If the American Community Survey is eliminated, fraudsters will try to fill the void
June 15, 2012 § Leave a comment
When the U.S. House of Representatives voted last month to eliminate the American Community Survey (an ongoing statistical survey of American households, administered annually by the Census Bureau), news organizations, academic researchers, and business associations (including the U.S. Chamber of Commerce) were united in their dismay. If government agencies, businesses, and nonprofits are rendered blind by the defunding of research programs that provide them with information about the socioeconomic makeup of communities across the country, how are they supposed to decide where to build new schools and roads, where to open new stores and hospitals, etc.? One does not have to be an expert on statistics to predict that even the “compromise” solution proposed by Senator Rand Paul – to make participation in the survey voluntary rather than mandatory – would inevitably result in a marked decrease in the quality of data collected by the Census Bureau. (I do suspect that this may come as a surprise to Representative Daniel Webster [R-FL], who revealed a troubling lack of knowledge of the basics of research methods when he suggested that defunding the survey would not be a big deal, because “in the end this is not a scientific survey. It’s a random survey.”)
Even more disconcerting is the prospect of fake data manufactured by unscrupulous political operatives filling the void left by the dismantling of legitimate data collection programs. And this prospect is neither far-fetched nor purely theoretical. Consider PolitiFact’s investigation of the widely reported finding that 85% of recent U.S. college graduates end up living with their parents after graduation. In countries where it is fairly common for unmarried adult children to live with their parents – Japan is one example of such a place – this statistic would hardly cause a stir, but in the U.S., it is nothing short of scandalous. So, it comes as no surprise that the figure has been prominently featured not only in an anti-Obama ad produced by Karl Rove’s PAC American Crossroads, but also in TIME Magazine, New York Post, CNN Money, U.S. New and World Report, and other news sources. Well, it turns out the statistic is fake.
PolitiFact’s inquiry (covered by NPR’s On the Media on May 11, 2012) strongly suggests that Twentysomething Inc., the now defunct “prestigious, world-renowned management consultancy” that was the apparent source of the made-up statistic, was itself an elaborate sham – complete with nonexistent employees and stock photos downloaded from the internet and passed off on the company’s web site as pictures of its staff. The only outstanding question is who paid for this fraud. (When Louis Jacobson of PolitiFact finally spoke with David Morrison, the founder and former president of Twentysomething Inc., Mr. Morrison stated that a nondisclosure agreement barred him from commenting on the “poll” that had supposedly produced the 85% figure or from revealing the identity of the client who had commissioned the study.) We can be sure about one thing: it is highly unlikely that this example of fraudsters-for-hire deliberately feeding disinformation to the media and the public is unique.
In 2008, I attended a public debate between two academics – one representing the liberal camp and the other speaking for the conservatives – at a large private university in the Northeast. The topic of the debate was whether the media should play the role of fact-checkers in presidential elections. The conservative scholar argued in all seriousness that fact-checking was not a proper task for the media because [a] journalists lacked the skills/training/expertise to make judgements about the veracity of candidates’ statements and [b] most of what candidates said on the campaign trail was not factual anyway. Judging by the failure of multiple news organizations to verify Twentysomething’s “data” before publishing them, many reporters and news editors have decided to take it to the next level and abandon fact-checking altogether. Supporting quality journalism and taking media companies to task when they fail to verify factual claims before reporting them is one way to counteract the not-so-subtle and, to date, largely successful push to create an environment that allows politicians to lie without fear of any real push-back from the media. More important, we have to ensure that decision-makers at all levels have access to comprehensive and accurate socioeconomic data.
Academic institutions and independent research organizations have an important role to play here. (One of the clues that the 85% estimate of the number of recent college grads living with their parents is fraudulent can be found in a recent report by the Pew Research Center, which puts the real numbers for all young adults at 12% for 25- to 34-year-olds and 40% for those ages 18 to 24 as of December 2011. Interestingly, the Pew report notes that the vast majority of 18- to 24-year-olds currently living with their parents or grandparents say “they did not move back home because of economic conditions [in fact many of them may have never moved out in the first place].” These figures may still be too high for comfort, but they are a far cry from the 85% figure concocted by David Morrison.) However, much of the analysis performed by nongovernmental research organizations relies on raw data gathered by the American Community Survey, the Economic Survey (which measures the health of the U.S. economy and which the House Appropriates Bill eliminates as well), and other government-sponsored projects. In the end, only the Census Bureau has the capacity and the mandate to survey 3 million American households each year and 5 million businesses every 5 years to produce massive datasets used by researchers, government agencies, nonprofits, and businesses across the country. Eliminating the American Community Survey project would undoubtedly please the paranoid fringe and their champions at the Cato Institute who claim that questions about monthly utility costs, level of education, commuting time, and the availability of flush toilets constitute an unconstitutional invasion of privacy and bring the country one step closer to a totalitarian dystopia. (It is hard to imagine why anyone outside Congress would seriously advocate for the elimination of the Economic Census.) The rest of us have every reason to tell Congress to leave these research programs alone and let the Census Bureau do its job.