Stephen B. Holt

Associate Professor of Public Management, University at Albany, SUNY

In this class, there will be two elements that will be important for you to develop a good research question: data on the outcome you think will be affected by a policy and data on the states or cities that implemented that policy. Below are datasets that can be helpful for the former. For the latter, you will have to do a little research on the policy you are investigating.

IPUMS Data Center - Census Microdata

IPUMS is a data cleaning project that creates linkages across annual datasets for several extant annual data collected by public agencies. This link takes you to Census and American Community survey data. This data includes demographic, economic, occupational, commuting, and health data. The data includes identifiers for metropolitan statistical areas, counties, and states (which is important for linking to policy data). This data is good for examining policies that affect housing (both access to housing and things like affordability), transit, employment (overall and of specific groups of people), and immigrants.

Urban Institute Education Data Portal

For anyone interested in working with K-12 education policy issues or higher education policy issues, this is a starting point resource. It includes data on things like school staffing and enrollment, district finances (both revenues by source and expenditures by type), biennial data on AP course taking, SAT/ACT taking, and bullying/harrassment exposure. In the realm of higher education, there is IPEDS, which includes data on all colleges on things like staffing, enrollments, majors, and financial aid and the college report card data which includes earnings data from graduates of various colleges.

NYC Open Data

For anyone exploring city or state policies, the NYC Open Data portal is almost surely going to have data worth examining. Anything from transit routes and on-time rates to crime data to police calls-for service to school data, nearly every public service in NYC reports data to the portal. I strongly suggest students planning to work with this data let me know their policy topic and outcome of interest so that I can confirm the data is available and usable for the project. The upshot of this portal is there is a lot of data; the downside is it can be messy and is not always capturing what you need it to.

American Time Use Survey - IPUMS Version

The ATUS collects time-diary data from a nationally representative sample of Americans that captures how Americans spent their time in a typical day. The data is linked to the Current Population Survey for the respondents which includes information on housing, income, demographics, education, family structure and occupations. This data is good for looking at policies that might affect time dedicated to work versus leisure, time dedicated to studying or taking classes, time spent with kids, and time spent commuting to and from work and other places. I have used this data to look at gender gaps in time spent studying, sector differences in time spent volunteering, and socioeconomic gaps in time spent waiting for service. State-level policies like paternal/maternal leave, unemployment insurance, and perhaps union status might be good policies to use in conjunction with this data.

National Health Interview Survey

The NHIS provides annual data from a national sample on both physical and mental health. This data is excellent for people interested in capturing state policies that target either of those health outcomes both overall and for specific communities (e.g., immigrant populations). Some good questions would be related to whether expanded eligibility or access to public insurance or public health programs improve health outcomes for those newly made eligible. Some other creative uses might be looking at changes in mental or physical health as a spillover from other policies - such as access to college or sanctuary from deportation.

Performance Monitoring for Action

Health data from countries with high fertility rates - typically developing nations in east Asia and Africa. Good for looks at policy changes that target environmental issues in the included countries; however, I caution that projects tackling questions in this area will need a clear statement of the policy change that is affecting countries within the sample.

National Incident-Based Reporting System

NIBRS is an FBI-led effort at collecting national crime data in a systematic and comparable way. Local law enforcement agencies opt-in to reporting crime incidents to the database in a standardized way and on a regular schedule. Because the database is incident reports, it includes a rich set of data on crime with a lot of information about a given incident (characteristics of those involved, the type of incident, the location, the outcome, etc.). Working with this data can be a little complicated so if you are working on a policy question where crime is the outcome of interest, please reach out and I will help you get started on this dataset.

Below are some repeat links and some links to additional data on things like government finances and so on. I strongly suggest students focus on a state or local government policy change that allows the use of the data sets above because they are more user friendly and will be good for learning how to work with data in this context.