Analyses: Self-reporting studies on barrier-free income giveaways unreliable
On Monday, Supe Ellenberg argued in SJ Spotlight that requiring sobriety of welfare recipients is an unscientific, fear-based frippery—that people can be trusted to “make their own determinations” with gov't funds—basing her conclusion on what turn out to be highly questionable, biased reports. Spoiler alert: relying on self-reporting about people's drug habits and spending is unserious and parochial. Various sources excerpted below.
A commonly cited, high-profile Stockton study (2019) on how guaranteed income recipients spend their money concluded it was mostly basic necessities. However, researchers couldn't definitively track 40% of funds, which were withdrawn as cash from recipients' debit cards. PBS provides an overview:
But critics say the experiment likely won’t provide useful information from a social science perspective given its limited size and duration. …
People in the program get $500 each month on a debit card, which helps researchers track their spending. But 40% of the money has been withdrawn as cash, making it harder for researchers to know how it was used. They fill in the gaps by asking people how they spent it.
Since February, when the program began, people receiving the money have on average spent nearly 40% of it on food. About 24% went to sales and merchandise, which include places like Walmart and discount dollar stores that also sell groceries. Just over 11% went to utility bills, while more than 9% went to auto repairs and fuel.
The rest of the money went to services, medical expenses, insurance, self-care and recreation, transportation, education and donations.
Read the whole thing here.
Similarly, Cleveland State Law Review (2013) pokes holes in barrier-free welfare studies often viewed as infallible by guaranteed income advocates:
Despite the fact that there is no constitutional requirement of a high prevalence of drug use to find a special need [to drug test welfare recipients], critics claim welfare recipients do not have higher rates of substance abuse as support that there is no special need for the warrantless search. However, recent studies have found these reports unreliable and inaccurate because previous studies relied on self-reporting for collection. “The exclusive reliance on self-report data is a serious limitation of these findings. Many experts now consider data on the prevalence of substance use drawn from the [National Household Survey on Drug Abuse] to be unreliable because of underreporting.” A 1998 study on welfare recipients in New Jersey, for example, “found that 12 percent self-reported cocaine use, but 25 percent tested positive for cocaine use based on hair sample analyses.” In addition to the self-report data, study results “have varied widely in their findings, with rates of between 4 and 37 percent reported. Much of the difference in prevalence rates found in these studies is due to different data sources, definitions and measurement methods, particularly the different thresholds used to define substance abuse.” …
[A report from the Foundation for Government Accountability in Florida] analyzed the first month of Florida’s implementation of the suspicionless drug testing program. Opponents cite the two percent positive test results as a evidence of a failed policy. However, as the report explains, this number was so low because “denials for incomplete applications due to missing drug test results do not appear until the following month.” The report explains
in July there were only 9 applicants denied for a drug-related reason, but the number of drug-related denials climbed to 565 in August (reflecting the one month lag). Of these 574 total drug-related denials, only 9 were for a positive test. Almost all remaining applicants never completed a drug test even though these individuals completed all other steps in the application process and were determined eligible once DCF received negative drug test results.
This data is more accurate because it reflects individuals who chose not to complete their application due to the testing requirement. Taking this additional factor into consideration, the study concluded “July 2011 represent[ed] annualized savings to Florida taxpayers of $922,992. The cost of reimbursing the 5,390 approved applicants with a negative drug test ($30 average for each) reduces this annualized savings figure by $161,700, for a net savings to taxpayers of $761,292 for the first month of the program alone.” The report estimated that “if these July trends continue[d] throughout the first year, the drug testing requirement w[ould] save Florida taxpayers $9,135,504 from July 2011 through June 2012.”223 Similar to the variance in unemployment rates, drug use prevalence will vary between different states. Therefore, it is best left to each state to conduct its own cost-benefit analysis according to its individual situation.
Read the whole thing here.
Related:
Mayor Breed: Homeless residents should have to pursue sobriety before receiving City money
Thriving TX homeless shelter a model contra disastrous Housing First doctrine, experts say
Follow Opportunity Now on Twitter @svopportunity
Opp Now enthusiastically welcomes smart, thoughtful, fair-minded, well-written comments from our readers. But be advised: we have zero interest in posting rants, ad hominems, poorly-argued screeds, transparently partisan yack, or the hateful name-calling often seen on other local websites. So if you've got a great idea that will add to the conversation, please send it in. If you're trolling or shilling for a candidate or initiative, forget it.