There are core datasets that both state and local governments should provide. These are pivotal datasets, the equivalent of connective tissue. Without these datasets, it is difficult to use that governments' open data in order to accomplish basic task. There are other datasets that are also very important, but these are the datasets that unlock the potential of many of those other datasets.
State Core Datasets
These are the essential datasets for state governments to provide.
Often held by the Secretary of State's office, these record every business that exists now and, preferably, those that no longer exist. This is pivotal data, because in many states, corporations can own land, lobby for and against legislation, be parties to lawsuits, contribute to political campaigns, and much more. Without these records, it's impossible to do anything with big swaths of those data sources. Records should include the corporation's name, date of creation, address, status, officers, major shareholders, and registered agent, at a minimum.
Legislation and Legislators
Held by the legislature, this is a record of the legislative activities of the state. This is necessary to understand campaign finance data and election results. It should include every bill that’s before the legislature, along bills from past session, with appropriate metadata: title, sponsor, date introduced, a history of actions taken on the bill, vote tallies, the text of the bill. Also, it should include every legislator that is now in office or that was in office during any past sessions for which bills are provided, providing each legislator's name, committee memberships, contact information, dates in office, and district.
Often held by the legislature. Law data is neccessary to connect regulations to enabling laws, to connect court rulings to anything at all, and to understand legislative data. It should include the section identifier, title (if applicable), the text of the law, and the statutory history of the law.
Every address in the state, and its latitude and longitude. This is essential for geocoding (the ability to know where an address is located on a map). The state's GIS agency frequently keeps these records. This is often part of E-911 records.
Who gives money to whom, and how that money is spent. This data is in a great deal of demand, because it's widely regarded as revealing. Without this data, it's difficult to understand much about legislator data, or to see the role that corporations play in public policy. This must be atomized to a per-transaction level, and updated as often as reports are filed.
Municipal Core Datasets
The boundaries of every lot in the municipality, who owns it, when it has changed hands, how much it was sold for, and its appraised value.
Every address in the municipality, and its latitude and longitude. This is essential for geocoding (the ability to know where an address is located on a map). The state's GIS agency frequently keeps these records. This is often part of E-911 records.