Comptroller Liu’s Pension Report: Optimism, Misdirection and Disaster

NYC Comptroller John Liu recently released a report and started an initiative to assure everyone that New York City’s public employee pension funds are fine. “Retirement Security NYC is a major initiative launched by Comptroller John C. Liu to protect the retirement security of public employees while ensuring the City's financial health,” with the retirement security portion intended to gain the support of public employee unions in a campaign for Mayor and the financial health portion intended to assure the city’s bondholders and wealthy and business taxpayers there is a limit to the extent to which their ox will be gored. The latest report is called Sustainable or Not: Pension Cost Projections Through 2060.

I have read the report, and have several problems with it. First, Liu piles up one rosy assumption after another. Next, his claims in big print in the front of the report, and in press releases, are much rosier than the detailed tables in the back of the report, even given the rosy assumptions. Third, those detailed tables lack the details required to figure out how the numbers were calculated. Finally, Liu endorses the Generation Greed position that it is perfectly wonderful if retroactive pension increases for those cashing in and moving out are offset by lower pay and benefits for future public employees. I get the feeling that the only way these reports can achieve their political objective is if no one actually bothers to read them, and write a post like this one. Or nobody bothers to read this post. My objections are detailed after the break.

First the assumptions, on page 2. “Three scenarios projected employer and City contribution costs using an 8.0 percent, 7.5 percent, and 7.0 percent rate of return, with modest growth in the active membership of the pension funds. There is also a fourth scenario that held active membership constant with an 8.0 percent rate of return. Unless otherwise specified, discussions in the report refer to projections based on an 8.0 percent rate of return with a 0.25 percent annual rate of growth in pension fund active membership. This is referred to as the ‘baseline’ scenario in the report text.”

Now I’ve argued in detail that assuming an 8.0% rate of return from today’s inflated asset values is wrong. And despite recent declines, asset values are still inflated, though they are less inflated than in 2007 when they were less inflated that in 2000. I won’t rehash this – you can read about it here. But I will say this – none of Liu’s alternative scenarios is as troubling as the most likely. That, temporary booms and busts aside, there will be little positive return on investments for several years, perhaps another decade, until the level of debt in the U.S. economy and executive pay fall, and interest rates and dividend yields rise, to more normal levels. And then there will be better returns from that point forward. Which may not matter to the NYC public employee pension funds, because no matter how high the rate of return is, you don’t get any earnings if you don’t have any money left.

On page 2: “The compounded annual average rate of return on the assets of the five pension funds combined was 9.37 percent for the 30-year period FY 1981 – FY 2010.”

That sure sounds like more than 8.0%, doesn’t it. But there are a couple big problems with it. First, on page 2 of his previous report dated April 6th, Comptroller Liu claimed “by far the largest contributor” to soaring pension costs for New York City “was lower investment returns, which accounted for 48 percent of the additional costs over the period. It added $3.1 billion to costs in FY ‘10, and accounted for $15.2 billion over the decade.”

Now think about what that means. From 1981 to 2000, the city’s pension funds assumed the rate of return would be 7.0%. This was changed to 8.0% in 2000. And during this period as a whole, when those public employees who retired recently or will soon retire were working, the pension investment returns were higher than assumed at 9.37 percent. Not lower. So how can low investment returns be the cause of the soaring pension burden, as claimed in one report, if investment returns were not low, as claimed in another report?

The answer is that low investment returns are not the reason public services are being gutted to pay for soaring pension costs. Retroactive pension enhancements, and past inadequate funding by politicians, are the to reasons for our diminished future.

Which leads by to my second problem with Liu’s statement about 1981 to 2010 returns. The years he selected to make the claim that an 8.0% return assumption is not excessive were years of excess returns, excess returns that left asset values inflated, leading to lower returns going forward. The period from 1982 to 2000 is what is known as a “secular bull market,” although there were cycles within it. In secular bull markets stock prices gradually inflate above their long-term value, and investor enthusiasm allows executives and Wall Street to rip them off. In secular bear markets stock prices fall back to, or below, their real value. The last one was 1966 to 1982, after the “nifty fifty” stock market bubble. That was 18 years. Japanese stocks, in a secular bear market, are far lower than they had been in 1994, a total of 17 years ago.

Stocks don’t crash to fair value all at once. There are too many suckers, and public employee pension funds, still buying – even with a dividend yield below 2.0%, and the federal government has sought to re-inflate bubbles with cheap money to preserve the assets of the affluent. Thus we had another run up, associated with the housing bubble, in 2007. Before stocks could get back to fair value in the subsequent crash, zero percent interest rates temporarily puffed them back up to their current levels. But zero percent interest rates won’t last forever.

Rather than measure long term returns from secular bull market peak-to-peak or bust-to-bust, as I had suggested in my analysis of pension returns, Liu has chosen an era that is mostly bubble. It isn’t quite as bad as using 1981 to 2000 or 1981 to 2007, but it is close. And this isn’t the first time Liu’s reports have played with years. New York’s public employee pensions are based on the wages earned in the last year before retirement. Police officers and firefighters are eligible to retire after 20 years of work. So do they work more overtime in year 20 and later, when more overtime might pad their pensions, than in prior years, when their pensions would be unaffected? Liu set out to pretend to answer that question while avoiding it. 

On page 23: “Uniformed police with 15-17 years of service worked an average of 251 overtime hours in FY 2010, while those with 18-20 years of service worked an average of 281 hours, a difference of about 11 percent. For firefighters, the difference was smaller; 253 compared to 239 overtime hours, or 6 percent.” Wait a minute, aren’t those the wrong years? Right. Why weren’t years 1 to 19 compared with years 20 and later? Because, I would guess, instead of a 6 or 11 percent increase, the difference might have been greater.

Back to the assumptions in the latest report. On page 4. “The long-term rate of municipal employee general wage growth was assumed to be 3.0 percent annually, of which 2.5 percentage points was assumed to be inflation and 0.5 percentage points real growth.”

The richer future workers are assumed to be relative to retirees, and the bigger the city’s budget is, the less the future burden of pensions relative to other things. And that is what Liu wants to show. Not only does Liu assume wage gains for public employees in excess of the rate of inflation, but he also assumes the number of public employees will either increase or, at worst, remain stable.

But as I have shown in other posts, based on NYS Department of Labor data, New York City’s local government employment is lower than it was in 1990. This is not just a change because the city is currently in a fiscal crisis, but the latest employment peaks have been lower than the ones before over the years, and the latest employment troughs have been lower than ones before as well. Cycle aside, NYC public employment has been going down. Liu is assuming this trend will reverse. Starting now, as layoffs are debated.

What has also been going down is the average earnings of U.S. workers, adjusted for inflation. It may not be so observable within the city’s boundaries, because better off people have been moving in compared with those dying off or moving out, the opposite of the 1950s and 1960s. But it is true nationally. And with public services re-collapsing to pay for pensions over the next decade, the better off may stop moving in.

It may not be unreasonable to assume that New York’s public employee unions will use their power to get richer in cash pay relative to inflation, even while everyone else gets poorer. But it is unreasonable to assume that at the same time, those poorer serfs will be able to afford an equal or rising number of better off government workers. Only one of those two things will be true, unless neither is.

With these rosy assumptions in hand, Liu asserts that pension costs will fall in the far off future, as a percent of the wages and salaries of active public employees. According to page 8 of the latest report, this is due in large part to “the phasing in of new pension plans, which are generally less favorable for employees and less expensive for the City. Some of these plans have been in place for some time, such as the ‘Tier 4, 57/5 Plan’ into which new civilian employees have been enrolled since June 29, 1995. This pension plan is less expensive to the City than earlier plans, primarily because of the higher rate of employee contribution required.”

Not by my calculations. In July 2010, I created a series of simple a one-employee pension system models to show what the city’s pension promises would have cost if enough money had been put in all along, for an employee who worked the minimum number of years, retired at the minimum age, and lived to age 80. The models were for a general city employee or teacher, a worker who gets to retire sooner because he or she has a physically demanding job such as a sanitation worker, and a police officer or firefighter. And they were for the pensions as originally promised to those approaching retirement or recently retired, and for the pensions as modified by some of the major retroactive pension enhancements I’ve been aware of. You can see my calculations in the spreadsheet attached to this post.

The 1995 deal allowed public employees to work for 25 years instead of 30 and retire at 57 instead of 62, a huge increase in costs on both ends – fewer years worked, more years paid to do nothing in retirement. But it also required employees to pay 1.85% more into the pension plans out of their own pay, with existing workers expected to “buy back” the years they had not contributed. Of course the city could not “earn back” the years of returns on that money after the fact, but let’s leave that aside.

Based on a lower assumed rate of return, I had calculated that the original 30/62 pension most NYC public employees had been promised when hired, if the right amount of money had been put in all along and the pensions were not in the hole to start with, would have cost 11.8% of payroll, with 3.0% put in by the employee and 8.8% by the taxpayer. But the 25/57, with the extra 1.85% put in from the start, would cost 20.4% of payroll. Even with the employee kicking in more, the amount the taxpayer would need to contribute over a career was 46.3% higher, and that doesn’t even include the added cost of retiree health insurance for eight years before age 65 when Medicare picks up some of the burden, rather than three.

Might Liu’s higher assumed investment explain the difference? I redid part of the spreadsheet shifting to an 8.0% assumed return. That spreadsheet is attached to this post. And I find that the cost of the 25/57 pension is 13.8% of payroll with the higher assumed rate of return, with 4.85% for the employee and 8.95% for the taxpayer. But the cost of the original 30/62 pension is just 7.7% of payroll if one assumes an 8.0% investment return, with 3.0% for the employee and 4.7% by the taxpayer. So the 25/57 pension, even with the extra worker contribution still costs more that 30/62. In fact the increase in the taxpayer’s cost is greater at 58.5% if the rate of return is assumed to be greater, because of five fewer years of investment returns at the higher rate.

In stating that the 1995 25/57 pension deal “saved money,” Liu is just repeating what former Mayor Giulani, who cut the deal (and presumably the city’s pension contributions as well based on those savings), said at the time. I simply assumed it was true then, because I hadn’t done the calculations myself, but now I know differently. Worse, the 1995 deal allowed thousands of city workers – including many teachers – to retire early without buying back the extra 1.85%. No additional money was kicked in to pay for that, either.

Looking at my spreadsheets, you can see the money paid into the pension each year, the rate of return, the money paid out, and the rate of return. You can see the formulas to figure out how each number was calculated. You can modify the assumptions, as I did to increase the expected rate of return and inflation rate, and balanced this by reducing the required taxpayer contribution as a percentage of wages and salaries.

The full actuarial model behind Liu’s report would scale up these numbers for tens of thousands of workers, with some dying sooner and some later, some leaving before the number of years required for a full pension and others working longer. Not all that information could be in a printed report. But there is no way to see the background data that I can find. And there is a column for money paid in every year, there is no column with the amount of money paid out every year, to understand where the balance is coming from. I couldn’t figure out what was going on.

I used the same assumptions for the originally promised 30/62 pensions and the 25/57 pension as upgraded in 1995, and found the latter cost more. If Liu used the same assumptions for each, he should have found the same.

Also on page 8, “more financially significant changes have been implemented for the Police and Fire pension plans in recent years. Since July 1, 2009, new uniformed Police and Fire Department hires are required to join a ‘Tier 3’ pension plan. Among the significant changes from the earlier Police and Fire pension plans, Tier 3 extends the years of service necessary to receive a 50 percent-of-salary pension benefit, changes the determination of final pensionable salary, and reduces disability benefits.”

So when was the Tier 3 pension plan enacted? In 2009? Wrong. In the 1970s.

It was always there, but legislation had been repeatedly passed to allow police officers and firefighters to join the more lucrative Tier 2. But in 2009, then Governor Paterson vetoed the legislation. What are the odds that police officers and firefighters will have Tier 2 retroactively reinstated in the future, perhaps in the run up to an election in exchange for campaign contributions and political support? Wouldn’t Liu himself agree to support such a change in exchange for an endorsement in a run for Mayor? You can bet on it.

We have had one retroactive pension increase after another, yet Liu’s analysis includes no money set aside for any in the future. Bu the fact is that at any moment, New York’s public employee pension plans can be retroactively enhanced by the state legislature, which is a geriocracy, and immediately become an irrevocable vested right no matter how severe the consequences. That right can never be taken back. It could happen tomorrow. It could have happened a 3 am last night. And for that reason, we can’t wait until some time after 2060 for our pension plans to get out of the hole.

What, however, if the police and fire pensions are not changed between now and the year 2060? According to Table 2 on page 9, the cost of Tier 3 pensions to the city (and thus the value to the police officers and firefighters) about 13.0% to 15.5% lower, as a percent of payroll, for those hired after 2009 that for those hired before. My figures for what those pensions cost, for both Tier 2 and Tier 3, would be much higher in each case, but let’s use Liu’s figures for a moment.

I don’t think the Tier 3 pensions are unfair compared with what most people get, as they allow a retirement with a full pension after 25 years of work (up from 20) at any age. But they are unfair compared with those who came before, who had their pensions retroactively increased through various automatic disability pensions, automatic cost of living increases, and pensions based on the last year’s pay rather than the last three, making pension spiking more likely. Why is it fair for future police officers and firefighters to be paid 13.0% to 15.5% less than those who came before? Are they expected to be less qualified? To do a less good job? Or are they just not as politically powerful?

My view is that if the value of pensions for future police officers and firefighters is lower by 13.0% to 15.5% of pay, then their cash pay should be 13.0% to 15.5% higher. But the trend has been the opposite, with the starting cash pay of future police officers and fighters far lower than those who came before, adjusted for inflation. To pay for a better deal for those cashing in and moving out. In all Mayoral administrations (but with the same unions). At one point the starting pay of NYC police officers and firefighters was cut 40.0% (Bloomberg, you idiot), and the city had to fight the unions to get that partially reversed. For most NYC workers, those covered by DC37, starting cash pay was cut by 15.0% compared with those cashing in and moving out. For teachers, there was a 6.0% cut in relative starting cash pay in a mid-2000s contract. Liu seems on board with all of this. I, of course, am not.

“Changes have also been made to teachers’ pension plans in recent years,” Liu noted on page 10 “Effective December 11, 2009, new teachers entering the Teachers’ Retirement System are required to contribute 4.85 percent of salary until they have 27 years of credited service, and 1.85 percent thereafter. This lowers the City’s costs, as 62/5 members were only required to contribute 3.0 percent of salary for the first ten years of service.”

Well I don’t know what Liu is comparing the latest teacher deal with. Looking at my spreadsheet, in the “enhanced” tab, following the 1995 deal the teachers got a deal in 2000 to cut their pension contributions to zero after ten years, and to add an inflation increase to the pensions even in years with no inflation. Instead of being 58.5% more expensive for the taxpayer than what was originally promised, now the pension was 114% more expensive, even without the option early retirement. In part because the employee contribution fell by more than three quarters, since they only contributed in the low earning early years.

For those who got to walk out the door immediately when 25/55 was offered in 2008, the cost of the teacher pension to the taxpayer was 149% more expensive that what they had been promised. But that is only if 2 ½ times as much money had been contributed all along. It was not. Now that money has to be made up, for all the 25 years it wasn’t there earning investment returns.

Under the latest revision Liu points to, future teachers will be required to pay in five times as much over the course of their careers to retire at age 57. The taxpayer cost is, in fact, less compared with the 25/55 deal those cashing in and moving out. But even the current pension plan for future teachers, if we can ever afford hire more, would still cost 70.5% more than the pensions today’s recent and near retirees were originally promised. Even with Liu’s higher rate of return. When talking about the recent deal saving money, of course, Liu is just repeating what Mayor Bloomberg said. But by the time HE said it, I already knew it wasn’t true.

On page 3, “Pension costs as a percent of the City’s budget will continue to run above typical historical levels until about 2022, at which point they will return to levels below those of the early 1980s.”

What, I may ask, is historically typical about the early 1980s? That was the worst – or at least the worst before the next decade to two. As those who have read my reports know, I have tabulated data on NYC revenues and expenditures, and comparable figures for the rest of New York State and the U.S. average, as far back as the Census Bureau has collected the data. For taxpayer pension contributions as a share of public employee wages and salaries, I have put the data in a chart in the attached spreadsheet “Census NYC Debts and Pensions.” And the data shows that the city’s pension contributions had soared to nearly 30.0% of payroll by the early 1980s, following the 1968 Lindsay Tier I retroactive pension enhancements and the secular bear market of 1966 to 1982.

And public services were gutted to pay for this. Does anyone remember what they were like? How are taxpayer pension contributions at nearly 30.0% of payroll “historically typical,” as Liu claims on page 3, when for the most part the city’s pension contributions have been far lower – and U.S. average taxpayer contributions to public employee pensions have consistently varied between 8.0% and 12.0% of payroll? As the chart shows.

But Liu doesn’t make the case that pensions actually cost what they did in the early 1980s. He argues that they in fact cost far less, in big bold letters.

Look at Table 2 on page 9. The “normal” taxpayer cost of (for example) the teacher’s retirement plan, if the plan wasn’t already in the hole due to the failure of the stock market bubble to inflate indefinately, is just 9.0% for females hired who got the pre-2009 plan and 6.79% for those hired under the new plan. That’s all it costs, say’s Liu, who earlier had called the early 1980s burden historically typical.

“As more senior employees enrolled in earlier pension plans retire and are replaced with employees enrolled in the newer plans, the real cost of pensions to the City will decline…If no further adjustments to pension benefits are made, other assumptions are accurate, and investment returns are equal to the assumed rate (in Chart 4, 8.0 percent), the gap between contributions as a percentage of salary and the entry-age normal rate will narrow.”

This table is the most dishonest thing in Liu’s report. Not just because it relies on a series of perfect world assumptions. But because it conflicts with the detailed tables the actuary has put in the back of the report, in small print.

According to the table on page 40, which is the “baseline” scenario for teachers, the required taxpayer contribution for NYC teachers would still be above 20.0% of payroll in the year 2040. With all the rosy assumptions. That’s 29 years from now, and given that most NYC teachers had the opportunity to get a deal that allows them to retire after 25 years, it is theoretically possible that every teacher with the favorable deal would have retired by that time. In the front of the report, Liu claims the pensions of those who got the enriched deals cost less than 10.0% of payroll, but in the back the cost is still over 20.0% nearly 30 years from now.

Moreover, in the small print table on page 40 the required taxpayer contribution is still 14.6% in the year 2060, more than double the requirement in Liu’s boldfaced table on page 9. By that year not only would all of today’s NYC teachers retired, but many if not most of their replacements would have retired too. So how much to those pensions really cost? When will the costs in Liu’s boldfaced table be the actual cost, under his own model? In the year 2100, if all of his rosy assumptions come true and things go well?

Does Liu really claim that taxpayer pension costs would still be double his boldfaced claim nearly 50 years from now — because of stock market decreases in 2008 and 2009 — even though the rate of return from 1981 to 2010 all told was above average? Come on!

Thus Liu’s statements are fundamentally misleading compared with each other, and not just compared with what a fair minded observer would claim. After nearly five decades of a combination of higher taxes and reduced public services, he reports, taxpayers and public service recipients would still be worse off, based on the fine print, than the report and press releases try to make it appear. And that’s under optimistic assumptions. Things will be very, very bad for a long time, and at the time that today’s five year olds are age 54 years old, they’ll still be bad. All so those who already had the longest and most generous retirements could become even better off. But that isn’t what Comptroller Liu, who is seeking their political support, has decided to say.

But what about the question posted in Comptroller Liu’s latest report: Sustainable or Not? A better question is, “sustainable at what price?” New York City’s state and local tax burden, as a percent of its residents’ personal income, would be the highest in the nation if it were a separate state – except for Alaska and Wyoming were most of the taxes are on oil, gas and coal revenues. As the city found out in the 1970s, at some point if the gap between its tax burden and other places is wide enough. the affluent and businesses will flee, shrinking the tax base and leaving the city with less revenue than before. Similarly, high crime rates and garbage in the streets in major business locations, such as Manhattan, and affluent neighborhoods might also send the tax base running out of town.

But both the executive class and the political class, Liu’s base of support, can tolerate the collapse of public services otherwise. The rich get corporate parking, limousines and taxis, while public employee unions drive to free parking spots reserved for them by placards; neither has to care about the subway. The rich can send their kids to private schools, the public employees can live in the suburbs, and the politicians can get their kids into the few NYC public schools that are any good. None have to care about the rest of the public schools. Those with good health insurance don’t have to go to public hospitals. The rich can fund the public parks in their areas directly through their donations and fees.

In theory, as long as the privileged are kept safe and happy and not taxed beyond the tipping point, the rest of the city’s budget could be used for pensions, retiree health insurance, interest on the debts, and sinecures for the politically connected. If no public services are required for anyone else, then of course that’s “sustainable.” And that’s where we are heading.

Think about it. After the debts and pension deals of the Lindsay era, the era the city’s public employee unions have recreated through the state legislature and a couple of Bloomberg and Giuliani deals, the city’s library system was gutted. Finally, more than 30 years later, five day per week library service was restored. It lasted, what, two years?

Now Comptroller Liu makes the case that everything is fine because the city’s budget might have almost recovered by the year 2060, if everything goes well. Who knows? Maybe there will be five day per week library service in 2070, if there is not a single retroactive pension deal between now and then. Maybe they’ll start talking about spending more in the classroom in the New York City’s public schools in 69 years. But really, so what?