Florida and Welfare Reform

Florida passed a law requiring applicants for welfare to take a drug test.  Applicants pay for the test but those who pass are reimbursed by the state.  Applicants found using drugs would be ineligible for welfare for a year although the children of the drug users would still be eligible through a third party.  Since the state began testing in July, 96% of the applicants were drug free.  Another 2% were disqualified for using drugs and the remaining 2% were not completing the application process for unknown reasons. 

Based on these numbers, Catherine Whittenburg (“Welfare drug-testing yields 2% positive results”) calculates the cost of testing versus the estimated savings in welfare payments to taxpayers. 

Cost of the tests averages about $30. Assuming that 1,000 to 1,500 applicants take the test every month, the state will owe about $28,800-$43,200 monthly in reimbursements to those who test drug-free.

That compares with roughly $32,200-$48,200 the state may save on one month's worth of rejected applicants.

The savings assume that 20 to 30 people -- 2 percent of 1,000 to 1,500 tested -- fail the drug test every month. On average, a welfare recipient costs the state $134 in monthly benefits, which the rejected applicants won't get, saving the state $2,680-$3,350 per month.

But since one failed test disqualifies an applicant for a full year's worth of benefits, the state could save $32,200-$48,200 annually on the applicants rejected in a single month.

Net savings to the state -- $3,400 to $8,200 annually on one month's worth of rejected applicants. Over 12 months, the money saved on all rejected applicants would add up to $40,800-$98,400 for the cash assistance program that state analysts have predicted will cost $178 million this fiscal year.

I like Whettenburg’s “back of the envelope” calculations and I have some thoughts but reach no conclusions.  Utilitarians and libertarians might find the savings insufficient to justify the cost in terms of return on dollars invested or government intrusion in private lives but I doubt that typical taxpayers would come to the same conclusion based on classroom experiences.  While covering the composition of federal, state, and local budgets in class, students frequently suggest that tax payers could save billions of dollars by eliminating welfare fraud. One time this happened, I decided that it was a good time to bring up the principle of tradeoffs. Fraud can be eliminated but at a cost. I took an informal survey of students and found that they view welfare fraud as being so morally reprehensible that they would increase payments for enforcement even if the cost of enforcement exceeded the reduction in welfare due to fraud. I repeated the survey in several other classes and found the same result. 

The reported statistics may miss important costs and benefits.  Potential applicants who use drugs will self-select out.  Why put up the money for the test if you are not going to pass?  The number of applicants should decline.  These non-applicants still need money.  Some will engage in or increase their participation in prostitution, illegal drug sales, and other criminal activity to support their drug use.  While a taxpayer may dismiss the cost of drug use on the user, it is more difficult to ignore the impact on the users’ children and the victims of their crime.  Other non-applicants might give up drugs to qualify for government assistance. 
Many economists have found that taxpayers in ethnically diverse communities are less willing to pay welfare than taxpayers in homogenous communities.  Raghuram Rajan summarizes this view in “Fault Lines” (page 95).

We should also not minimize the importance of population heterogeneity.  “There but for the grace of God go I” offers a powerful rationale for social insurance.  People are more willing to be taxed to benefit others if they believe that the benefits go largely to people like themselves, and not disproportionately to groups they do not identify with.  This may also explain why Americans give generously to charities: they have more control over who the beneficiaries are.  Politicians who want to derail benefits legislation have often been quick to raise the specter of hard-earned taxpayer money going to the undeserving, irresponsible, and lazy, and such demagoguery is especially potent when the bogeymen look and behave differently from their constituents.

By demonstrating that welfare recipients are not drug users, taxpayers in heterogeneous communities may ironically be more willing to fund welfare programs.  Requiring welfare applicants to pass a drug test will create many interesting questions for economists to answer in their research.