Does Bureaucratic Embeddedness Improve Legibility? Evidence from Peru

• In Event: Ethnicity, State Capacity, and Democracy in Historical Latin America

Abstract

Recent scholarship on state capacity has highlighted the critical role of the state’s knowledge about its citizens in a wide array of governance processes, ranging from conscription to welfare provision. The predominant view, however, has studied legibility --the production of standardized data captured in censuses, birth certificates, IDs, and property registries-- as an attribute of a monolithic state. This perspective overlooks how data-gathering activities depend on a variety of routines implemented by an army of street-level enumerators and clerks, who, in countries without civil service systems, might exhibit uneven characteristics.
This paper focuses on one of these characteristics: bureaucratic embeddedness. It argues that legibility projects, designed to codify and standardize population, depend on the informal routines of street-level workers embedded in the communities in which they work. For this purpose, I focus on social registries in Peru. These administrative tools assess the socio-material status of rural and indigenous households, thereby determining their eligibility for inclusion in welfare programs. I argue that local officials with strong ties to the community have low information costs: they know their neighbors’ idiosyncrasies, local migration patterns, and geographic constraints. Based on this knowledge, officials can circumvent administrative challenges, develop informal and creative practices, and use them to collect people’s information. These practices contrast with generic and abstract directives issued by the central state. The outcome is a more effective extraction of population data.
Quantitative and qualitative evidence supports these claims. To test the effect of embeddedness on legibility, I leverage two administrative datasets of local bureaucrats and a matching design. To measure embeddedness, I use an indicator that reflects whether a frontline worker’s surname is exclusive to the district. For a variable that reflects informational capacity, I collect individual information on data-entry errors made by the bureaucrats when gathering data in the terrain. A set of matching variables at the district levels helps construct comparison groups. Individual level variables, such as education, salaries, and experience, are also added as control. Qualitative data, complementarily, sheds light on the causal mechanisms that connect local knowledge with informal practices of legibility, including what we call “mundane” creativity: bureaucratic practices that deal with unexpected situations of the implementation of legibility projects.
This paper makes some contributions. It frames information capacity as an administrative process rather than an indicator of state capacity, as recently conceptualized in the literature. In this vein, we revisit a central concern of Scott’s framework: how state engineering plans depend on, or even parasite, informal practices built by local actors –in our study, street-level officials. Secondly, the paper discusses an unexplored connection between legibility and bureaucratic embeddedness. It empirically shows how embeddedness is a source of effective practices that ease the implementation of legibility. By showing this effect, the paper demonstrates that in certain scenarios -i.e., legibility projects designed to incorporate millions of rural dwellers into the welfare system- state legibility plans can circumvent the frequently assumed conflict between standardized orders and local execution.

Author

• Luis García, Temple University