The State Networks dataset is a compilation of many state-to-state relational variables, including measures of shared borders, travel and trade between states, and demographic characteristics of state populations. The 2,550 units in the dataset are dyadic state-pairs (e.g., Alabama–Alaska, Alabama–Arizona, Alabama–Arkansas, and so on, for each state plus the District of Columbia). The data were collected from multiple sources and incorporate measures of similarity drawn from data in the Correlates of State Policy Project.
This dataset is the first version of a collection of important variables that are commonly employed in state-networks research. The dataset includes directional ties where available and appropriate. All of these variables are directed inward toward the State1 variable. For example, the IncomingFlights variable is the number of flights from State2 with a destination in State1. Therefore, each pair of states is listed twice in the dataset, as the dyad “Alabama and Alaska” has different values than the dyad “Alaska and Alabama.” In addition to these and other inherently relational variables, the dataset contains variables based on similarities (or differences) in state attributes. Some similarity measures represent directional differences, with negative values indicating that State2 has a larger value for the variable than State1. Other similarity measures, such as RaceDif and ReligDif, are the total absolute value of the differences between states for multiple variables (for example, RaceDif is the absolute value of the sum of LatinxDif, WhiteDif, BlackDif, AsianDif, and NativeDif). In these cases, smaller values indicate greater similarity between State1 and State2. With few exceptions, the dataset comprises the most recently available information for each variable. The IRS_Migration and Income variables represent totals from the years 1993 to 2010. The policy_diffusion_tie variable aggregates total ties during the period from 1960 to 2015.
The wide range of variables compiled in this dataset allows for numerous comparisons and opportunities to study the economic and political relationships between the states. For example, if researchers wanted to identify the factors associated with the number of policy diffusion ties from State2 to State1 (defined in the dataset as the total number of times over the 55-year period that State1 used State2 as a policy source), they could begin by exploring similarities in the racial and religious composition of the states’ populations, in their levels of economic and social liberalism, and the extent of their trade relationship.
The number of policy diffusion ties from one state to any other ranges from 0 to 56. Michigan, for instance, has most frequently used California, New Jersey, New York, and Wisconsin as policy sources, while Delaware, Maryland, Ohio, and South Carolina have most frequently used Michigan as a policy source. The state most frequently used overall, California, has a total of 1,731 ties to the other states, while the state least frequently used as a policy source is Alaska, with a total of 347 ties for the period.
The results of a basic statistical model indicate that larger differences in liberalism between two states are strongly and negatively associated with policy diffusion ties, meaning that states with larger differences seem to use each other as policy sources less frequently. The results are similar when we consider differences in racial and religious composition between two states—again, larger differences are associated with fewer policy ties (though the association is weaker for racial-group-size differences than religious ones). The Imports variable, or the annual value of trade from State2 to State1, does not appear to be related to the number of policy ties.
This simple example highlights just a few of the research possibilities for policy scholars and practitioners. Policy diffusion ties may also be related to other economic indicators or measures of state partisanship.
Financial support for compiling this database was provided by Michigan State University’s Institute for Public Policy and Social Research (IPPSR) (East Lansing, Michigan), as a component of the Correlates of State Policy Project.
Olson, Shayla F. 2019. State Networks Database v.1.0. East Lansing, MI: Institute for Public Policy and Social Research (IPPSR).
Jordan, Marty P. and Matt Grossmann. 2017. The Correlates of State Policy Project v.2.1. East Lansing, MI: Institute for Public Policy and Social Research (IPPSR).