At a recent meeting at the Child Care and Early Education Policy Research Consortium (an annual meeting run by the US Department of Health and Human Services), issues around the implementation of childcare subsidy programs occupied a group of social science researchers to good effect. It is certainly true that implementation issues are one of the least understood and the least studied aspects of human service policy. It has been getting some attention of late, as that discussion demonstrated, but it’s a thorny issue that deserves much more attention than it has received.

It was Michael Lipsky of MIT who coined the phrase “street-level bureaucracy” in his book of the same name in 1980.While childcare subsidy case managers have a certain amount of discretion in how they interpret and enforce program policy, the level of scrutiny of their actions certainly does not compare with that exercised over other street-level bureaucrats, such as police officers, or even teachers. Childcare subsidy case managers’ latitude in decision making informing and shaping attitudes about the program (street-level PR) is no doubt quite large. Probably larger still is the power of local managers to emphasize, for example, their choice as to whether a hard line of enforcement (calculated to minimize potential fraud on the part of heads of household whom they mistrust) or a softer stance (favoring the child’s needs for care, even at the chance of missing some fraudulent behavior) should be applied.

To cite another example, in a study of the behavior of cash welfare policy implementation in three counties in Michigan, Norma Riccucci found that “the discretionary power of street-level bureaucrats is an important determinant in the implementation of public policy . . . .” Furthermore, “Ultimately, street-level bureaucrats in the welfare offices may not be implementing the goals intended by state-level officials. Moreover, each office may be unique in terms of the delivery of TANF [Temporary Assistance for Needy Families program] services and benefits at the front lines, despite the fact that Michigan’s welfare system is highly centralized.”[1] Front line workers responded to surveys indicating large differences from management in perceptions of the most important goals of their program. They also differed from staff in other offices sampled in the study.

Note that Riccucci measured attitudes with questionnaire responses, not behavior. Yet another level of complexity involves what the front line workers are actually doing, as well as thinking.

How would we even begin to find out? Could we expect to get a true sense of the detail of local operations through a questionnaire sent to managers? Well, many difficulties appear. Does the manager know what his workers are doing? Some do, no doubt, but how could you tell the well informed from the ignorant? Why would they even bother to answer the questionnaire, if there wasn’t a level of trust? They might well wonder whether dire effects on their salary might follow. Could the researcher cultivate a handful of trusted local managers as informants? That may work, at least in some situations, but the number of such cultivated contacts must necessarily be small. How would you ever be sure of getting a good sampling of opinion? What about going directly to the front line workers? The attitudes and behaviors of case managers, or social workers are even more obscure. Metrics based on data collected for other purposes, such as eligibility data or voucher payment data, can show patterns, but such patterns, without explanation, must remain inconclusive.

Blame human nature if you like, the territorial instinct, passive aggression, perhaps, and the desire for ownership of an important function as well, but the fact is, the question of front line worker behavior is a very important one.

[1] Norma Riccucci, “Street-Level Bureaucrats and Intrastate Variation in the Implementation of Temporary Assistance for Needy Families Policies” in Journal of Public Administration, Research and Theory, 15 (1), Jan 2005.