This isn’t a question of fiat. Variation will happen while still complying with the letter of the law

More broadly, the results presented above highlight the importance of distinguishing between the letter of the law and the actual implementation on the ground, both for deriving theoretical predictions about the impact of policies and for testing these predictions empirically. Analyses that focus on formal statutes, policies, and regulations alone, while ignoring the significant discretion that comes into play during the implementation process, risk drawing the wrong inferences about the causes and consequences of public policy. For example, the problems of uneven implementation are generally overlooked in the expansive literature on welfare migration. Motivated by the provocative argument made by Peterson and Rom (1989), scholars have examined whether states with relatively generous welfare regimes—measured in terms of spending, eligibility requirements, and benefit coverage—become “welfare magnets” that attract impoverished migrants from stingier neighbors, generally finding no evidence of such migration. If poor people make locational choices based on differences in actual welfare outcomes rather than formal welfare rules, however, they may be reluctant to move based on statutory promises alone if they expect actual implementation to falls well short of what the laws provide. While one may interpret the evidence from existing studies to mean that poor people do not respond cross-state differences in welfare policy, the null results may also indicate that the intrastate variation in implementation significantly weakens the attraction of even the most generous statutory welfare provisions by shaping the expectations of potential beneficiaries.