My interaction with politicos has been very limited over the years, probably to our mutual benefit. I can count the times I have been the same room with an elected official on two hands, and the times I have actually spoken with one on one hand. But about four and a half years ago, I found myself in the back of a room next to Alan Hevesi, who was then running for New York State Comptroller, and I asked him for something. Handing over a sample, I asked if he, as Comptroller, would compile and publish the kind of comparative state and local revenue, expenditure, employment and payroll data that I have been producing privately over the years (see here and here). I had previously tried to convince some of his future underlings to do the work, but had gotten nowhere, which is no surprise given my own history as a government underling and inability to accomplish anything. But here I was taking to the future top guy. The reason I wanted the Office of the State Comptroller to compile and publish the information isn’t that I didn’t feel like doing it myself. For reasons I’ll explain after the jump, that office is in a position to provide more detailed and timely data than I am. Well, despite some positive noises at the time, I never did get my data.
I get the data from the U.S. Census Bureau, which collects it from state and local governments around the country. Every five years, most recently in 2002, it surveys them all, and produces a comprehensive “census of government” with local government data summed up for every county in the United States. In other years the Bureau only contacts some local governments, not all of them, and uses this partial survey to produce estimates for local government for each state as a whole. Since the City of New York and Port Authority of New York and New Jersey are always included, I am able to download data for these entities to produce local government data for New York City, and can then get the rest of New York State by subtraction. But I cannot produce local government finance information for other parts of the state individually. Worse, to save money, the Bureau didn’t even publish local government data at the state level in 2001 and 2003, breaking a series going back to the 1950s. A national estimate for all local governments in the country is all that was made available.
It turns out, however, that for New York and other financially sophisticated states the Bureau does not, in fact, have to survey individual units of government. Instead, it gets the data for all governments on tape each and every year, in New York’s case from the Office of the State Comptroller. The Bureau chooses, in non-census years, to take a sample of this and provide more limited data to make it comparable with other states. But data could be compiled for every county, because it is available. I asked the Bureau if it would consider doing a “special tabulation” of public finance and employment data for New York State each year, if I could raise the money. The Bureau refused, lacking the staff and concerned that other states would demand equal treatment.
The Bureau did, however, offer to provide a “crosswalk” between its classification codes and those of New York State, to allow the Office of the State Comptroller to spit the comparable information out itself. This would permit an annual comparison between local government in every county of New York State and the national average, every year. Hence my request. Some time next year, the U.S. Census Bureau will release a national estimate of state and local revenues and expenditures for the United States for 2005. The Office of the State Comptroller could, if it chose, produce comparable data for every county in the state. It has made a non-decision to not do so. Instead, it continues to release less detailed, un-comparable local government finance data that excludes New York City, and is not even adjusted into per-capita or share of personal income ratios so that it makes sense compared with the national average.
Which is too bad. I’ve often said that just because it is the national average doesn’t make it right, but it does mean that substantial deviations, in either direction, should be justified. The Comptroller’s audits have tended to focus on the trivial, such as whether or not a government-financed business phone was used to call relatives, rather than the significant, such as why some places have three times the non-instructional staff in their public schools relative to population. While being something of a finding in itself, comparative data also indicates where to look – where savings could perhaps be found, and were additional resources may perhaps be required – and questions to ask.
To be fair the Comptroller, I have also asked the City of New York’s Independent Budget Office, repeatedly, to at least work with what the Census Bureau provides. I provided one of its staff members with data on revenues and expenditures for the City, and the local government total in each state, each year going back to the late 1960s, with the data in the same format classified under the same Census Bureau codes. But the IBO preferred to release data for the city alone, with no data for other places, instead. Comparing what was raised and spent last year with what was raised and spent this year, rather than a normalized comparison with other places, contains an implicit assumption that last year’s spending and expenditures are the best guide to what ought to be done this year. This locks in the winners and the losers. They have certainly been locked in in New York State. In any event, it is possible that the IBO staffer I met personally did try to make comparable information available and was thwarted by the higher ups, on the grounds that some fact might upset some interest and threaten the agency. Based on my own experiences, that wouldn’t surprise me.
To elected officials, a spreadsheet of comprehensive data in the wrong hands (such as mine) is a bad thing, because it allows the losers to be identified. And politicians represent the winners. Regardless of party, there is always a side of the story they don’t want told. The internet brought an explosion of detailed information made readily available, by organizations like the U.S. Census Bureau. More recently, the tendency over time has been for less data to be made available, rather than more. Politicians would rather produce carefully screened press releases than spreadsheets that allow people to draw their own conclusions. Or so it seems.
Some pols don’t even bother to release carefully screened and misleading data. To explain why I was sitting in a public agency doing very little that was ever actually used, I developed the “nuclear deterrence’ theory of policy research. The other side might launch a fact. So you need analysts in their silos so that you can launch a fact back. But once the facts start flying, you get mutually assured destruction, because it would soon become apparent to the outsiders that the actual deals have nothing to do with the facts, and the pols aren’t really interested in them. So the analysts stay in the silos.
Well, for the moment, Alan Hevesi is still State Comptroller. On the other hand, it is possible he will be replaced. Either way, I’m still willing to set things up with the Bureau, get those cross-walks, explain how I compile the information and why I compile it that way, and how to do it. And, as for the Spitzer Administration, I’m not looking for a job, or a favor, or a consulting fee, or anything else. If I can help in my spare time without taking too much time away from something else, I’ll do so. After all, if the OCS started compiling and releasing comprehensive, comparable public finance and employment data I’d save time in the long run. Doing so myself probably costs me about 40 hours a year, a very frustrating hobby given the bad news that inevitably awaits when the data is compiled.
In Fiscal 2004, the State of New York and its local governments spent over $2.5 billion! in the Census Bureau’s Financial Administration category (see attached spreadsheet), so isn’t it outrageous that state residents have to rely in me, in my spare time, to produce comparable data? Perhaps not so outrageous given that spending only equaled $3.36 for every $1,000 of personal income earned by a New York State resident, compared with a national average of $3.65. But hey, if they aren’t going to make comprehensive, comparable data readily available in usable form, perhaps those financial agencies deserve a misleading attack based on limited facts.