If you want to convince Arizonans to reform PSPRS, you don't go to New York

With thanks to a colleague who informed me of this article, Coaxing Fire and Police Staffs in Arizona to Cut Own Pensions, by Ken Belson in the August 11, 2014 New York Times, I was hopeful that there was going to be some new ideas from Professional Fire Fighters of Arizona (PFFA) President Bryan Jeffries, who recently took over for Tim Hill.  Unfortunately, the article was a disappointment.

The article details Mr. Jeffries' attempt to sell the PFFA reform plan, discussed in several recent posts, as a noble concession by Arizona's current firefighters to help PSPRS, and by extension, taxpayers and future firefighters.  The article has errors*, and the writer does not seem to be very familiar with PSPRS or its recent travails.  I am not sure how the New York Times got involved in this or why they found it newsworthy, but the Arizona Republic, which has done excellent reporting on PSPRS, would have certainly been more accurate and probably given the claims of Mr. Jeffries and others quoted in the article a little more scrutiny.

For me the key passage in the article is this:

To put the plan into effect, Mr. Jeffries wants to change the Constitution to allow for this one-time fix. This would reassure workers that lawmakers could not make even more drastic changes later.

I do not know if the second sentence is a paraphrased quote from Mr. Jeffries or inferred from the desire to embed the PFFA reform plan in the Arizona Constitution.  Either way, this statement rather boldly declares that the PFFA strategy is basically an end run around the Arizona legislature when it comes to PSPRS.  It is arrogant enough for the PFFA to present a final reform plan to the legislature as if it was the perfect solution.  It is another thing to go to the New York Times, hardly an Arizona-friendly paper, portray your organization as a protector of Arizona's taxpayers, and give the impression that you plan to treat the Arizona legislature as a non-entity in PSPRS reform.

As I detailed in recent post the PFFA reform proposal is no good.  I do not know how involved the PFFA was in the drafting of SB 1609 three years ago, but their public position was that no reform was necessary at the time.  Now they are all-in for reform but think that the legislature should just rubber-stamp their proposal, even though it is no improvement over the Arizona legislature's own bill, SB 1609.  So what was Mr. Jeffries' goal in talking to the New York Times?  I do not know, but if he thought a write-up in an elitist east coast paper would help PFFA's case, I think he is badly mistaken.

*PSPRS only takes half of any returns over 9% and places them in the COLA fund, and only part of SB 1609 was overturned, not the entire law.

A modest proposal for PSPRS COLA reform

I have spent several posts criticizing the Professional Fire Fighters of Arizona (PFFA) pension reform proposal, and I have pointed out the fundamental flaws with both SB 1609 and the current excess earnings model in how they calculate COLA's, especially when we consider the possibility of inflation.  (And, once again, for consistency I will again use the acronym COLA throughout instead of the less familiar but more accurate term permanent benefit increase (PBI).)  So what would I propose to do about COLA's?

I think the solution is simple and does not involve any significant change to PSPRS.  It would simply be to eliminate COLA formulas and allow the PSPRS Board of Trustees ("the Board") to determine the next year's COLA based on input from its actuary.  This is such a basic idea that I find it hard to believe that I am the first to propose it, so if someone has already floated this idea in the past, I apologize.

The main opposition to this idea would be that the Board could not be trusted to award fair and appropriate COLA's.  The Board might favor employers (and taxpayers) over employees and retirees, or vice-versa.  In its current statutorily mandated makeup, the Board is made up of two employee representatives--one fire and one law enforcement, two employer representatives--one state and one local, an elected official or judge, and two citizen representative, all appointed by the governor.  As can be seen, the Board is constituted to have representation from all stakeholders in PSPRS: employees, employers, politicians or judges, and taxpayers.  This allows for a diversity of opinion and interests and forces some give-and-take from Board members.  Furthermore, the Board is already responsible for other more critical decisions, particularly how and where funds are invested, so entrusting them to decide COLA's seems reasonable.

The Board would also be required to take input from its actuary, whom they already rely on for the other critical decisions they make.  The actuary does the number-crunching, simulations, and projections that determine funding ratios and annual required contributions.  Within a certain range based on the actuarial calculations, the Board could be restricted in how large or small a COLA could be awarded, and this is where the give-and-take among the Board members could take place and would not allow them to award COLA's based on unrealistic assessments or personal bias.

If we take a historical perspective, this method would have worked fine in the past.  When PSPRS was overfunded and returns were high, COLA's would have been more generous (even exceeding the rate of inflation) and allowed the benefits of excess earnings to be distributed equitably among retirees, current employees, and employers.  However, as PSPRS became more and more underfunded and returns over time were not sufficient to recoup past losses, COLA's could have been reduced or eliminated.  Looking ahead, if PSPRS begins to recover and inflation becomes a problem, COLA's could surpass the 4% or 2% thresholds in place or proposed.

This method of awarding COLA's would avoid the long-term mistakes engendered by hard formulas put into law or, even worse, into the Arizona Constitution.  COLA's would be based on the current financial condition, not from when times were better or worse.  Unions that negotiate for employees have a vested interest in having wages and benefits determined by hard formulas.  These hard formulas create certainty and stability in their contracts that do not allow the negotiated terms to be changed arbitrarily by elected officials.  This works fine with a collective bargaining agreement (CBA) that may last only one or two years, but the use of hard formulas becomes a problem when you are talking 20 or 30 years.  Inflation and returns fluctuate over time and setting in stone a COLA formula based on only your most recent experiences is short-sighted and dangerous.  Would unions accept a 20-year CBA based on the economic conditions that exist now?  Of course not, so why would they think that locking down a permanent, unchangeable COLA formula is a good idea?

The PFFA proposal is bad, and I am thankful that the Governor did not call a special session to consider it.  It is a shame that PSPRS wasted money on an actuarial analysis of such an absurd proposal.  I know that there are multiple labor organizations representing Arizona law enforcement, and I was not sure their feeling about this proposal.  However, a letter copied in Tucson City Council Member Steve Kozachik's Ward 6  June 18, 2014 newsletter indicates support for the proposal from at least one law enforcement organization, the Arizona Lodge of the Fraternal Order of Police (FOP).  The letter says, "Those of us who have worked on the issue believe the Firefighter's proposal has merit and may protect our members from more draconian measures.  This proactive effort represents the best chance for public safety to protect important benefits, our employers, our pension fund, our retirees, our actives, and our future members."  It is disheartening to know they were on board with the PFFA proposal.

While the PFFA pension reform proposal looks dead for now, it was important to analyze it closely to show its flaws.  Even if it never rises again, the PFFA proposal served a useful purpose in exposing the chronic myopia that afflicts the PFFA leadership and, possibly, law enforcement union leadership as well, when it comes to PSPRS.  For someone who is already in the DROP or close to retiring, dedicating 4% of your future career earnings to retirees may not seem like a sacrifice.  If you have many years left to retire or are just starting your public safety career, it will greatly affect the financial well-being of you and your family.  For someone who is expecting a six-figure DROP payout, committing to 2% COLA's in the future may not seem like a hardship.  For someone who retired before the DROP went into effect, a 2% COLA could mean your retirement benefit could get eaten away by inflation.  Looking at pension reform only from the perspective of someone with 20-30 years of time in PSPRS and access to the DROP is not only financially perilous, it is grossly unfair.  Unless the PFFA leadership can start looking at the cost and benefits of reform from the perspective of all affected constituencies, no reform effort proposed by them should ever be taken seriously.

PSPRS and its broken COLA system may be worse than you think

In recent posts I discussed the Professional Fire Fighters of Arizona (PFFA) reform plan to PSPRS.  If you have read those posts, it is clear that the PFFA plan is a non-starter which provides zero improvement over the reforms implemented by SB 1609.  However, the cost of living allowance (COLA) reforms of SB 1609 have been reversed for retirees and are virtually certain to be reversed for all PSPRS members hired before 2012.  This leaves us with an unsustainable COLA system that will remain in place for decades.  So what is to be done about COLA's?

A Little Backstory

Obviously, this is a difficult question to answer, so the best place to probably start would be at the beginning.  I have mentioned this several times before, but we have to remember that what we are discussing are not really cost of living allowances.  PSPRS confuses the issue by interchanging the term COLA and permanent benefit increase (PBI) in its documentation.  PSPRS does not increase retiree benefits based on any measure of inflation like the consumer price index (CPI), which tracks the prices of a basket of goods, to raise Social Security benefits annually.  The PBI awarded by PSPRS is based strictly on a highly flawed excess earnings model, and recent PBI's have been much more generous than Social Security.  This can be seen in the post here that looks at PBI's since 2000.  Another way that PSPRS differs from Social Security is that PSPRS awards a flat monthly dollar amount not a percentage of the existing benefit.  This helps those with lower retirement benefits as it represents a higher percentage increase than those with larger benefits.  As a final note, the Arizona Revised Statutes also do not mention the term COLA and uses the term "permanent increase in the base benefit."

The flawed excess earnings PBI model was based on an even more flawed and misguided notions. The excess earnings model could only have been designed and advocated by those who believed that PSPRS could never become underfunded, that it could never suffer large investments losses, or that the bill for systemic underfunding of benefits through bad policies and programs would never come due.  I suppose these notions came about from an extended bull market from the early 1980's until about 2000, a textbook case of financial short-sightedness.  This was all capped off with the passage of Proposition 100, which added a "pension protection clause" to the Arizona Constitution, in 1998, and the flawed excess earnings model became permanent.

What we have seen over the past 13 years is an erosion of PSPRS' aggregate funding ratio from a high of 126.9% in fiscal year 2000-01 to 57.1% in fiscal year 2012-2013.  This erosion has been steady and unrelenting yet, except for the brief interlude after SB 1609's passage and the resolution of the Fields case, PBI's continued to be paid, and with the Fields decision, retroactive PBI's have recently been paid out.  The most powerful and most expedient way for PSPRS to increase its funding ratio is through the compounding of market returns.  Legend has it that no less than Albert Einstein said that compound interest is the most powerful force in the universe.

Let's see exactly how powerful this would be with some calculations done with the compound interest calculator available here.  We will use two annual excess earnings of  $100 million and $200 million to show the difference to the balance in PSPRS' underlying fund under scenarios where half ($100 million) of the amount over 9% goes to the PBI fund (as in the current excess earnings model) fund versus all ($200 million) earnings remaining in PSPRS' underlying fund (as in the SB 1609 model with PSPRS under 60% funded).  $200 million compounded monthly for five years at the current ERR of 7.85% would earn the PSPRS' underlying fund about $48 million more than $100 million compounded at the same rate.  This $48 million is in addition to the $100 million that was never placed in PSPRS' underlying fund in the first place, so instead of having an additional $296 million in it, the underlying fund has only an additional $148 million, with the remaining $148 million in the PBI fund.  Of course, PSPRS is not actually losing the $148 million total since it exists in a different account that PSPRS controls, but funds in that PBI account do not count toward PSPRS' funded ratio since those funds can only be used to pay PBI's.

While this is a simplistic scenario it illustrates how destructive the current excess earnings PBI model is.  The millions not going into PSPRS' underlying fund mean higher contribution rates for employers, and consequently, lower pay and benefits for current employees and diminished services for citizens.  Keep in mind that the current excess earnings PBI model is even worse than it appears at first glance because excess earnings are paid into the PBI fund regardless of PSPRS' past performance.  Using the previous examples, PSPRS could suffer a cumulative loss during the five year period but still pay excess earnings into the PBI fund because some individual years may have exceeded a 9% return.  In this case, not only does PSPRS lose an opportunity to help clear its unfunded liability, it actually increases its unfunded liability.  We do not need a compound interest calculator to know that this is not a sustainable method of paying PBI's. 

A Closer Look at PBI's


For PSPRS retirees, the payment of PBI's has been a great deal.  The most recent PBI retroactive for fiscal year  (FY) 2014 (July 1, 2013 to June 30, 2014) was $121.19/month or $1,454/year.  The FY 2013 PBI was $159.13/month or $1,910/year.  For someone with a monthly retirement benefit of $4,000/month, the PBI would mean a percentage increase of 3.0% for FY 2014 and 4.0% in FY 2013.  The average PSPRS pension for FY 2013 was $49,571/year.  The following shows the COLA's paid by the Social Security Administration over the past 40 years:

7/1975: 8.0%     1/1985: 3.5%    1/1995: 2.8%     1/2005: 2.7%
7/1976: 6.4%     1/1986: 3.1%    1/1996: 2.6%     1/2006: 4.1%
7/1977: 5.9%     1/1987: 1.3%    1/1997: 2.9%     1/2007: 3.3%
7/1978: 6.5%     1/1988: 4.2%    1/1998: 2.1%     1/2008: 2.3%
7/1979: 9.9%     1/1989: 4.0%    1/1999: 1.3%     1/2009: 5.8%
7/1980: 14.3%   1/1990: 4.7%    1/2000: 2.5%     1/2010: 0.0%
7/1981: 11.2%   1/1991: 5.4%    1/2001: 3.5%     1/2011: 0.0%
7/1982: 7.4%     1/1992: 3.7%    1/2002: 2.6%     1/2012: 3.6%
           *               1/1993: 3.0%    1/2003: 1.4%     1/2013: 1.7%
 1/1984: 3.5%     1/1994: 2.6%    1/2004: 2.1%     1/2014: 1.5%

* Change from a fiscal to a calendar year

The following show PBI's paid by PSPRS and their percentage of the same $4,000/month retirement benefit:

2004: $111.90/2.7%    2008: $134.34/3.4%  2012: $153.06/3.8%
2005: $116.82/2.9%    2009: $138.60/3.5%  2013: $159.13/4.0%
2006: $121.76/3.0%    2010: $146.74/3.7%  2014: $121.19/3.0%
2007: $127.06/3.2%    2011: $152.84/3.8%  2015: TBD

Comparing the two charts shows that since 2004 PSPRS has exceeded the COLA percentage paid by the SSA in all but two years, and in two years when SSA paid nothing, PSPRS paid 3.5% and 3.7%.   Under the current excess earnings formula, retirees are eligible for a PBI of up to 4% of the average normal retirement benefit, so it is possible that PBI's could be less than the rate of inflation.  However, with inflation kept pretty tame over the past 30 years, PBI's have been a boon for retirees, especially when we note that the average annual pension in FY 2004 was only about $35,000/year versus $49,571 in FY 2013.   Someone with a $3,000/month pension in 2004 would have seen a 4.0% increase in his pension that year, and in 2014 would now be getting a monthly benefit of $4,484, a nearly 50% increase in his benefit check in just 11 years, which makes his annual pension $4,000 higher than the FY 2014 average.  Using the SSA's COLA percentages would give us a monthly, inflation-adjusted benefit of $3,915, $570 less than what PSPRS actually awarded

Inflation

I would hope that everyone would agree that robbing the underlying PSPRS fund to pay PBI's will hurt everyone, whether retired or not, even if it has been, up until now, a real good deal for retirees.  SB 1609 addressed some of the problems by tying PBI's to PSPRS' funded ratio and raising the threshold for paying PBI's.  Of course, this is now a moot point since that part of SB 1609 has been ruled unconstitutional.  So where does this leave retirees?

Let's start by taking a few steps back and looking a little closer at some economic issues, particularly inflation.  When one considers an investment, it is necessary to determine the real interest one is earning.  This is simply the nominal interest rate minus the inflation rate.  This is equally true when considering an increase in pay as someone getting a 2% pay raise in an economy with 3% inflation is actually poorer (but of course better off than if he had gotten no raise).  For a retiree, whose only source of income may be his retirement benefit, knowing the real interest rate is very important.

If we look at the SSA chart again, we see a seven-year period between 1975 and 2002 where inflation was quite high, reaching 14.3% in 1980.  As stated before the current excess earnings model for PBI's allows for a maximum PBI of 4% of the average normal benefit.  Translated into actual dollars this could be more or less than 4% of an individual's benefit, depending on whether the individual's benefit was north or south of the average.  If we again use our $48,000/year pension and calculate a PBI benefit at the maximum 4% per year for four years, that annual pension would be $56,153 at the end of that four-year period.  It gets interesting if we begin to contemplate the possibility of inflation exceeding 4% per year.

If inflation is 6% per year, a retiree would find himself $10,597 poorer (with annual losses of $960, $2,016, $3,175, and $4,446) at the end of the four-year period based on his -2% real interest rate with a pension 7.3% lower than the first year (having only $56,153 available but needing $60,599 to keep up with the increased cost of goods).  At 8% inflation, he would be $21,613 poorer with a pension 14.0% lower than the first year.  At 10% inflation, he would be $33,061 poorer with a pension 20.1% lower than the first year.  We could go on with higher inflation rates or longer periods of time, but I think that the insidious effect of inflation is quite clear in these examples.

If you are a retiree, high inflation is your kryptonite.  Neither the current excess earnings model or SB 1609 allow for a PBI higher than 4%, and the PFFA proposal limits PBI's to 2%.  The danger here is the same one that has afflicted PSPRS for years: a sense of complacency that everything will continue on as it always has.  High returns will continue indefinitely--wrong.  Equities have always made us money--wrong.  Serious market crashes can be prevented--wrong.  Inflation can be controlled--right?  I don't know about the last one, but the track record has not been very good on the others. 

My final point on inflation deals with PSPRS itself.  It may seem like PSPRS would also suffer with inflation, but in fact under all the current PBI scenarios, inflation would be helpful in returning PSPRS to financial health.  Inflation benefits debtors who can pay off debts with cheaper dollars.  If inflation were to take off, PSPRS would most likely see an increase in market returns on its investments.  Wage inflation for employees and higher tax revenue coming to employers would bring in more contributions.  Higher market returns and greater contributions would go a long way to erasing PSPRS' unfunded liability, but retirees would be limited to the 4% or 2% PBI's. I have already expressed my disdain for the PFFA pension reform proposal as a current employee, but retirees should look askance at it as well when we consider the potential for harm to those on a fixed income.  The 2% limit on PBI's proposed by the PFFA were to be embedded in the Arizona Constitution making them nearly impossible to alter.

 Conclusion

So the past is no guide to the future.  In years of low inflation PSPRS was paying PBI's that exceeded inflation, yet PSPRS has no mechanism for dealing with extended periods of high inflation.  This is the legacy of treating PSPRS as some kind of magic money machine that only has to be programmed with the right formulas to keep paying out in perpetuity.  The next post will have some modest proposals to deal with this.