The Census of Governments Has Your Number
Michael Maciag’s walk-through of this under-utilized goldmine
When poring over a city budget or a government’s website, you may have found the data to be less than adequate for the story you’re trying to tell. In many cases, a limited snapshot of a government’s finances or payroll just doesn’t provide enough context.
Perhaps you’d like to know which areas a government’s budget have grown over time. Or, if a fire chief proposes hiring more firefighters, it would be helpful to compare current staffing levels to those of other jurisdictions.
Every five years, the Census Bureau conducts a survey to help answer these and other questions. The Census of Governments covers more than 91,000 units nationwide, providing a comprehensive overview of local, state and federal government.
Most of the data for the 2012 survey was published earlier this year. Here’s an overview of the three main components of the survey:
|Employment and Payroll||Released||Counts of public employees and payroll totals for each government, by function|
|Lists and Structures of Governments||Released||Numbers and types of all governments nationwide|
|Government Finance||State government data released; finances for local governments and public schools scheduled for late 2014||Tax collections, revenues, expenditures, and other finance statistics|
Some of this same type of data is available in financial statements, such as a Comprehensive Annual Financial Report (CAFR), or you may file a records request for a government’s most current information. Where the Census of Governments proves particularly useful, though, is that it encompasses many years of data for all states and most larger local governments – all in one place in a standardized format. This makes it possible to identify historical trends and compare multiple jurisdictions, provided that you understand the limitations of the data.
In this guide, you’ll learn how to take advantage of this under-utilized resource.
Employment and Payroll
In an era of tight budgets, many state and local governments are forced to reduce expenses, and often this means trimming payrolls. In all, state and local governments shed about 600,000 jobs nationwide dating back to early on in the recession. Cuts haven’t been applied evenly, though. While some individual agencies implemented layoffs or hiring freezes, others weathered the recession much better or even expanded payrolls.
When a governor or a mayor proposes changes to public employee staffing levels, it’s a good idea to consult the Census of Governments.
Using these datasets, you can provide readers with a detailed portrait of where public employees are working and what they do. For each unit of government, the survey reports detailed figures for more than 30 different employee classification types.
You may also want to see how a state or municipality stacks up to others. On a per capita basis, some state governments employ significantly more workers than others.
You need to be careful how you compare jurisdictions, though (more on that later), and will need to do further reporting to better understand the numbers. Wyoming, for example, has the highest rates of government employment, but that’s primarily because of the large number of public hospitals in the state.
Data published in the employment and payroll section of the survey includes employee counts, number of hours worked, payroll totals, and average monthly earnings. Listed payroll amounts are snapshots for a single month and are not adjusted for inflation; March 2012 was used for the most recent survey.
For local governments, it’s a different process. Download the corresponding .ZIP file for your state under the “Local Government” heading (they’re broken into 13 parts); then open the Excel file.
In the downloaded spreadsheet, you’ll find data for all counties, municipalities and special districts within a state. You’ll probably want to filter the data. Try removing some of the top rows above the column headers and then turn on AutoFilter in Excel.
Not all local governments respond to the survey, so you’ll need to verify whether the listed counts and payroll amounts are actual reported figures or estimates that the Census Bureau computed.
For each data value in the spreadsheet, there’s a corresponding flag. Flags coded as C, R, T, U or V denote data reported by the government or agency. Totals, merely the sum of all listed categories, always have a “Z” flag. Others are imputed using a variety of methods we won’t discuss here. One of the more common flags you’ll see is “G,” referring to numbers the Census Bureau computes by applying a growth rate to prior-year data. (See this reference table for a description of each flag)
As an example, here’s employment and payroll data for Autauga County, Alabama:
|Government Name||Government Function||Full-time employees||Full-time employees flag||Full-time pay||Full-time pay flag|
|AUTAUGA||Other Government Administration||5||R||10974||R|
|AUTAUGA||Parks and Recreation||2||R||5224||C|
|AUTAUGA||Police Officers Only||28||R||79461||C|
|AUTAUGA||Solid Waste Management||2||R||4360||C|
|AUTAUGA||All Other and Unallocable||7||R||16083||C|
|AUTAUGA||Other Police Employees||2||R||5008||C|
Autauga County employs 163 full-time workers. We know the data shown refers to numbers provided by the county since all government functions listed below the total have an “R” flag. Full-time pay amounts also all represent reported data.
You’ll notice that some jurisdictions list more government functions than others. To understand what each specific category refers to, you’ll want to get acquainted with the survey’s classification manual. See chapter 12 for a definition of each category and whether it’s published at the state, local, or federal level.
When writing about public employment, it’s best to refer to the full-time equivalent (FTE) totals rather than the full or part-time figures. Smaller jurisdictions may actually employ more part-time workers. Also, as a rule of thumb, education workers generally account for about half of a state government’s total employment. Hospitals and public safety are two of the other larger segments of the workforce.
Given how large the dataset is, it may be tempting to compare totals for different jurisdictions. This is doable. But you need to be careful how you compare.
When comparing individual governments, you’ll want to know how services provided differ, along with other factors that may affect employment and payroll. Here are a few caveats:
- It’s best to compare jurisdictions of similar size.
- Departments’ responsibilities may differ greatly from jurisdiction to jurisdiction. For example, local governments perform much of the state highway maintenance in some states.
- Agencies may contract out, so consider total budgets as well.
- If education employment is not relevant, exclude it from your calculations as it accounts for a significant share of the workforce.
Counts of Government
In some areas, local officials often lament how hard it is to coordinate efforts with so many relatively small jurisdictions within a county or larger metro area. Many argue taxpayers would be better served by more sharing of services or even consolidation of entire governments.
When talks of consolidation come up in your community, Census of Governments data can illustrate just how fragmented governments are in a given region. Consider Allegheny County, Pennsylvania, for example. The county, which includes Pittsburgh and many smaller suburban jurisdictions, encompasses 260 total general-purpose governments and special districts. While that’s an extreme example, other urban areas are home to similarly high numbers of governments.
The Census of Governments publishes total counts and types of governments for states and local areas.
Select a table from the dropdown menu on the Lists & Structure of Governments page to display the latest 2012 data. Table 13 lists counts for all types of government units within each county. Tables 1, 4, and 5 show historical trends for types of government for each state.
Here’s an overview of common terminology used:
General-purpose governments: Provide general services; referring to municipalities, counties, and townships.
Special districts: Provide a single or limited number of services. Some of the more common types are sewer authorities, water districts, and fire districts.
Independent school districts: These are broken out separately from general-purpose governments and special districts.
Types of governments, particularly special districts, can vary a lot from state to state. Fortunately, there’s a document summarizing – in great detail – types of governments operating in each state.
Another possible story one could tell would examine how the number of governments has changed over time. Tallies of governments generally remain fairly stable, so you’ll likely need to go back at least a couple decades to see any notable changes. Special districts, in particular, have multiplied quite a bit in some regions.
Illinois (6,968), Pennsylvania (4,905) and Texas (4,856) recorded the most local governments in the 2012 survey. On a per capita basis, North Dakota is home to the highest tally because of the many civil townships spread across the state. In some cases, just a few families live in a single township!
The Census Bureau also publishes a series of financial datasets, assessing everything from property taxes to liquor store revenues.
Let’s say you want to write about property tax revenues. The Census figures show how the extent to which cities rely on property taxes varies greatly, from less than 10 percent to around half of total revenues for some. These revenues took a hit in recent years with the bursting of the housing bubble, so many jurisdictions are now looking to diversify revenues more.
Financial data for state and local governments is released separately. The Census Bureau published 2012 state government finances on FactFinder earlier this year. From the Advanced Search page, search for table ID FIN009. Then select your state from the Geographies menu on the left.
In the table, you’ll find the same type of figures you’d expect to see in a government’s annual budget. These include detailed revenue and expenditure totals along with short- and long-term debt.
Local government financial data from the 2012 survey is slated to be released in December. For the most up-to-date financial data for a single government, though, you’re better off referring to its latest comprehensive annual financial report (CAFR) — typically posted on a government’s website.
The Census Bureau also conducts several annual and quarterly surveys on tax collections, pension systems, and public school finances. We won’t discuss them in detail here, but this list summarizes what’s available and when it’s released.
It’s not uncommon these days for lawmakers to contend that a government has grown too large and that it’s time to axe programs. In response, an agency might argue that it has already suffered cutbacks.
To cut through the rhetoric, we can refer to historical census data. One of the strengths of the Census data is that many larger governments have reported several years of data, allowing for comparisons over time. It’s also useful in that you’re often (but not always) able to compare historical trends to those of similar jurisdictions.
As mentioned, the Census of Governments is conducted every five years, with the last two releases published for 2007 and 2012. In between these five-year periods, the federal government administers annual surveys measuring the same types of data, providing key information to better illustrate trends so that you’re not limited to every five years.
Employment, finance and most other government data is available online dating back to the early ’90s. The next round of survey results for 2013 won’t be published until next year.
Each annual survey draws on a sample from the Census of Governments including all jurisdictions meeting defined thresholds, along with randomly selected smaller governments. If you are interested in a state government or larger jurisdiction, chances are you’ll be able to find what you’re looking for.
The annual surveys on employment, payroll, and government finances include the following:
- Federal civilian employees
- All state governments
- County governments with populations of 100,000 or more
- Municipalities with populations of 75,000 or more
- Townships with populations of 50,000 or more
- Larger school districts and special districts
- Sample of other smaller units
If you’d like to see employment trends over time, there are a few different ways to pull historical data.
For state government employment data, I’ve also posted counts for all major job functions dating back to 1960.
Compiling other historical employment and payroll data for states and individual local jurisdictions is a bit more time-consuming. Start by selecting a year in the dropdown menu and then downloading the appropriate .ZIP file for your state as before. Larger governments meeting population thresholds and a sample of smaller units are included. You’ll also want to review each field’s data flag as before to verify whether it’s a reported number or estimated by the Census Bureau.
For historical financial data, refer to this Census Bureau table and select a year and state from the dropdown menu. You’ll find state government and aggregate data for localities dating back to 2004. Downloadable data for prior years is also posted separately. Individual local government financial data is not yet available online, but the Census Bureau tells me they’re working on a solution to post the data eventually.
In general, data should be comparable over time. But there are some questions you’ll want to answer first:
- Was data imputed for any years (as indicated by the flags)? If so, you’ll either want to exclude it or make a note of it.
- Did a jurisdiction’s geography remain consistent over time, or did a government merge or consolidate services? Such changes would greatly influence the reported totals.
- Did a government begin providing a new service at some point? If so, expect to see an increase in payroll and employment. Conversely, any move to privatize services should result in a dip in public employment and payrolls.
Other Data in Factfinder
Some (but not all) of the data discussed here can also be downloaded from American FactFinder . From the Topics menu on the Advanced Search page, expand the datasets and select “2012 Census of Governments.” You’ll see a list of 71 tables to browse from.
With all the datasets available, you’ll probably want to start by focusing on the one most relevant to your reporting. Be sure to read the documentation and understand the limitations of what’s published. Once you become familiar with the data, you’ll find it to be a great resource.
Mike Maciag is the data editor for Governing magazine, a policy publication covering state and local governments. He lives in Arlington, Va.