Whose Voice Do We Hear in the Marketplace? Evidence from Consumer Complaining Behavior

Abstract

Victims from heavily minority areas complain to the government much less than those from nonminority areas, relative to their degree of victimization.

Which Communities Complain to Policymakers? Evidence From Consumer Sentinel

“…As Figure shows, differences in the number of complaints per victim could be because of differences in the consumer's recognition of victimization, in the company's response to complaints directed to them, or in the decision to complain to consumer protection authorities for consumers who realize they were victimized. To examine how demographics might affect victimization, I construct an implied victimization rate by multiplying the complaint rates in Consumer Sentinel by a set of weights developed in Raval () as the predicted victim rate divided by the predicted complaint rate . Because Raval () used data from only nine consumer protection cases, for external validity, this exercise requires that estimates of the propensity to complain for these cases extend to a broader set of complaints.…”

“…In addition, consumers alienated from mainstream institutions are less likely to choose to interact with them, including by complaining. In Raval (), I assemble considerable evidence that minority consumers in particular have less social trust—they are much less likely to answer affirmatively to questions on social trust in the General Social Survey, and the qualitative literature in sociology and marketing has also found evidence that members of minority groups experience more alienation from society, and have less trust.…”

“…Complaint rates are the combination of the victimization rate and the propensity to complain given victimization, so differences in complaint rates may not reflect differences in victimization. In a companion paper, Raval (), I examined how the propensity to complain conditional on victimization varied across demographics by combining company supplied data on victims together with Consumer Sentinel complaints for nine consumer protection enforcement cases. In these cases, all purchasers were considered to be defrauded.…”

“…In contrast, this paper examines the demographics of the per capita complaint rate using all complaints to Consumer Sentinel. To remove demographic differences in the propensity to complain from the complaint rate, I use a set of weights constructed in Raval () as the ratio between the predicted victimization rate and complaint rate for each zip code based on demographics. I then construct implied victimization rates as the aggregate per capita complaint rates on fraud cases multiplied by these weights.…”

“…I then construct implied victimization rates as the aggregate per capita complaint rates on fraud cases multiplied by these weights. Doing so requires one to extrapolate that the differences between victims and complaints for the nine cases used in Raval () hold for fraud‐related issues more generally.…”