The U.S. Census Bureau released its FY 2017 public education finance data for school districts across the country on May 21st.
Of the 50 states, New York ($23,091), the District of Columbia ($21,974), Connecticut ($19,322), New Jersey ($18,920) and Vermont ($18,290) spent the most per pupil in 2017. Of the 100 largest school systems based on enrollment in the United States, the five school systems with the highest spending per pupil in 2017 were New York City School District in New York ($25,199), Boston City Schools in Massachusetts ($22,292), Baltimore City Schools in Maryland ($16,184), Montgomery County School District in Maryland ($16,109), and Howard County School District in Maryland ($15,921).
As is the case most years, I’ve downloaded this data and compiled it into tables showing per student revenues and expenditures by category for New York, New Jersey, and nearby states, four sections of New York State (New York City, the Downstate Suburbs, Upstate Urban Counties and the Rest of NY State), the U.S. average and other places that interest me. And every individual school district in New York and New Jersey. And since this is a Census of Governments year, when the Bureau will later be compiling data on every function of every state and local government in the country, I’ve done exactly the same compilation for FY 2007 and FY 1997, to make comparisons over time possible.
As is my custom, I’ve going to provide an explanation of where the data comes from and how it was compiled, and make the spreadsheets with the data, tables and charts available for download, in this post. Before providing my take on it later, after I’ve thought about it for a while. So open-minded and curious people (all three or four of you, apparently) are able to download the spreadsheets, look at the numbers and charts, and make up your own mind, rather than having me tell you what it means. Moreover, I certainly won’t be writing about every individual school district in New York and New Jersey. But the data is provided in such a way that anyone else can write about their own.
The 2017 data as released by the Census Bureau may be found here.
With prior years also available going back to 1991. The Bureau no longer produces a report on school finance, but all the tables previously published in PDF form are located in the “Summary Table” spreadsheet on the link above.
However, I download the data in the “All Data Items” spreadsheet, with information for every school district in the country, add up national, state and in New York sub-state totals, and calculate my own per student revenue and expenditure numbers.
Unlike the per student data in the Census Bureau’s “Summary Table” spreadsheet, which mostly focuses on “current expenditures,” I also include capital expenditures and interest payments. But I exclude private school payments, and payments to private charter school operators, and non-elementary and secondary school expenditures (adult education for example) from my per-student calculations –because their students are not included in the Census Bureau’s fall student count. According to the Census Bureau, charter schools chartered by state and local governments are supposed to be included in the general data, with their revenues and expenditures reported by category (teacher wages, food services, etc.) rather than in the aggregate.
Within New York State, as I define them, the “Downstate Suburbs” are Nassau, Suffolk, Westchester, Rockland and Putnam Counties. The “Upstate Urban” counties are Albany, Broome (Binghamton), Duchess, Erie (Buffalo), Monroe (Rochester), Niagara, Oneida (Utica), Onondaga (Syracuse), Orange, Rensselaer (Troy), Saratoga and Schenectady. The “Rest of New York State” is mostly rural and small town, but includes small cities such as Elmira, Jamestown, Cortland, Oneonta, Amsterdam, etc.
The data for individual school districts in New York and New Jersey are in alphabetical order by counties, except that in New York State I put adjusted figures for the Downstate Suburbs first, and the Bureau puts the big-city districts next. Aside from “dependent” school districts in those cities, the Berne Knox-Westerlo Central School District in Albany County is the first listed in New York State.
The spreadsheet with data for all areas for FY 2017 is here.
My per-student revenue and expenditure data for each area are to the left, with the data as downloaded from the Bureau on the right, staring in column AD.
Apparently not everyone in high-spending New York wants people to see just how much is being spent on public schools here, and the fact that I make this information available makes many people displeased. Many of those same people are apparently also displeased with Governor Cuomo.
Ever wonder where all that school spending goes? Parents, students and taxpayers in Troy, Schenectady and several other Capital Region districts can get a detailed idea about expenditures with the release this week of the state Budget Division’s School Funding Transparency site. It allows the public to see and compare funding on a school-to-school basis for 76 districts.
Eventually, cost comparisons for schools within all 674 of the state’s districts will be online at the site at
The data release, however, isn’t without controversy. Senate Democrats, like much of the education establishment in New York, continue to oppose the mandated reporting and are calling for its abolition as part of this year’s budget negotiations.
They are against telling the serfs just how much they are paying. What does that tells you? I’d like to see a summary of the data at the District level, compared with the U.S. average and the average for nearby states, handed out to every parent and teacher, and put on a postcard and mailed to every taxpayer in the state. The data in my compilation would make that very easy. For New York City, the postcard sent to everyone might read something like this:
In FY 2017, according to the U.S. Census Bureau, the New York City School District spent $28,495 per student. That was higher than the average for the Downstate NY suburbs at $28,027 per student, and far higher than the average of $21,033 for New Jersey, $21,095 for Connecticut, or $18,047 for Massachusetts. These are all high wage, education-oriented states on the Northeast Corridor. Downstate NY is a relatively high-wage area, but even adjusting New York City per student spending down for this factor down to $22,386 per student, it is still 57.6% higher than the U.S. average of $14,202 per student. On instructional (ie. teachers) wages and salaries alone, the New York City School District spent $16,998 per student, which is $339,960 per 20 students or $203,976 per 12 students.
How about it Governor?
Notes the Census Bureau’s press release:
These statistics are not adjusted for cost of living differences between geographic areas.
As noted above, however, for some parts of the high wage high cost Northeast I do make such an adjustment.
The cost of living adjustments are based on the average private sector wage, excluding the overpaid finance and insurance sector, in Downstate New York (NYC and the suburbs), New Jersey, Connecticut and Massachusetts. In 2017, according to Employment and Wages data from the Bureau of Labor Statistics, the payroll per private sector worker in Downstate New York (finance excluded) was 27.3% above the U.S. average. Since 1/1.273 is 0.786, adjusted figures for New York City and school districts the Downstate Suburbs are multiplied by 0.786 to make them comparable with the U.S. average, and with Upstate NY (where the mean private sector pay per worker is well lower than the U.S. average).
In the spreadsheet an asterisk (*) is used to identify states, areas and school districts where per student revenue and expenditure data that has been adjusted downward for a high average private sector wage and therefore presumably a high cost of living. Unadjusted figures are also provided for these areas. (Thus far I haven’t adjusted any per student spending figures higher in areas where the general pay level is lower).
Using mean private sector wage and salary payroll per worker to make a cost of living adjustment provides an extremely generous downward adjustment for Downstate New York, including New York City. The data includes all the highest paid private sector wage and salary workers outside of finance – those who work for Google, doctors, laywers, CEOs and other management. That pulls up the average. Census Bureau data on median earnings – what a typical person gets – shows the gap between New York City and the U.S. average is much smaller than 27.3%, in part because an increasing share of private sector workers – especially in later-born generations – have been forced in contingent, day laborer-type contract and freelance status.
According to the American Community Survey, the median worker residing in New York City earned just 4.9% more than the U.S. average in 2017. Thos with graduate degrees earned 8.6% more here, those with just college degrees 15.6% more, those with some college or an associates degree just 2.5% more, and those with a high school degree or less earned less than the U.S. average. And yet my adjustment implies that these workers should pay 27.3% more than the U.S. average for schools just for those working in the schools to be treated fairly. The anecdotal evidence implies very little concern with fairness in the other direction.
On the other hand, the cost of living adjustment I use has gone down over time, because even among private sector wage and salary workers, those in the Tri-State area are not as well off relative to the U.S. average as they once were. In 2007, those in Downstate NY were paid 32.3% more than the U.S. average, not 27.3%. The drops are even more dramatic for New Jersey – from 27.3% above the U.S. average in 1997 to just 14.0% in 2017, and for Connecticut – from 25.0% to just 12.0%. The pay of Massachusetts private sector workers is about as high as it had been, relative to the U.S. average.
One reason that Vermont’s statewide school spending per pupil is so high is that there are many school districts with just a handful of children, or none at all, that still have expenditures. As I measure it, Vermont districts that are kept in place despite having fewer than 500 students averaged $29,832 spent per student (including capital and interest on debt) in FY 2017, compared with $19,862 for those with 500 students or more. I use the latter figures, for school districts with 500 or more students, in summary tables and charts that include Vermont.
One can also find, however, school districts within New York State that also have very few students and very high per student expenditures. These tend to be located in rural areas with large numbers of second homes. The New York school district with the highest expenditures per student is Kiryas Joel at $290,143. That district’s issues have certainly been discussed enough in the media for me to leave it aside. Second is the Fire Island Union Free school district at $229,160 per student. It has just 25 students. Some high cost NY school districts apparently have been merged since 1997, but not many.
The all-in U.S. average elementary and secondary public education expenditures per student was $14,202 in FY 2017, as noted. That includes operating expenses, capital expenditures, and interest on debts. The lowest spending school district in New York State was General Brown Central School at $14,788, followed by Victor Central School District at $15,181. Victor was actually below the U.S average some years back, but not anymore. With a downward adjustment for the higher general wage level/cost of living downstate, however, one New York State school district did come in below the U.S. average – Franklin Square Unified in Nassau County at $13,673.
Here are the identical spreadsheets for FY 2007 and FY 1997.
The data from column AD and over to the right is just as I downloaded it from the Census Bureau, with no adjustment into today’s money. But all the per-student revenue and expenditure data have been adjusted upward for inflation to make them comparable with FY 2017. You can see how most of the cells are multiplied by the inflation adjustment factor in cell B3, which is based on the Consumer Price Index.
Bear in mind that in my data and charts the increases you see over time From FY 1997 to FY 2007 and from FY 2007 to FY 2017 are after adjustment for inflation– in an era in which most workers’ pay has been falling behind inflation. When you see just how much public school spending per student has gone up in New York City, you might find that hard to believe.
A spreadsheet with summary tables for selected areas, for FY 2017, inflation adjusted data for FY 2007 and FY 1997, and the dollar and percent changes between those decades, is here.
Since my focus is New York City, and I won’t be writing much about other individual school districts in the state (just regions), much of what I write in future posts is in this set of tables. Although it is worth noting that New York City’s school spending per student for FY 2017, at $28,495, was once again higher than Darien ($22,201), Westport ($24,432) and Greenwich ($25,304) Connecticut. Albeit still lower than Chappaqua, NY ($31,712).
Here is another version, with data items limited to per student federal revenues, state revenues, total expenditures, instructional (teacher) wages and salaries, instructional employee benefits, and non-instructional spending, and with data for the U.S., different regions of NY state, selected other states and districts, and all the individual districts in downstate NY after adjustment for the cost of living — for FY 1997 in $2017, FY 2007 in $2017, and FY 2017.
The data shows stunning inflation-adjusted increases in per student school spending in NYC at time when most workers saw their compensation fall behind inflation — and even Wall Street pay, though excessive, is down from what it had been. Total inflation adjusted spending per student doubled in NYC from FY 1997, when the Campaign for Fiscal Equity lawsuit was starting, to FY 2017. Can you think of anything else whose real cost has doubled, any other large group of people other than founders of social media companies whose compensation has doubled? The real increase of 99.1% for NYC total spending per student was an increase of $14,183 — about the same as the total expenditures per student in the U.S. as a whole in FY 2017, at $14,202.
Real state aid per student for the NYC schools increased by 112.1% over 20 years, more than doubling. Real NYC instructional wages and salaries per student increased by 81.9%, nearly doubling. Real NYC instructional employee benefits per student increased by 227.3%, nearly quadrupling. Four times as much, after adjustment for inflation. Real NYC non-instructional expenditures per student increased by 85.8%, nearly doubling.
The one or more posts that I write using this data will be illustrated by a large number of charts. My experience, according to what WordPress tells me, is that relatively few people download the spreadsheets and look at the tables. But lots and lots of people download individual charts. That’s why I take the trouble to include them, but they are a lot of work. If there are typos I may make corrections later, and I might also include a few extra charts as I think of them. But otherwise you don’t need to wait a week or two to see most of the charts I will use. You can see them by clicking on the tabs at the bottom of this spreadsheet.
As for the financial management of America’s schools, it is worth noting that for public education every year is a Census of Governments year, with data submitted to and tabulated by the Census Bureau for every school district in the country. And that data is in vastly greater detail than any other government function is required to provide. Vastly greater detail, in fact, than I even see necessary – there are plenty of columns I deleted, especially with many, many specific categories of federal, state and local government revenues. And the education finance data is more timely than the data for other government functions. We’ll be lucky to get the government finances data from the Census of Governments for FY 2017 before the end of 2019, and finalized data may not be available until some time in 2020.
In the end, however, the data is only as good as the people submitting it are honest. I’ve explained in other posts my concern that the New York City schools, now that their per student expenditures have rocketed to the moon, might be tempted to under-report or mis-report certain categories of expenditures, notably retirement benefits. I compared this year’s data with last-year’s data to see if that might be happening, and found no evidence. But I’d like it to be known that someone is, in fact, checking.
So please download the tables and look at them, while I take a week or two to try to come up with something I didn’t already say about this data last year
Or the year before.