We live in an era when people do not seem to want to do surveys anymore. Response rates have been falling since the 1970s, and even high-quality telephone surveys now often have response rates in the low one figures. The various countermeasures survey researchers have taken to combat this trend-particularly making more attempts to contact sample members-have sharply raised data collection costs. For example, the per household cost of conducting the US Decennial Census was seven times greater (in constant dollars) in 2010 than in 1970.3 The twin phenomena of falling survey response rates and increasing survey costs are well documented but poorly understood. It is clear that people are much more likely to refuse to take part in surveys now than they were in the past, although it is not clear why. And the trend appears to be affecting the entire developed world, not just the United States. Because of the increased risk of bias and the higher costs of doing surveys, some researchers have advocated turning to nonprobability samples, such as volunteer Web panels, or abandoning surveys altogether in favor of administrative data. At the same time, the demand for accurate information on a range of topics has only increased. Despite the unfavorable environment for surveys, some researchers have taken up the challenge of conducting studies of hard-to-survey populations, such as small ethnic minorities or itinerant populations, like the Irish Travelers. The description of their study at the National Center for Health Statistics by Galinsky et al. (p. 1384) is an extremely useful addition to the literature on hard-to-survey populations. Populations can be difficult to survey for a variety of reasons. They can be difficult to sample (e.g., because population members are rare or highly mobile); they can be hard to identify (e.g., because members are reluctant to admit to being part of a stigmatized population); they can be hard to locate (e.g., because members move frequently) or to contact (e.g., because gatekeepers bar access to members of the population); they may be hard to persuade to take part (e.g., because members mistrust the authorities); or they may be hard to interview (e.g., because of language barriers). The primary challenges for the Native Hawaiian and Pacific Islander (NHPI) National Health Interview Survey were the rarity of this population (which comprises less than 1 in 200 Americans) and their potential mistrust of the researchers.