A new study from researchers at the University of Maryland School of Public Health published in Epidemiology assess the accuracy of self-reported U.S. Department of Housing and Urban Development (HUD) assistance in the National Health Interview Survey (NHIS). Using individually linked administrative records, the researchers found a relatively high rate of false reporting in the NHIS but no evidence that it biased estimates of the association between health and rental assistance.
The researchers underscore that by improving housing affordability, quality, and stability, rental assistance from the HUD’s rental assistance programs have the potential to improve the health of low-income populations. To support evidence of this, the researchers highlight the potential of federal surveys to estimate the effects of HUD assistance on health but identify that “Self-reports of rental assistance are considered to be inaccurate.”
Using a three-pronged approach to assess the accuracy of self-reports of rental assistance, the researchers “Conducted a record-check study using a version of the 2004–2012 NHIS that was linked to HUD administrative records.” Only NHIS respondents who did not refuse the rental assistance question and provided complete information on the variables used in the linking process were included in the sample.
The researchers first created new weights for the NHIS sample to ensure that the sample was representative of the NHIS target population on every variable included in their analysis. They then calculated the rate of false-negatives for the total population and for subgroups. Lastly, they examined the extent to which false-negative reports biased estimates of the association of HUD assistance and self-reported health.
The researchers estimated a fairly high false-negative rate of 22 percent and found that misclassification was higher among public housing residents. Comparing self-reported, administrative, and edited measures, they found no large discrepancies in the differences in self-reported health between assisted and non-assisted respondents.
The researchers add that “Although the level of misclassification is concerning, misreporting did not appear to substantially bias estimates of the association between assistance and health status.” They state that “This suggests that data users can have some confidence in estimating correlations between assistance and health using multivariable regression.”
The paper, “Misclassification of Rental Assistance in the National Health Interview Survey: Evidence and Implications,” was written by Drs. Michel Boudreaux, Andrew Fenelon, and Natalie Slopen from the UMD School of Public Health.