded state jobs or a state welfare program with that same amount of money over that period of time, the people in question would have received 43% less ($25000 – $14286/$25000) than they currently do working for Amazon. Private enterprise works, folks.
After rereading the article, I realized that the 2500 jobs are what Amazon has promised to create, not the number of workers onsite over the past 7 years. A quick Google search about the possible closing of the distribution center brings up this article from the Texas Tribune which says the center employed 120 people at the time of its writing.
While my original point about the equivalence of spending and tax incentives still applies, obviously the above makes the math a lot worse for Texas depending on what the state considers payoff. If the objective of the incentive is to increase private payroll, the tax incentive is the $250M, and the average earnings of a warehouse worker is $25K/year, then
in the worst case it takes $269000000/(2500 * $25000) = 4.3 years 250000000/(2500 * 25000) = 4 years for the state to recoup its investment in terms of money pumped into the local economy. Given that the last tax agreement between Amazon and the state lasted 5 years from 2005 through to 2009, it’s safe to assume the deal will result in at least 2500 * 25000 * 5 = $312.5M of payroll before something else comes up. In lay terms, Texas as a region stands to make much more in increased payroll than the state government spent on creating the corresponding jobs. This comparison is meaningless by accounting standards but provides 2 light years of mileage to any politician who wants to claim responsibility for it.
But on a lost taxation revenue basis, assuming these workers save nothing at that income level, all their expenses are subject to Texas’ 6.25% sales tax rate, Texas continues to have no income tax,
and that both inflation and wages rise at the current inflation rate of 2.16%, the payback equation becomes (2500 * 25000 * 0.0625 * (1.0216^(2x))) = 269000000 250000000/(2500 * 25000 * 0.0625). The solution to that equation is a whopping 99 198 years 64 years. It doesn’t take a rocket scientist to see that deals like this without some alternative revenue source (which includes cost cutting elsewhere) are disastrous for state government finances.* This would explain the state cutting $5.4B from education as Story says.
However, Story’s argument that Texas’ incentives are hurting education doesn’t hold water: the state’s high school graduation rate is 3rd in the nation.
*The calculations get a lot more involved if you factor inflation in. In fact, my attempt to account for inflation is the reason behind the numerous strikethroughs in the preceding calculations. For this case, we’ll have to resort to good ol’ calculus.
Letting S(t) = salary at a given year t and r being the rate of inflation, the rate of change in salary per year is dS(t)/dt = rS(t). Integrating that expression to gives us S(t) = Ce^(rt), where C is a constant. Since the salary at the outset is $25000, we know that S(0) = 25000 and therefore C = 25000 by substitution. So now we have an equation for salary vs. time: S(t) = 25000e^(rt). An expression for the total cumulative (i.e. summing all the previous years up) salary earnings per worker is found by integrating S(t) from 0 to t, which gives 1.15741 * 10^6 * e^(0.0216 * t) – 1.15741 * 10^6. If we treat the $250M state expense as a zero interest loan, then we can solve for the number of years for Texas to recoup the tax loss via sales tax on salary by: 0.0625 * 2500((1157410e^(0.0216t)) – 1157410) = 250000000. The solution to that is 40.2 years. That’s a long time, but it’s not inconceivable for a company to keep operating in the same place with a minimum number of workers for 4 decades. Also, zero interest government loans are a common practice.
However, if we apply the time value of money to the $250M expense, the right hand side of the final equation above changes to (250000000 * (1.0216^t)) using the compound interest formula. The solution to that new equation is 1408 years. No further comment needed there.
We can take our calculation a step further and ask how large the tax incentive should be if Texas wants to recoup the money spent within 5 years, assuming the $250M is effectively a zero interest loan. The equation for that is 0.0625 * 2500((1157410 * e^(0.0216 * 5)) – 1157410) = x, where x is the tax incentive. The result is $20.625M.