Florida: Drug Testing Welfare

Drug Test Welfare

Much ado has been made of Florida’s attempt to drug test welfare applicants.  Florida’s idea is that if an individual has enough discretionary income to afford drugs, that person has enough income to be disqualified from a need based program.

Lately, the left has been trying to make the case that conservatives should be against the initiative based on fiscal calculations:

From July through October in Florida — the four months when testing took place before Judge Scriven’s order — 2.6 percent of the state’s cash assistance applicants failed the drug test, or 108 of 4,086, according to the figures from the state obtained by the group. The most common reason was marijuana use. An additional 40 people canceled the tests without taking them.

Because the Florida law requires that applicants who pass the test be reimbursed for the cost, an average of $30, the cost to the state was $118,140. This is more than would have been paid out in benefits to the people who failed the test, Mr. Newton said.

So, I did my own math.  And what I found was interesting.

The Times article above uses the number of failures at 2.6%, however, other sources I’ve read use a number more like 2%.  Because that number helps the liberal cause, I’ll use 2%.  Because Florida has a 6.7% drug use rate, the 2% failure rate is indicative  of a self-selecting population.  That is, drug users are not applying for benefits knowing they will fail, and have to p ay for, a test.

Further, once an individual fails a test, he is ineligible for benefits for 12 months.  A savings not taken into account.  With all of that said, here is my data:

Drug Test Welfare data

Assuming a constant rate of applications and a conservative 2% failure rate the data shows that 25 people fail each month.  Because Florida hums along at a 6.7% drug use rate, I calculated that a certain number of people would simply not even try to test.  Again, to be conservative, I used a 6% rate and found that 50 people each month did not apply.  Finally, I begin to sum the failure and count the benefits that they WOULD have been given as savings.  I do this on a rolling 12 month period.  Additionally, I assume that this month’s drug users will not be using next month and only count 50 Missing Failure each month; I don’t sum them.

To be fair, the short term cost to the state is negative.  For the first 8 months of this system, the state loses money.  However, by September of the of the first year, or month 9, the state begins to realize savings and finally is in the black12 months after that.

The numbers would be even great if we used a normal 6.7% drug use on the welfare population and didn’t make the assumption that this months drug users will not continue to use next month.

As much as I like it when people use numbers and data to make their point, it’s important that such numbers reflect reality.

4 responses to “Florida: Drug Testing Welfare

  1. Unless you can show that the drug testing regime caused a 4% drop in welfare applicants, I think it’s a pretty silly assumption that 4% of the at risk population were drug users who would have been receiving benefits if they weren’t scared off. After all, if drug users aren’t going to apply for benefits because of the testing, why are you still seeing a 2-3% failure rate?

    Here’s an explanation for why the failure rate on TANF is much lower than the general population: drugs cost money and people who get benefits are largely people without disposable income for drugs.

    “As much as I like it when people use numbers and data to make their point, it’s important that such numbers reflect reality.”
    Well, yes, when you get the numbers through statistical sampling or direct observation. But when you take a proportion of the general population and assume it applies to a discrete sub-population, then you are on very shaky ground.

    I’ve known a lot of poor people, rich people, and people in between, and with the exception of alcohol and tobacco, the poor people I’ve known use less drugs than the other two groups.

    • Unless you can show that the drug testing regime caused a 4% drop in welfare applicants, I think it’s a pretty silly assumption that 4% of the at risk population were drug users who would have been receiving benefits if they weren’t scared off.

      I’ve made no assumption on drug use. All that I’ve done is take Florida’s 6.7% drug use rate, trunc to 6% and extrapolate. I make no claim that the poor are more likely or less likely to use drugs.

      My guess? I would guess that poor people spend less in absolute dollars than the wealthy, but that the percentage of their income spent on drug use is higher.

      Here’s an explanation for why the failure rate on TANF is much lower than the general population: drugs cost money and people who get benefits are largely people without disposable income for drugs.

      I agree. That’s a legitimate explanation.

      But when you take a proportion of the general population and assume it applies to a discrete sub-population, then you are on very shaky ground.

      I stipulate that I applied the general ratio to the sub-group. However, I applied it at a discount and counter to my own biases.

      I’ve known a lot of poor people, rich people, and people in between, and with the exception of alcohol and tobacco, the poor people I’ve known use less drugs than the other two groups.

      Not counting college, mainly because it’s a temporary existence that doesn’t describe a real action/consequence situation, I’ve had two life experiences:

      1. Bartending
      2. Corporate manager

      In both cases I’ve found that people have adequate “pocket money”. No one was poor or destitute, all had access to food and shelter. I drove a Chevy Cavalier back then and a Volvo station wagon now. The Chevy was older and had mechanical windows and locks; the Volvo is newer and has many amenities.

      In short, I have the same worries now as then, just that the things are upgraded. I worried about rent then, I worry about my mortgage now. I used to worry about feeding me then, I worry about feeding my family now. I travel virtually the same then as now. I work more hours now but the benefits are better.

      In nether case did I hang with extremes; no one was destitute then and no one has FU money now.

      Drug use in my bartending days was almost universal. Drug use among my peer group now is almost universally opposite.

      • The entire point of the conservative argument is, as you put it, “if an individual has enough discretionary income to afford drugs, that person has enough income to be disqualified.” But then you assume that the people who otherwise qualify for need-based aid are just as able to afford drugs as the rest of the population in order to make your argument. That’s inconsistent. I’m not saying you’re necessarily wrong, but you are assuming something that seems to be unreasonable, and you really need to back it up with more than mere extrapolation.

        In both cases I’ve found that people have adequate “pocket money”. No one was poor or destitute, all had access to food and shelter. … In short, I have the same worries now as then, just that the things are upgraded.

        I don’t know enough about your situation to say this with any certainty, but it seems that you might not have really been exposed to the level of poverty that make one eligible for TANF. You need to be making less than $22,000 a year, you need to be keeping less than $11,000 per year as take-home after expenses, and you need to have less than $2,000 in assets. If you have 2 people in your household, then your combined goes up to $30k in gross income, $15k in net, and the same $2000 asset limit. I mean, that’s a very low bar to get under. You also need to participate in work related activities for 20 hrs a week, etc. These aren’t junkies selling their food stamps for an eight ball, these are just very poor people. Honestly, in a more caring society, these people would probably be eligible for therapeutic weed to deal with their lives.

        • But then you assume that the people who otherwise qualify for need-based aid are just as able to afford drugs as the rest of the population in order to make your argument. That’s inconsistent. I’m not saying you’re necessarily wrong, but you are assuming something that seems to be unreasonable, and you really need to back it up with more than mere extrapolation.

          http://www.samhsa.gov/data/nhsda/1997main/nhsda1997mfWeb-119.htm#Table13.1

          Illicit Drug Use in Last Year – by Income:

          < $9,000 16.6% $9,000 - $19,999 11.2% $20,000 - $39,999 11.3% $40,000 - $74,999 10.2% > $75,000 10.8%

          Further down they break it out by welfare: Receive or Not Receive

          18.0% of Welfare recipients reported using an illicit drug in the past year.

          I went down from 6.7% to 6. I should have gone UP to 18%.

          Florida will be saving a bunch of money.

Leave a Reply

Your email address will not be published. Required fields are marked *