Census Bureau Public Employment Data: Public School Employment and Payroll in March 2016 and March 2006

In the wake of the Great Recession, the number of people working in public education fell relative to the overall population in much of the country, and inflation-adjusted teacher pay fell in many parts of it.  There was only a partial reversal from March 2014 to March 2016, and it was mostly in places where school staffing and pay were already relatively high.

Several factors contributed to the trend.   First, in some states were the tax burden was already low relative to the personal income of state residents, it was reduced further by anti-tax ideologues.  The result was reductions in public services for the non-elderly across the board, including, in the end, elementary and secondary education. Second, with the large Baby Boom Echo (aka millennial) generation exiting school, and smaller generations entering school, the actual need for school workers fell.   But in some places with high and rising taxes the demand for school jobs increased, as other high-compensation alternatives for politically influential college graduates diminished.   Third, the cost of retired public employees soared, as a result of past taxpayer underfunding in low-tax right-wing states.  And as a result of retroactive pension increases scored by powerful public employee unions in (if by “right-wing” people mean having public policy favor the otherwise advantaged at the expense of those with less power) the other type of right-wing, high-tax states, those generally self-described a “progressive.”  A review of the data follows.

This is the second post in a series on state and local government employment and payroll in March 2006 and March 2016.   The first post, which describes where the data comes from and how it was compiled, is here.


The spreadsheet, with tables and all the charts that will be used in this and other posts, is once again here.

Local-State Gov Emp 2016

Let’s start with local government elementary and secondary school instructional employees, which in this data set includes not only teachers but also other “pedagogical” employees such as principals, superintendents, and guidance counselors.


For the U.S. as a whole, the number of such local government employees fell from 1,547 per 100,000 people in March 2006 to 1,451 in March 2014, before edging up to 1,458 in March 2016.   For New York City there was a steeper decline from 1,471 per 100,000 city residents in 2006 to 1,352 in 2014, followed by a bigger rebound to 1,424 in 2016.

The fact that New York City’s local government employment is below the U.S. average in this category is misleading, as will be discussed later.   What is not misleading is the fact that instructional employment in the Rest of New York State was already much higher than the U.S. average at 1,890 per 100,000 residents in 2006, and yet it sill increased to 1,915 in 2014 and 1,953 in 2016, during a period of falling enrollment.  One finds the same trend in Connecticut, where instructional employment barely fell from 2006 to 2014 and then increased to more than 2,000 per 100,000 state residents in 2016.   Last I checked (2015 data – 2016 in a month or so) New Jersey’s enrollment decrease was less, as a percent, than the Rest of New York State or Connecticut, but its instructional employment per 100,000 state residents fell from 1,970 in 2006 to 1,886 in 2014 to 1,778 in 2016.


One finds a similar trend for non-instructional local government employees, such as custodians, school bus drivers, and cafeteria workers.  For the U.S. as a whole their number fell from 680 per 100,000 residents in March 2006 to 626 in March 2014, and was about the same in March 2016.   New York City was much lower at 337 per 100,000 city residents in 2016, down from 389 in 2006.  But NYC contracts out school bus service and food service, and those jobs therefore don’t show up as local government employment.  Expenditures data (including those contracts) shows cost of these services in NYC to be very high for school buses and average for food service, respectively, adjusted for the higher average wage/cost of living here.   NYC’s cost of custodial services is also extremely high; those services are also contracted out in some buildings.

The number of non-instructional staff in the Rest of New York State was very high, at 873 per 100,000 area residents in 2016, but it is down from 942 in 2006.   Perhaps because schools elsewhere are becoming more efficient, or perhaps because the non-instructional unions lack the political power of the New York State United Teachers, and the political influence of administrators.  Non-instructional employment equaled a below-average 557 per 100,000 state residents in New Jersey in 2016, down from 656 in 2006. Connecticut went the other way, with an increase from a below-average 587 to an above average 656.

New York City’s relatively low instructional employment per 100,000 residents is misleading because the city’s number of public school children per 100,000 residents is also relatively low, for two reasons.

First, the city has relatively few children.   Children age 5 to 17 accounted for about 16.8% of the U.S. population in 2016, according to American Community Survey data, but just 14.7% of the NYC population.  Many NYC parents move to the suburbs when their children reach school age, due both to the city’s historically poor public schools and its relatively small housing units.  Young adults without school age children, meanwhile, account for a relatively large share of NYC residents.

Second, New York City’s children are more likely to attend private schools.


U.S. private school employment per 100,000 residents increased from 204 in 2006 to 243 in 2016.   In New York City, however it spiked from an already high 412 to 623.  It is possible that city pre-K contracts for Mayor DeBlasio’s “pre-K for all” program account for much of this increase, though the “charter school” movement also continues to grow – even as the Catholic School system shrinks.

The Rest of New York State and New Jersey are somewhat above the U.S. average in private school employment, but like the U.S. only had a moderate increase from 2006 to 2016.   But Connecticut was double the U.S. average at 500 private school employees per 100,000 residents in 2016, up from 432 in 2006.  Children ages 5 to 17 account for a below average 16.3% of that state’s population.


I grabbed some FY 2016 public school enrollment data from the New York State Department of Education to measure the more relevant ratio of instructional employment to public school students, rather than instructional employment to the overall population.   It was just 8.4 students per instructional employee in New York City in 2016, down from 8.5 in 2006 and the same as the 8.4 in the Rest of New York State in 2006.   New York City had 9.4 students per instructional employee in 2014, well below the U.S. average found when I did this analysis last, but evidently not low enough for those working in the schools.   Meanwhile, in the Rest of New York State the number of students per instructional employee fell to just 7.8 in 2014 and just 7.4 in 2016.

We’ll get a 2016 update for the U.S., New Jersey and Connecticut when the Census Bureau’s public school finance data, with enrollment data for school districts across the country for each year, comes out in a month or so.

I get it.  In Upstate New York and Connecticut the population is stagnant, the young people are moving away, and school enrollment is shrinking.  You have college graduates but fewer college-level jobs with above average compensation.  Shrinking enrollment would mean more college graduates lost their jobs, instead of more being hired.  So politicians respond.

After all the hostility to “welfare” for those on the bottom, however, we find a more lucrative and growing welfare system for those in the upper middle class (and even, with the financial bailouts in 2008 and zero interest rates for years after, those at the top).   But something is lost and someone else is hurt, now or in the future, to pay for it.  A jobs program admitted to be such would never pay that much.  People in jobs programs are generally not better off than people who are made to sacrifice more to pay for them.


Speaking of pay, New York City may have the highest state and local government tax burden in the country, low population states with extensive mineral tax revenues aside, but its public employee cash payisn’t really out of line.  Because cash pay is what everyone can see and easily compare with themselves, New York’s public employee unions seek to keep it as low as possible – particularly for new hires.   In March 2006, NYC’s payroll per instructional employee elementary and secondary school employee was 25.9% above the U.S. average, down from 34.2% above average in March 2014 (when fewer new teachers had been hired to bring down the average) but up from 20.6% above average in March 2006.

As a comparison, the mean payroll per private sector worker in Downstate New York (excluding the overpaid Finance and Insurance sector) was 28.0% above the U.S. average in 2016, comparable to the 25.9% above average for NYC instructional workers.

There are two caveats.   First, in March 2016 NYC teachers were toward the end of a long period of zero or low pay increases, other than the automatic increases that come with seniority, a period that ended with a huge pay increase at the end of their contract, which I believe is this year.  So NYC teachers are likely earning much more relative to the U.S. average right now than this data shows.

Second, New York City’s private sector mean pay is pulled by very high pay at the top, even outside the Finance and Insurance sector.  I did up a quick spreadsheet of American Community Survey data in 2006 and 2016 on medianearnings per worker.

Median Earnings NY-US-2006-2016

It showed that the median worker living in New York City (public and private) earned just 8.6% more than the U.S. average in 2016, with the median worker with a bachelor degree earning 13.9% more and the median worker with a graduate degree earning 8.1% more.   NYC females, however, earned 18.5% more than the U.S. average overall, 21.1% more for those with bachelor’s degrees, and 17.2% more for those with graduate degrees, closer to the 25.9% above average payroll per worker for NYC instructional employees.

Adjusted for inflation, from March 2006 to March 2016 the mean payroll per instructional employee decreased 0.6% for the U.S. but increased 3.8% in New York City – despite the big deferred pay increase that came afterward.  For all workers, from 2006 to 2016 the median earnings per worker decreased 0.9% in the United States and decreased 1.1% in New York City – which frankly looks a lot better than it did comparing 2005 with 2015.

Looked at another way, the 2016 median pay per worker for those living in New York City in 2016, including those working part time, was $58,853 for those with a bachelor’s degree, and $75,808 for those with a graduate or professional degree.  The mean March 2016 payroll per NYC local government elementary and secondary employee of $5,891, when multiplied by 12, equals $70,692 – but with a big raise coming, and summers off.  Most NYC teachers have a graduate or professional degree, but it is the one that takes the least amount of time to get.

It all looks about right, on average.  But the United Federation of Teachers makes sure teachers working with more difficult children in more difficult schools, and with more in-demand abilities such as math and science, are not paid more, and new hires are paid as little as possible relative to those cashing in and moving out to Florida.  And New York State residents are being sued for cheating those who work in the NYC schools out of $billions more.

The payroll per employee data is much less useful for the Rest of New York State in non-Census of Governments years than it is in Census years, such as FY 2017, because data can’t be calculated separately for the high cost of living Downstate Suburbs and the low cost of living (except heat and taxes) Upstate New York.   Even including Upstate New York, however, mean March payroll per instructional employee was up 4.9% from 2006 to 2016 in the Rest of New York State as a whole, to a level that was 34.1% higher than the U.S. average – and about the same as New York City.   New Jersey’s mean payroll per instructional worker increased just 13.0%, but to a level that was 46.5% above the U.S. average.   But Connecticut’s mean payroll per instructional employee decreased 3.0% to a level that was just 13.5% above the U.S. average.  Connecticut, it seems, was padding employment, not pay.

While the mean payroll per instructional employee fell just 0.6% in the U.S. as a whole, similar to the 0.9% in median earnings for all workers, a richer deal for teachers in places such as New York was balanced by a worse situation in other states.  In some of those states, teachers have gone on strike this year.  But in all states, however, a rising share of school spending, and other spending on state and local government services, is going to the retired.  According to the Associated Press


Among the biggest reasons for lagging pay is one of the least understood: The rising cost of state pensions.

Public pension systems nationwide face record levels of debt, totaling $1.4 trillion, according to a recent Pew Charitable Trusts study. That puts downward pressure on wages and benefit checks as governments struggle to close the funding gap. It suggests the recent outcry over teacher pay could spread in coming years, whether pension costs are widely acknowledged as a driving factor or not.

“I think what you see happening in the state and local and municipal sector is it has now become very, very clear how expensive defined benefit plans are. I think we’re headed for a big crisis across the country,” said Olivia Mitchell, executive director of the Pension Research Council at the University of Pennsylvania. “Pensions are now becoming the tail that wags the government dog, if you will.”

In Colorado, school district payments to the public pension fund have roughly doubled since 2006, to 20 percent of payroll. That has squeezed budgets when the state also was cutting funding during the recession.

The equivalent figure is 40 percent of payroll for pensions for NYC teachers.  And that is not enough to pay for past pension increases, let alone those the state legislature is likely to pass in the future.

For Colorado teachers, the pension’s role in school funding struggles adds insult to injury.

Additional contributions are not padding the benefits of current teachers. The bulk of the money that school districts spend on the Public Employees’ Retirement Association is paying off $32 billion in benefits previously promised to public-sector retirees but never properly funded.  Meanwhile, teachers are being asked to work longer hours at lower pay than their predecessors.

Current teachers indirectly paying for their predecessors’ retirement is not unique to Colorado. A 2016 study by Bellwether Education Partners found that $14 of every $20 that school districts contribute to the pension of each educator nationally are paying off unfunded debts, with just $6 going to the worker’s retirement.

Colorado teachers don’t get Social Security.  Neither do teachers in California, Illinois, and many other states.  New York teachers get Social Security as well.

In Colorado, public employee pension increase is taxable under the state income tax in excess of $20,000 for those ages 55 to 65 and $24,000 for those 65 and over, the same exclusion as for private sector pensions.  In New York State, all public employee pension income is fully exempt from state and local income taxes, no matter how young a worker collects, no matter how high pension or total income is. Private sector retirees get an exclusion of $20,000 after age 59.

Click to access taxonpensions2011.pdf

It is because the exclusion of all public employee pension income from taxable income is a constitutional provision in New York State that New York’s public employee unions went all out to prevent a constitutional convention from happening.  Remember those bumper stickers – Vote No – Taxes!   If you are a private sector worker who voted no because you were worried about your taxes, the joke was on you (and believe me, they are snickering).

In FY 2014 Colorado had a below-average state and local government tax burdens as a percent of its residents’ personal income, at 8.7%, about the same as in FY 2004.  New York City’s burden was 15.8% of its residents’ personal income in FY 2014, up from 14.0% in 2014.  The national average was 9.9% and 10.0% for the two years.


Meanwhile, the mean payroll per New York City non-instructional elementary and secondary school employee fell 18.6%, adjusted for inflation, from March 2006 to March 2016.  From a level that was 61.4% above the U.S. average to a level that was just 27.8% above the U.S. average — comparable to the percent above average the U.S. for NYC instructional employees, and non-Wall Street private sector workers.

There is no discussion I can recall in the press or politics associated with this change.  The one thing I could think of is the reverse – Mayor DeBlasio’s policy of reversing Mayor Bloomberg’s attempt to reduce the cost of school bus services and custodial services, each of which were sky-high.  But the school bus drivers are not government employees in any event, and do not affect this data.  When something like this occurs, it is often the result of a hiring spree, with an increased number of low-earning new hires bringing down the average.   But the number of NYC non-instructional employees decreased from 31,091 in March 2006 to 28,780 in March 2016.

In the Rest of New York State, the mean March payroll per non-instructional employee increased from 9.8% above the U.S. average in 2006 to 19.6% above average in 2016.   The increase in New Jersey was from 23.8% above average to 52.3% above average, while Connecticut was little changed relative to the U.S. average, at 23.0% higher in 2016.

To cut the unpaid workload I had intended to combine elementary and secondary school employment and payroll data with higher education employment data in a single, shorter post.  But this post is already too long.  A separate post on higher education will follow.   But remember, by downloading the spreadsheet, one could see the tables and charts right now.