State governments face an incentive to reduce their welfare program expenditures because overly generous welfare programs would potentially create unsustainable spending. States are likely to engage in a “race to the bottom” by not adopting generous program benefits so as to avoid becoming a “welfare magnet”. However, simply not adopting a new policy does not represent an actual “bottom” of the race, but the ultimate bottom of the race is termination of existing policies and programs. This proposition is an open one: no prior empirical work has examined the “diffusion of termination” yet because termination cases are rare.

Since the implementation of Temporary Assistance for Needy Families (TANF) program in 1996, the General Assistance program is the last resort of government support for those who are not eligible for the benefit of TANF or Supplemental Security Income – which presents a challenge to opponents of General Assistance program. As a result, elimination of a state General Assistance program would create harm for one of the most vulnerable populations – and this creates a counter-incentive to termination.

This study seeks to explore why the diffusion of General Assistance program termination has become more prevalent in the last two decades. I will examine the factors contributing to the diffusion of General Assistance program termination in this study. The diffusion of termination of the General Assistance program across states from 1990 to 2011 is analyzed through a modification of the lens of the diffusion of innovation framework.

Hypotheses for this study are grouped into three different levels of diffusion determinants: internal determinants, horizontal diffusion, and vertical influence. For internal determinants, financial imperative, political ideology, and expansion of alternative program are canvassed; for horizontal diffusion, program termination in the neighboring states and in the ideologically similar states are examined; and vertical influence would be measured by federal aid for state welfare programs. Event history analysis is employed to test these hypotheses. Collectively, how each of these determinants affects the diffusion of General Assistance program will provide critical insights into the diffusion of program termination.