Strangers in the Field: A Methodological Experiment on Interviewer-Respondent Familiarity.

Evaluating a long-term methodological norm - the use of interviewers who have no prior social relationship to respondents - we compare response patterns across levels of interviewer-respondent familiarity. We differentiate three distinct levels of interviewer-respondent familiarity, based on whether the interviewer is directly acquainted with the respondent or their family, acquainted with the research setting, or is a complete outsider. We also identify three mechanisms through which variability in interviewer-respondent familiarity can affect survey responses: the effort a respondent is willing to make; their level of trust in the interviewer; and interview-specific situational factors. Using data from a methodological experiment fielded in the Dominican Republic, we then gauge the effects of each of these on a range of behavioral and attitudinal questions. Empirical results suggest that respondents expend marginally more effort in answering questions posed by insider-interviewers, and that they also lie less to insider-interviewers. Differences in responses to "trust" questions also largely favor insider-interviewers. Overall, therefore, local interviewers, including those whom, in blatant violation of the stranger-interviewer norm, have a prior relationship with the respondent, collect superior data on some items. And on almost no item do they collect data that are measurably worse.

Voulant évaluer une pratique méthodologique ancienne ­ le recours à des enquêteurs qui n'ont pas eu d'interactions préalable avec la personne interrogée -, nous comparons les réponses apportées à un questionnaire en fonction du niveau de familiarité entretenu entre les deux protagonistes. L'article se fonde sur trois catégories distinctes de familiarité : si l'enquêteur est en relation directe avec la personne interrogée ou sa famille, s'il connaît le terrain d'enquête dans lequel il intervient, ou s'il en est complètement étranger. Nous identifions trois mécanismes au travers desquels ce lien entre un enquêteur et un enquêté peut affecter les réponses : le degré avec lequel la personne interrogée est prête à faire des efforts pour répondre, son niveau de confiance envers l'enquêteur, ainsi que des facteurs spécifiques à la situation d'entretien. A partir de données d'expérimentation méthodologique conduites en République Dominicaine, nous avons mesuré les effets de chacun de ces mécanismes sur des réponses à des questions portant sur des comportements et attitudes. Les résultats empiriques suggèrent que les répondants font à la marge plus d'efforts pour répondre lorsque les questions sont posées par un enquêteur familier ; de la même manière, ils lui mentent moins souvent. Les réponses apportées aux questions « de confiance ¼ tendent également largement en faveur d'un enquêteur connu du répondant. Plus généralement, le recours à des enquêteurs locaux, c'est-à-dire issus du terrain et incluant donc, en violation des règles habituellement suivies, des enquêteurs ayant déjà eu des relations avec la personne interrogée, produit des données plus fiables sur plusieurs items. Sur la plupart des items, les données collectées ne sont, en tous les cas, pas moins bonnes.

Can incentives improve survey data quality in developing countries?: results from a field experiment in India.

We report results of an experiment designed to assess whether the payment of contingent incentives to respondents in Karnataka, India, impacts the quality of survey data. Of 2276 households sampled at the city block level, 934 were randomly assigned to receive a small one-time payment at the time of the survey, whereas the remaining households did not receive this incentive. We analyse the effects of incentives across a range of questions that are common in survey research in less developed countries. Our study suggests that incentives reduced unit non-response. Conditionally on participation, we also find little impact of incentives on a broad range of sociodemographic, behavioural and attitudinal questions. In contrast, we consistently find that households that received incentives reported substantially lower consumption and income levels and fewer assets. Given random assignment and very high response rates, the most plausible interpretation of this finding is that incentivizing respondents in this setting may increase their motivation to present themselves as more needy, whether to justify the current payment or to increase the chance of receiving resources in the future. Therefore, despite early indications that contingent incentives may raise response rates, the net effect on data quality must be carefully considered.

Own-Choice Marriage and Fertility in Turkey.

Goode's foundational work on the fertility transition identified own-choice marriage as a factor driving fertility decline, part of a widening repertoire of choice pertaining to marriage and childbearing. Yet research supporting this connection in today's transitional societies is scarce and somewhat contradictory, and it is unclear how other marital traditions, such as consanguineous marriage, shape this relationship. This study evaluates Goode's theorized connection using pooled Demographic and Health Survey data from Turkey, comparing children ever born, use of contraception, and parity progression across four types of marriage: own-choice and arranged marriage; and marriage to a cousin versus an unrelated spouse. Results are largely consistent with the idea that a move towards own-choice marriage reflects a widening repertoire of choice that also leads to fertility decline. However, they also show that hybrid models like own-choice marriage to a cousin tempers these effects.

Land cover change and fertility in West-Central Africa: rural livelihoods and the vicious circle model.

The vicious circle argument, rooted in a neo-Malthusian tradition, states that resource scarcity increases the demand for child labor and leads to higher fertility. The rural livelihood framework, on the other hand, contends that households employ multiple strategies, only one of which involves adjusting their fertility levels as a response to environmental pressures. This study provides a unique test of both theories by examining the relationship between land cover change and fertility across hundreds of rural communities in four West-Central African countries. The findings reveal a complex relationship between natural capital and fertility. In communities where natural capital was initially low, a further decline in that capital is associated with both higher fertility preferences and levels. However, we find that fertility preferences and behavior are often discordant, with notable within-community differences in response to decline in natural capital across levels of household wealth.

The exclusion from welfare benefits: Resentment and survey attrition in a randomized controlled trial in Mexico.

Public policy programs must often impose limits on who may be eligible for benefits. Despite research on the impact of exclusion in developed countries, there is little evidence on how people react to being excluded from benefits in developing societies. Utilizing repeated waves of data from an experimental evaluation of Mexico's foundational PROGRESA antipoverty program, we examine the impact of exclusion and distinguish two separate forms. "Statistical exclusion" occurs where determination of benefits is based on randomized assignment to a treatment and control group. "Needs-based exclusion" occurs when benefits programs are designed to be selective rather than universal, basing eligibility on characteristics, like relative poverty, that are difficult to measure simply and accurately. Focusing on temporal variation in survey non-response as our behavioral outcome, we show that needs-based exclusion has much greater negative effects on continued participation than statistical exclusion. We also show that these effects are concentrated among the wealthy, that is, those furthest from the eligibility cut-off line. These findings reinforce general concerns about the validity of evaluation studies when incentives are at work. We discuss both the behavioral explanations that might underlie these findings as well as some potential approaches to reduce threats to evaluation validity.

A test of the stranger-interviewer norm in the Dominican Republic.

We offer the first empirical test of the 'stranger-interviewer norm', according to which interviewers in social, demographic, and health surveys should be strangers-not personally familiar with respondents. We use data from an experimental survey in the Dominican Republic that featured three types of interviewer: from out of town (outsiders); local but unknown to the respondent (local-strangers); and local with a previous relationship to the respondent (insiders). We were able to validate answers to up to 18 questions per respondent, mainly by checking official documents in their possession. Contrary to expectations derived from the stranger-interviewer norm, respondents were more reluctant to show the documents needed for validation when the interviewer was an outsider. Furthermore, and again at odds with the stranger-interviewer norm, we found no difference in accuracy by type of interviewer. Our results have important implications for the selection of survey interviewers in less developed and non-Western settings.

Family Planning for Strangers: An Experiment on the Validity of Reported Contraceptive Use.

Sterilization levels reported in the Dominican Republic appear well above what we would normally expect given prevailing patterns in the region. We suspect that the use of strangers as interviewers--the normative approach in data collection in both developed and developing country settings--may be partly responsible for this result, and may underlie a long history of bias in family planning data. We present findings from a field experiment conducted in a Dominican town in 2010, where interviewer assignment was randomized by level of preexisting level of familiarity between interviewer and respondent. In our data, sterilization use is higher when the interviewer is an outsider, as opposed to someone known to the respondent or from the same community. In addition, high sterilization use is correlated with a propensity of respondents to present themselves in a positive light to interviewers. These results call into question the routine use of strangers and outsiders as interviewers in demographic and health surveys.

Military westernization and state repression in the post-Cold War era.

The waves of unrest that have shaken the Arab world since December 2010 have highlighted significant differences in the readiness of the military to intervene in political unrest by forcefully suppressing dissent. We suggest that in the post-Cold War period, this readiness is inversely associated with the level of military westernization, which is a product of the acquisition of arms from western countries. We identify two mechanisms linking the acquisition of arms from western countries to less repressive responses: dependence and conditionality; and a longer-term diffusion of ideologies regarding the proper form of civil-military relations. Empirical support for our hypothesis is found in an analysis of 2523 cases of government response to political unrest in 138 countries in the 1996-2005 period. We find that military westernization mitigates state repression in general, with more pronounced effects in the poorest countries. However, we also identify substantial differences between the pre- and post-9/11 periods.

Hotspots and Coldspots: Household and village-level variation in orphanhood prevalence in rural Malawi.

We explore the characteristics of households and villages in which orphans are resident in two areas of Malawi. We first review pertinent themes in qualitative data collected in our research sites. Then, using spatial analysis, we show how positive and negative clusters of orphans - which we term orphanhood "hotspots" and "coldspots" - can be found at the village and sub-village levels. In the third and longest section of the paper, and using multilevel analyses with both simple and complex variance structures, we evaluate the relationship between the presence of orphans and a range of individual, household and village-level characteristics, including households' spatial relationship to each other and to other local sites of significance. This series of analyses shows that the most important covariates of orphan presence are household size, wealth, and religious characteristics, with all measured simultaneously at both household and village-level. In addition, most of these have heterogenous effects across villages. We conclude by reviewing some difficulties in explaining causal mechanisms underlying these observed relationships, and discuss conceptual, theoretical and programmatic implications.

Estimation of AIDS adult mortality by verbal autopsy in rural Malawi.

OBJECTIVE: To estimate AIDS-related adult mortality using verbal autopsies (VA) in a rural sub-Saharan African setting with high HIV prevalence. DESIGN: VA data were collected from relatives and neighbors that described the circumstances leading to death for 92 adults in three rural areas of Malawi. These adults were initially interviewed in 1998 as part of a longitudinal social survey of a sample of 1554 women and 1126 men, and their deaths occurred between then and a 3-year follow-up interview in 2001. METHODS: AIDS-related deaths were identified using a standard World Health Organization (WHO) algorithm and this observed number of deaths was compared with the expected number. The expected number is derived from a comparison of adult mortality in this sample with the pre-AIDS mortality levels measured in the 1987 Malawi census, giving an 'excess mortality factor' that can be considered to be AIDS related. It was calculated that 74.9% of observed deaths in 1998-2001 would be from AIDS. RESULTS: According to the WHO algorithm, 75.5% of the VA deaths could be categorized as AIDS deaths. The gender-specific figures were 75.1% for females and 76.3% for males. CONCLUSIONS: Results are consistent with those found in studies conducted in other areas of sub-Saharan Africa at earlier stages in the epidemic. They confirm that VA data can be used to estimate with a reasonable degree of confidence the distribution of AIDS- and non-AIDS-related deaths in the aggregate, even in a rural population with relatively low levels of education.

Lateral and vertical intergenerational exchange in rural Malawi.

Contemporary empirical literature on family resource flows in developing countries focuses on vertical flows between parents and children. Using data from the 1999 Family Transfers Project in Malawi this article examines a broader set of flows between adult respondents and their surviving parents, and paternal and maternal aunts and uncles. It compares the frequency and value of material and monetary flows, and the frequency of provision of other services, among these relatives. It also explores variation on these parameters across three ethnic groups, each of which has discrete normative patterns of descent, inheritance and postmarital residential arrangements. Results suggest that: (i) intergenerational support networks in Malawi are both vertical and lateral; (ii) in their transfer relationships, working aged adults have a net loss to parents, but a net gain to uncles and aunts, implying the existence of an institutionalized network for the transfer of resources among branches of the family; and (iii) lineal structures privilege kin of certain gender for certain roles. Maternal and paternal aunts are the largest source of material transfers among the matrilineal Yao, and paternal and maternal uncles are the largest source among the patrilineal Tumbuka.