Out of the mouths of experts

April 10th, 2013 · No Comments · Blog, Featured Home Page Post

Below are public comments former Social Security Administration Commissioner Michael Astrue made before the Social Security Advisory Board on March 8, 2013. But read first just a portion of this powerful (and lengthy) speech, which sums up a smarter approach to solving the Social Security disability “problem” without throwing the proverbial social insurance baby away with the bathwater:

“So let me return to the question I posed before: what problem are we trying to solve? I submit to you that the academics, the policy experts, and the advocates of all stripes have not demonstrated an overarching issue in the system that could be the foundation for sweeping reform. Most of the time Social Security disability does exactly what the American people want it to do, which is to provide a safety net for people whose medical problems dictate that they legitimately cannot work for a year or more. I can tell you better than almost anyone that the system is far from perfect, but the problems result from many very different failings throughout a highly complex system, not one or two overarching issues. In my opinion, there is no magic bullet solution, only scores of detailed answers to very specific problems.”

Until we engage in this kind of approach, our nation’s conversation on social insurance will be mired in status quo.

Here is the entire text of former Commissioner Astrue’s speech:

Straight Talk about “Disability Reform”

I want to thank the SSAB for inviting me back to Washington. It has involved some sacrifice for me—I’m going to miss the Early Bird Special this afternoon…but I should be back in Boston for bingo tonight.

As the Chief Actuary and others discussed this morning, we are just three years away from Title II disability checks being cut by an automatic 20% or so. Whether we like it or not, these imminent cuts should rouse the Congress to try to legislate about disability insurance for the first time since Ticket to Work in 1999.

Unlike in the past, we should not expect that this Congress will do a quiet and efficient reallocation of the Trust Funds to cover the OASDI deficit. We do not have a sufficient number of the Daniel Patrick Moynihans & Bob Doles, the Bill Archers & Barbara Kennellys, enough leaders to lead a civil, informed debate on a hard topic. What we must try to do now to is start a process of education and dialogue that could let us avoid another inflamed, last-minute public spectacle, like the fiscal cliffs and sequestration, driven by what Stephen Colbert calls truthiness rather than actual truth. Truthiness tends to involve plausible but unsupportable leaps of logic from actual facts, and there is an increasing amount of it in Washington.

Congress worked itself into two inconsistent frenzies of truthiness during my reign of terror as Commissioner. When I started, Members were howling about backlogs and excessive stinginess in allowances. Members stated on the record that they believed that we denied all applications and made everyone who applied appeal, despite longstanding data showing that SSA allows about one-third of all initial applications.  I was told by one Member in a hearing on my third day on the job that I was an immoral person simply by being the head of the agency. Staffers made sure that these kinds of comments disappeared from the Congressional Record, of course, but they happened.

With slow but steady progress on backlog reduction over several years, that frenzy subsided. I then made the mistake of allowing myself a small sigh of relief. Almost immediately, a new frenzy began from the other end of the spectrum. In the deluded world of today’s Capitol Hill, SSA had abruptly moved from an agency of Ebenezer Scrooges to an agency of Robin Hoods—Robin Hoods who were indifferent about fiscal responsibility and simply rubber-stamped applications. Moreover, SSA was also suddenly indifferent about fraud and suddenly every street in America had someone living high on the hog with disability checks.

Let’s talk facts. Neither critique had any connection to reality. Both critiques were in large part diversions to distract attention from the fact that, in order to pay for sexier items, Congress had drained the SSA administrative budget so badly over two decades that the inevitable occurred—backlogs increased not only for hearings, but also for the continuing disability reviews that are the agency’s main weapon against improper payments. Having said that, the level of actual fraud is still extremely low, probably some fraction of one percent of the people receiving benefits. Critics of the agency have resorted to distorting data that relates to the perfection of documentation—not fraudnot even improper payments—in order to support overheated claims of massive fraud. The Baltimore Sun just ran a screed this week premised on exactly that distortion. Despite the “truthiness” of the massive fraud claim, it is a widely held belief by many vocal Members on the right and many discreet Members on the left privately confess that they share that  view.

It is also untrue that the agency has pushed its processes either to the Scrooge side or the Hood side in response to these pressures. As the Chief Actuary laid out for you, there is no sudden crisis in the disability. We face exhaustion in 2016, exactly as the Trustees projected two decades ago.

The number of people on the rolls is rising substantially, but that rise is a predictable consequence of the baby boomers reaching their disability-prone years, not the product of new rules or management. I handled the disability application for my father when he was hit with fatal brain cancer at the age of 52. I was diagnosed a year ago with severe rheumatoid arthritis. It is the frailty of the aging human body, not ideology or mismanagement that has caused almost all of the increase in the rolls over the last decade. It is basic arithmetic— we get a lot sicker in our fifties than we do in our twenties and we have a bulge of people in their fifties in our work force.

If you adjust statistically to compensate for the increasing age of the work force, the agency has been allowing cases at a just slightly higher rate over the last decade. I’ll let the social scientists try to figure out the reason for that small increase, but in my opinion it is likely to be the combination of many factors, such as: 1) health care providers systematically sending patients to third-party representatives, such as Allsup, in order to ensure they get Medicaid or Medicare; 2) continuing judicial activism in some parts of the country, particularly in California, that has inappropriately expanded the treating physician rule and other doctrines; 3) cultural changes that have made it more likely that claimants will present claims of mental illness along with their physical illnesses; and 4) the rise of obesity, which causes many disabling diseases, both mental and physical.

In determining what needs to be done, one also should look at the agency’s improvement in quality over the past six years. Due to better staffing, training, policy clarification and IT support, particularly the e-Cat system that guides examiners through a case and stops shortcuts, DDS quality on initial decisions has risen substantially to about 97-98%.

Everything is in motion for quality to continue to improve in the coming years. The biggest source of time, money and error is the fragmentation of medical records. Progress on consolidated electronic medical records is glacial, but it is starting to happen, and this year SSA went from pilot to practice with its partnership with Kaiser Permanente. Once the private and public sectors move to consolidated electronic medical records, disability determinations will become significantly faster, cheaper and more accurate.

Likewise, decision-making will become more accurate once the agency scraps the antiquated Dictionary of Occupational Titles and moves to a new vocational tool developed substantially by the Bureau of Labor Statistics. The blogosphere has been filled with conspiracies about this effort, but the simple truth is that SSA is still using a book that was written for other purposes in 1938, hasn’t had a significant update since 1977, and never included the mental aspects of performing occupational requirements. We don’t have buggy whip assembly line inspectors any more—we have IT specialists and teleservice representatives; claimants deserve a system that reflects modern realities of those kinds of jobs. By contracting with the Bureau of Labor Statistics to supplement the existing O*Net system, the agency has shaved tens of millions of dollars off the cost of replacing the DoT. It may be doable for an additional cost of 10-20 million dollars and periodic costs for updating—a small price to pay for accuracy in a multibillion dollar program. You should insist that this initiative continue to completion, which could be as soon as 2015.

So, my bottom line is that the agency is only getting more efficient and more accurate within a statutory framework that is a time-worn balance between competing worldviews in Congress. I sincerely doubt that Congress has the stomach for disrupting that balance, and I am quite concerned that reforms they might embrace could end up creating more problems than we’re solving. Indeed, the first question to ask a would-be “reformer” is: What problem are we trying to solve?

I’ve looked at “reform proposals” that people have proposed. Due to the increased costs of such proposals, I see no likelihood that Congress will adopt a partial disability system similar to the VA’s, and I think it would create substantial litigation, expense and delay—the VA system is much smaller than SSA’s and the federal courts will never give SSA the deference they give the VA.

I also see no likelihood the Congress will fund a move toward “temporary disability”—it isn’t willing to fund continuing disability reviews now, which save the trust funds $10 for every dollar invested. How could they possibly get themselves to vote for another million reviews of beneficiaries each year when they can’t be motivated by a 10-for-1 return on CDR’s?

I am very concerned, though, about the possibility of another well-packaged but ill-considered reform initiative driven by the OMB institutional views, which relies on the fallacy that we can solve problems created by bureaucracy and complexity with ever-more complex bureaucracy and complexity. Like Ahab chasing Moby Dick, they reject all the evidence and experience to date and insist that they can harpoon a system that will be so much more accurate up-front that Congress will be impressed by the back-end savings and fund the added bureaucracy.

This whale hunt failed with the Clinton Administration’s prototype initiative, and it failed more miserably with the Bush Administration’s DSI initiative. In fact,  DSI almost brought the agency to its knees.  That effort blithely assumed that more than tripling the cost of reconsideration would produce huge, more-than-offsetting downstream savings, an assumption for which there was no empirical support whatsoever, and which did not happen in practice.

Moreover, with quality already dramatically improved, and almost certain to continue improving with electronic medical records and new vocational standards, mathematically there isn’t enough room left for OMB to assume whale-like improvements in upfront quality that could conceivably produce the programmatic savings for which they are desperate.

OMB should drop its harpoon, let that imaginary whale swim into the mist, and focus its attention on actions that could actually produce real program savings: two-year rather than one-year appropriations cycles; stronger support by the Appropriation Committees for the IT carry-over fund; and CDR funding formulas that would allow CBO to score the program savings in a way that would shelter associated administrative costs.

So let me return to the question I posed before: what problem are we trying to solve? I submit to you that the academics, the policy experts, and the advocates of all stripes have not demonstrated an overarching issue in the system that could be the foundation for sweeping reform. Most of the time Social Security disability does exactly what the American people want it to do, which is to provide a safety net for people whose medical problems dictate that they legitimately cannot work for a year or more. I can tell you better than almost anyone that the system is far from perfect, but the problems result from many very different failings throughout a highly complex system, not one or two overarching issues. In my opinion, there is no magic bullet solution, only scores of detailed answers to very specific problems.

My ringing call for incrementalism is hardly sexy, but if you care about supplying a compassionate but cost-effective safety net for the disabled, you will need to have a laundry list of concrete reforms to offer Congress in the next year or so. As I said before, Congress is not going to reallocate trust fund money from the retirement and survivors’ fund to the disability

fund without extracting “a fix” to disability. If a thoughtful bipartisan consensus doesn’t emerge in the coming months about improvements, Congress may well listen to the uninformed and, without hearing or debate, insert provisions that would be damaging to taxpayers and the disabled. In short, you need to build bridges with others and develop a legislative shopping list.

Let me “help” you with this list.

Let’s start with the front end of the process. There are claimants clogging up the process who shouldn’t be there in the first place. Typically, budgeteers in state agencies have figured out that they can shift their program costs to the administrative budget of SSA by requiring a decision on disability from SSA before allowing people to collect benefits from TANF or some other state-funded program. These policies are at best inconsistent with Congressional intent if not illegal, they create untold waste at SSA, and they are unfair to people who qualify for public assistance. It is time that Congress closed this loophole and made it explicitly illegal to condition receipt of any benefit upon an SSA disability determination if that benefit is partially subsidized by the federal government.

Another similar area for action is concurrent application for both unemployment and disability. There has been a lot of truthiness on this subject lately in Congress, which wants to blame Labor and SSA for not coordinating to bar this practice. Fortunately, GAO stuck to principle and made it clear in a report to Congress that the federal courts had determined that Congress had not actually written the law clearly enough for the agencies to coordinate in that way. I do believe, however, that there is something fundamentally wrong about representing to Labor that you are actively looking for work while you are representing to SSA that you are incapable of working for twelve months or more. Congress should stop the blame game and close this loophole.

Congress can also help claimants help themselves by authorizing SSA to direct the state agencies to develop  comprehensive, up-to-date websites in each state that include: 1) sources of assistance for free pharmaceuticals; 2) locations of DoL one-stop offices; 3) information about services available from the state vocational rehabilitation agency; 4) disease groups that offer supportive services; and 5) a primer about looking for work on the Internet and information about data bases that list available jobs, both locally and nationally. For claimants who are on the borderline of qualifying for benefits, this kind of information can be of enormous help while also holding down program costs. One has to be skeptical of labor-intensive solutions when the agency has shrunk from almost 70,000 employees to just over 60,000 in three years— and there is pressure to cut even deeper.

Let’s move on to the state agencies that make the initial determinations, the DDS’s. Some Members of Congress and public employee unions have argued that the DDS’s should be federalized, a billion dollar per year solution in search of a problem. In a time of ever- tightening budgets, the smarter thing to do would be to spend that billion each year on staff and technology, and then deal with the occasional jurisdictional issues between the states and SSA on a more surgical basis.

Let’s talk briefly about the most important of those jurisdictional issues. Congress, for all its complaining about DDS furloughs, never lifted a finger on SSA’s anti- furlough bill for DDS employees. Here is an issue where DDS management and unions, claimants and taxpayers are all in alignment, and still Congress did not see fit to even hold a hearing. For a state to furlough a DDS worker is lazy budgeting. It makes no economic sense for a state to lay off a fully reimbursed state worker—it slows payments to people in that state and diminishes tax revenues with no benefit at all to state governments.

Congress also needs to give the Commissioner the mandate to begin over the next five years an orderly elimination of the reconsideration step that occurs in most, but not all, states. Reconsideration as a crude quality control tool made sense when accuracy rates were lower than they are today, but accuracy has improved to such an extent in many states that it is only an annoying three-to-five month delay on the way to a final DDS decision. The cost of elimination of reconsideration should no longer be premised on flawed and outdated data from the prototype initiative—Congress should allow the Commissioner to eliminate reconsideration in any state that has maintained an accuracy rate of 98.5% for two consecutive years.

The cost implications of eliminating reconsideration can be tracked in real time, and the Commissioner should have the discretion to stop the transition if there are unexpected cost consequences, which I think is unlikely. The bottom line is that we can free up hundreds of millions of dollars for faster initial decisions and CDR’s while cutting three to five months off the total waiting time for claimants without significant change to outcomes. 

OK, on to hearings. In the past six years SSA successfully challenged the conventional wisdom that ODAR was inherently unmanageable. It moved to fully electronic processes, it handles more than 20% of its cases by video, the quality and training of staff has improved, and representatives can even obtain data on the progress of cases without distracting ODAR employees. Quality and efficiency are up dramatically.

Congress likes to take credit for the agency backlog reduction successes by pointing to the additional resources it provided in fiscal years 2008-2010, which were an enormous help. However, Congress takes far too much credit when it makes such assertions. Without ODAR’s increase in efficiency, its backlog reduction efforts would have led to a plateau of waiting times rather than a serious decrease. Let me punctuate that point: about half of the backlog reduction came from more resources and about half came from increased productivity. Of course, those added resources are long gone, and today the higher staffing at ODAR comes from sacrifices made elsewhere in the agency, sacrifices made possible by maintaining average annual increases in productivity of almost 5%.

In the last few years we also disproved the four-decade -long mythology that the Merit Systems Protection Board would not seriously discipline administrative law judges, and we removed more judges—4—than all previous Commissioners combined—3. However, by taking a stand on judges who assault women and infants, distribute pornography from government computers, and steal by holding two federal jobs, we have received dozens of resignations from bad actors who did not want to experience public exposure for their actions. The arrogance that leads a person to do such things also correlates highly with poor decision- making, so when took a tough stand on conduct we got the bonus of losing a lot of lazy and sloppy judges. As a result, we probably have the fewest number of outlier judges who refuse to follow agency policy that we have ever had, a change that is saving the trust funds hundreds of millions of dollars each year.

Just before my departure we received the first favorable decision from an MSPB ALJ terminating an SSA ALJ for lack of productivity. This case involved an experienced judge who decided about 100 cases per year—20% of our minimum expectations. He responded to counseling and remedial training by cutting his productivity in half to just 50 cases per year. His termination, which I am confident will be sustained by the full MSPB and the federal courts, marks the first time ever that a judge has been removed for simply failing to do his job. In most organizations, such an event would be unremarkable; in ODAR it is historic.

SSA accomplished much of this progress over the intransigent resistance of the Office of Personnel Management.  They steadfastly refused to reopen the list for over ten years—imagine what your hiring would look like if you could only hire off a list that was ten years old! I sat in one meeting where a senior OPM official told John Berry and me that there was no reason to consider new applicants until SSA hired from the last person left on the list. What does that statement say about OPM’s commitment to quality in public service?

What makes it worse is that OPM is spending a lot of trust fund money doing very little. OPM has interpreted the ALJ statute wrongly—as authorizing them to unilaterally design whatever tests and procedures they want, to withhold all information about what they’re doing, and to provide no justifications for what was done—just a bill, last year for over $2 million in a year in which they did not administer a test.

I believe the ALJ position is the only job left in the civil service that requires a test, and it is a test designed by people who have no understanding of the administrative process and no desire to learn. It has become a budget game to pad the OPM budget at the expense of the trust finds. For instance, not long before I left I learned third-hand that trust fund money is going to develop a new on-line test announced this week will be administered by little cartoon figures on the screen. Little cartoon figures…

Where are the oversight committees on this critical function? OPM’s work is excruciatingly slow, outrageously expensive, and of unacceptably poor quality— and it has been so since at least the Clinton Administration. OPM damages SSA’s efforts to supply timely, high- quality justice, and it is time for Congress to move OPM’s responsibilities in this area to a federal agency that understands administrative law—either the Department of Justice or the Administrative Conference of the United States.

To close my thoughts about ODAR, I want to take a moment to note how important it was that we killed the DSI plan to abolish the Appeals Council. When you are running what is by many measures the largest system of justice in the world, it is insane to junk your primary tool for maintaining quality and consistency. Chief Judge Jonas and Deputy Chief Judge Ray have done a great job refining the role of the Appeals Council so that for the first time there is professional dialogue between the Council and judges who have been overruled. With the establishment of the Appeals Council’s first quality unit, ODAR for the first time also systematically interacts with other SSA components to eliminate recurring errors. It is no wonder that Chief Judge Jonas was the first SSA winner of the prestigious Deming Award.

Before I close and respond to questions, I would like to mention a few areas of possible standard tightening that might allow Congress to accept a reallocation of the trust funds so that beneficiaries and recipients do not receive a dramatic cut in their benefits in the last year of this Administration.

First, Congress needs to look closely at the recent ACUS report on the treating physician rule. The agency has struggled with this rule for decades—in order to make it at all workable it has become increasingly baroque, and thus it has become a major source of error in ALJ decisions. It has also been an invitation to federal judges, particularly in California and other locations, to turn the “controlling weight” language into a conclusive presumption in favor of the opinion of the claimant’s physician on the ultimate issue of disability. It seems to me that we hire judges to make judgments, and the law should give the judges with expertise substantial discretion to weigh evidence as they see fit in accordance with time-honored principles of ascertaining witness credibility. A judge should show deference to opinions on the diagnosis from a true consulting expert; he or she should also discount down to zero the weight of an opinion on the ultimate issue of disability that comes from low-end mills that do high volume business for one or a small number of representatives. That is not how many courts interpret the law now, and Congress should take a hard look at this area if it decides to control program costs.

Second, Congress needs to streamline its statutes on return to work. Congress has tended to have an unrealistic view as to how many recipients can return to work. Somewhere around 60% of the people on the rolls have no realistic expectation of returning to work—they have fatal diseases, like ALS or Huntington’s, that are certain to continue debilitating the person until he or she dies a premature death. Some people have horrific brain damage from car accidents or military service.

For many of the 40% for whom a return to work would be difficult but perhaps possible at some point, the layers of new incredibly complex statutes to encourage work have become counter-productive. It is a shame that Congress has ignored the agency’s work incentive simplification proposal, known by the acronym “WISP,” which would have eliminated many of the confusing and overlapping rules, and would have created a simplified route back to work. There would be many other ways of achieving the same result, including BOND, but it is time for Congress to thin out the failed ideas of the past.

One key theme for efforts to promote return to work should be charging the agency to focus attention and resources on those who we know are most likely to work. A person with the same form of rare brain cancer that killed Senator Kennedy and my father will never do a substantial amount of work again. On the other hand, someone who is bipolar or schizophrenic might work simply by getting the right drug treatment ahead of the 24-month Medicare waiting period. The same would be true for some people with certain progressive autoimmune diseases and genetic protein deficiencies. Both the agency and Congress tend to be culturally attracted to a one-size-fits-all approach, but if you’re serious about return-to-work, you have to be serious about segmenting the claimants and focusing scarce resources on those who realistically might work.

When Congress gets serious about addressing the 2016 insolvency of the trust funds, there will be bad ideas floating around too. The one that has the most currency baffles me, which is another try at making hearings adversarial. I was stunned when I first answered questions before Congress on this proposal because none of its proponents knew that the agency had piloted this proposal in the 1980’s and that it failed miserably. It was expensive—probably several hundred million dollars to implement fully in today’s dollars—and it made no difference in outcomes while simultaneously undermining public confidence in the agency. Moreover, a primary rationale for the pilot, that government reps could find medical evidence that judges could not, will be unsupportable within five years when we enter the new world of electronic medical records.

Before I conclude, I would like to add a cautionary note about a recurring area of controversy, children receiving SSI. When the story broke a few ago in Boston that parents had killed their four-year old daughter by drugging her to get her fraudulently onto SSI, we looked at the data we had to try to figure out whether we had a problem on our hands or not. 

The truth is we couldn’t tell. On the one hand, there was nothing in the data that suggested a recent dramatic change or extensive fraud. On the other hand, there were anomalous patterns in a handful of cities around the country which tended to suggest that an unusually high number of children are on SSI, even adjusting for income and other factors. That is a legitimate source of concern.

My answer was to get an in-depth study done, a proposal I thought would be noncontroversial, but initially drew fire from both OMB and the advocacy community. As everyone calmed down and talked, I think a broad consensus developed that a study needs to be the predicate for policy changes. Statutory changes based on bias or guesses are likely to hurt both children and the taxpayers in the long run.  

Let me reiterate my thanks to the Board for allowing me to express many of my cantankerous opinions. You all have been very patient, for which I am grateful, and so before I head out to the airport to fly home for bingo, I would be happy to take your questions.

March 8, 2013

SSAB Forum




Leave a Reply

You must be logged in to post a comment.