It wasn’t easy and it wasn’t fun, but I’ve compiled a spreadsheet of local government employment and payroll data from the U.S. Census Bureau’s 2007 Census of Governments in time for Governor Paterson’s emergency budget session. The attached spreadsheet has two summary tables that print in two pages; one shows the number of full time-equivalent workers in various categories (police officers, sanitation department workers) per 100,000 people in March 2007, and the second shows their level of pay at that time, relative to the national average. In each case, the summary tables provide data for all local governments in the United States; New York City; the downstate suburban counties (Nassau, Suffolk, Westchester, Rockland, Putnam); the mostly highly urbanized Upstate NY counties (Albany, Broome, Dutchess, Erie, Monroe, Niagara, Oneida, Onondaga, Orange, Rensselaer, Saratoga, and Schenectady); the other more rural Upstate New York counties; the total for New Jersey, and Fairfield County in Connecticut. In a new feature, I have put one additional column in the summary tables — “your county here.” By blocking that column, and replacing text in the existing formula with the column letter (s) for whatever county one is interested in, one can put any other county in the table. The choices are all the counties in New York and New Jersey, almost all of the counties elsewhere in the country with the most total private-sector employment in 2006, and a few others that interest me, because I know people who live there. A discussion of where the data came from and how to interpret it follows.
The U.S. Census Bureau conducts a Census of Governments every five years, as part of its economic census series, and has done so since the 1950s. Data is collected from every state and local government on its employment, payroll, revenues, expenditures and debts. The employment and payroll data will be published this fall; the financial data is promised for next summer. The home page for the 2007 Census of Governments is here. While the Bureau has not published its handy tabulations of the employment data yet, highly detailed data is out at its ftp site, which is located here.
Local governments are organized various ways; Nassau County has dozens in all the local government categories — a county government, cities, other municipalities, school districts, and other special districts. New York City really has just two; the City of New York and the Port Authority of New York and New Jersey. This makes it impossible to compare one place to another using just data for, say, cities. But the Bureau provides a convenient file that adds of the data for all the local governments within each county in the U.S. It is this data, along with the local government totals for several states and the entire United States, that is in the spreadsheet.
Except for the data for one special district — the Port Authority of New York and New Jersey. The Census Bureau assigns the entire employment and payroll of that organization to New York City local government, despite its substantial operations and customers in New Jersey. To improve comparability, I deducted one-third the agency’s Transit and Airport employment and wages, two-thirds of its Water Transportation employment and wages, and half of its Highways and Other employment and wages from New York City, and shifted it to an adjusted New Jersey state total.
Transit and public elementary and secondary school data required additional adjustments, as I discussed in my prior post on 2007 state government employment here . Public schools are quintessential local government activities, but in some states, notably New Jersey, the state has take control of some school districts. The Census Bureau counts their employment and wages as “state government” rather than “local government,” making the counties in which they are situated not comparable with counties where all schools are counted as “local government.” For the United States and New Jersey, I’ve added all state elementary and secondary school employment to the “local government” total, fixing the comparability problem at the state and national level. Note, however, that the included county level data for some counties in New Jersey, and perhaps some other counties across the country, may understate local public school employment per 100,000 residents due to state takeovers, as I was unable to allocate the state-level data to individual counties.
I did, however, do an allocation for mass transit employment in New York State, where (for historical reasons) the Bureau counts New York City Transit and the PATH system as local government agencies, and almost all the other mass transit agencies (including other parts of the MTA such as the Long Island Railroad) as “state government.” I divided up the New York “state government” figure between upstate transit agencies and downstate suburban agencies such using similar data from the National Transit Database (NTD). I’ll need to divide up the Census Bureau’s payroll data based on NTD payroll data to create a proper payroll per employee estimate for these areas, but have not yet done so. New Jersey Transit is counted as a state agency, but I made and adjustment and included it in the local government Transit total for New Jersey. U.S. state government Transit employment and payroll is also counted with local government.
The county area data file assigns each government to a county, but many large transit agencies cover more than one. Rather than dividing the employment of multi-county transit agencies by the population they are headquartered in, I have substituted in the total population of their service areas, as obtained from the NTD.
All these adjustments can be found in formulas in the spreadsheet, in one wished to know exactly what I did to modify the data as downloaded. The population on which employment per 100,000 residents is based is the Census Bureau’s estimate for July 1, 2007, excluding the transit cases as described above.
Local government employment and wages can be difficult to compare from place to place not only because of the allocation of responsibility between state and local government, but also because of the allocation between local government and the private sector. New York City’s public mass transit employment, for example, is high relative to its population because many New Yorkers substitute transit for the ownership of one or more automobiles. New York had 579 full-time-equivalent transit employees per 100,000 residents compared with the national average of just 79. But New York had just 481 private sector workers per 100,000 residents in industries that sell and service automobiles, parts and fuel, compared with a national average of 1,431. The extra public money spent on transit workers is more than offset by money saved on personal motor vehicles.
To allow the reader to take these patterns into account, I have provided Quarterly Employment and Wages data on private sector industries that either substitute for public employment (as when the local government contracts with a solid waste management firm rather than hire its own sanitation workers to pick of the trash) or are substantially funded by the federal, state and local governments (as are the health care and social services sectors and the “heavy construction” industries). The data, taken from the federal Bureau of Labor Statistics (BLS) and several state Departments of Labor, is not strictly comparable — it tabulates every worker as one rather than pro-rate part timers into full time equivalents — but provides an idea. And the information is needed as elementary and secondary education and private health care are the two largest parts of the New York State budget, and the budgets of most other states. The private-sector data is generally for the second quarter of 2007, and is only available for the broader areas in the summary table, not for individual counties elsewhere (except where provided by friendly state departments of labor and not suppressed for disclosure reasons).
Note that in many places fire protection is provided by volunteers, reducing their fire department employment per 100,000 residents compared with places with professional protection, such as New York City.
According to full-year 2007 data from the BLS, excluding the extremely high paid Finance and Insurance sector, private sector workers in Downstate New York earned 32.3% more than the national average that year, a ratio that has been fairly constant in recent decades. The cost of living is similarly high here, so it is not unreasonable for local government workers to expect above average pay in Downstate New York as well. And, in fact, the average local government worker in New York City earned 28.2% more than the average local government worker in the U.S. in March 2007; the average local government worker in the downstate suburban counties earned 33.4% more than the U.S. local government average.
That ratio, however, varied substantially for workers in different categories. It can also vary greatly over time. In March 2002 New York City’s police officers earned 50% more than the U.S. average, but only because their pay was inflated by post-9/11 overtime. In March 2007 they earned 1.6% less than the U.S. average, but only because they had not received a raise in years as part of protracted contract dispute. When the contract was settled they received back pay for March 2007, pay that is not counted here.
Bearing in mind the possibility of such temporary effects, the summary table expresses March 2007 payroll per full time equivalent employee as a percent above or below the U.S. average. The private sector averages, total and excluding the Finance and Insurance sector, are only available for individual counties in New York and New Jersey.
I’ve been compiling and looking at this data a long time, so what jumps out? Local government employment, particularly public school employment, has always been sky-high in the rural counties of Upstate New York. Particularly given that many residents of those counties don’t have public water, public sewer, public trash collection, and many communities have volunteer fire departments. It has gotten higher. How can they afford it? Perhaps they can if someone else is forced to pay for it. That might provide an understanding of why “Republican” State Senators are pushing against any reduction in state school spending. It has to do with excess jobs, not education.
Just in case anyone is interested in the facts, however, they are in the spreadsheet — for every part of the state compared with other areas, and each county of the state with a few strokes of the keys and clicks of the mouse.