Why Health Cost Growth Increases after Estimators Say it’s Slowing: Observer Effects and Feedback LoopsPosted: August 3, 2015
“Health cost growth has slowed down, we think. So let’s increase health costs.” This is the federal government’s apparent response to some recent sanguine estimates about the future of health cost growth. We might call this response a policy version of the “observer effect,” where the mere observation of reality changes that reality. In this case, the observation that health care costs may be increasing more slowly than expected creates a political reality in which fewer efforts are exerted to keep costs under control.
Projections based on past historical trends are fraught with danger. The influence of government policy sits near the top of that danger list. Since federal and state spending plus tax subsidies now cover about 60 percent of the health care budget, government legislation decides much of what the nation will pay for health care. Speaking technically, policy is endogenous to—or influential on—the past trends we measure.
Logically, then, future legislation too has a powerful effect on the direction of health costs. But possible policy changes are an unsteady foundation for cost projections. Government agencies like the Congressional Budget Office try to get around this dilemma by treating policy as exogenous to—or not influencing—cost projections. That way, those agencies can display the implications of current laws, even when those laws imply a growth in cost that is unsustainable.
But health researchers, the public, or elected officials who conclude that past trends will simply continue often fail to account for how policy decisions feed health costs and vice versa. US health care insurance, including that provided or subsidized by government, still offers fairly open-ended access, allowing consumers to spend more and providers to earn more at others’ expense. If policymakers interpret slower cost growth to mean they need fewer new cost-reducing measures and can even rescind some old ones, then as a consequence health costs are going to rise faster.
It now seems clear that Congress and the Obama administration have responded to these new estimates by taking a more lackadaisical attitude toward controlling costs. Two pieces of evidence:
- Congress has long squabbled over how to deal with legislation that tried to make the growth in Medicare costs sustainable by cutting payment rates to doctors for certain procedures. Past efforts led to “doc fixes” that held off such cuts for a while and let them accumulate. All was not lost: according to the Committee for a Responsible Federal Budget, these annual forestalling actions involved significant other health cost cuts to pay for each fix. In 2015, however, Congress threw in the towel: it abandoned the old requirements on doctors while providing limited real offsets.
- Many Republicans and some Democrats now oppose a tax imposed under Obamacare that requires insurance companies to pay a tax on high-cost insurance plans (plans whose value exceeds $27,500 a year for a family and $10,200 a year for an individual in 2018). Admittedly an imperfect device, the tax did address both conservative and liberal concerns, backed by solid research, that offering a tax subsidy for costs above a cap mainly led to higher health care costs while doing little to expand coverage. Abandoning the high-cost plan tax would effectively increase health costs even more.
Harder to substantiate are cost-controlling initiatives that are abandoned or never undertaken. For instance, President Obama has removed some of the health saving initiatives that used to be in his budget—such as limits on state gaming of Medicaid matching rates—presumably because he thought these initiatives were unlikely to get through Congress, or he had enough health care fights on hand. How much have payment advisory commissions felt that they could let up on new suggestions to reduce prices? In general, how does a perceived reprieve from pressure lead any of these actors to kick the can down the road to their successors or at least until after the next election? Paul Hughes-Cromwick of the Altarum Institute also asks whether something similar doesn’t go on in the private sector: for example, would specialty drug makers price new entries so aggressively (e.g., Sovaldi for Hepatitis-C or even Jublia for toenail fungus) if we weren’t simultaneously coming off historically low spending growth?
My advice to estimators: include a feedback loop to demonstrate how your estimates affect the behavior of those making decisions on the basis of your estimates. Your projections of lower health cost growth may end up increasing health costs.
In testimony yesterday before a joint hearing of two House subcommittees, I urged Congress to modernize the nation’s social welfare programs to focus on early childhood, quality teachers, more effective work subsidies, and improved neighborhoods. One way lawmakers can shift their gaze is by considering the effects of combined marginal tax rates that often rise steeply as people increase their income and lose their eligibility for benefits.
While some talk about how we live in an age of austerity, we are in fact in a period of extraordinary opportunity. On a per-household basis, our income is higher than before the Great Recession and 60 percent higher than when Ronald Reagan was elected President in 1980.
A forward-looking social welfare budget should not be defined by the needs of a society from decades past. Two examples of how our priorities have shifted: Republicans and Democrats didn’t always agree on the merits of Aid to Families with Dependent Children (AFDC) or the Earned Income Tax Credit, yet they agreed on the need to shift from welfare to wage subsidies. Ditto for moving from public housing toward housing vouchers.
I sense that both the American public and its elected representatives are united in wanting to create a 21st century social welfare budget. That budget, I believe, should and will place greater focus on opportunity, mobility, work, and investment in human, real, and financial capital.
However, for the most part, our focus has been elsewhere. As I show in my recent book Dead Men Ruling, we live at a time when our elected officials are trapped by the promises of their predecessors. New agendas mean reneging on past promises. Even modest economic growth provides new opportunities, but the nation operates on a budget constrained by choices made by dead and retired elected officials who continue to rule.
For instance, the Congressional Budget Office and others project government will increase spending and tax subsidies by more than $1 trillion annually by 2025, yet they already absorb more than all future additional revenues—the traditional source of flexibility in budget making.
I am concerned about the potential negative effects of these programs on work, wealth accumulation, and marriage of combined marginal tax rate imposed mainly on lower-income households. To see how multiple programs combine to reduce the reward to work and marriage, look at this graph.
For households with children, combined marginal tax rates from direct taxes and the phasing out of benefits from universally available programs like EITC, SNAP, and government-subsidized health insurance average about 60 percent as they move from about $15,000 to $55,000 of income. This is what happens when a head of household moves toward full-time work, takes a second job, or marries another worker.
Beneficiaries of additional housing and welfare support face marginal rates that average closer to 75 percent. Add out of pocket costs for transportation, consumption taxes, and child care, and the gains from work fall even more. Sometimes there are no gains at all.
While there is widespread disagreement on the size of these disincentive effects on work and marriage, there is little doubt that they exist. One solution: Focus future resources on increasing opportunity for young households. Make combined tax rates more explicit and make work a stronger requirement for receiving some benefits.
This post originally appeared on TaxVox.
Remember when people complained that hedge fund managers and private equity firm owners paid a lower tax rate than many workers? Or when Warren Buffett said he shouldn’t pay a lower rate than his secretary? In those cases, investors were benefiting from low capital gains rates, but at least they paid some tax.
Now, thanks to Roth accounts, a special form of retirement savings, investors and even their children pay close to zero tax on huge sums of income. While this scheme can create significant future government budget shortfalls, my attention here is on how our tax code favors those who understand how to play this Roth-account game.
You got a hint of what might be possible when Governor Romney disclosed that he had over $100 million in his individual retirement account. However, this money was in a traditional IRA, taxable upon distribution, not a Roth IRA, where once a modest tax is paid up front, then all future income from the account escapes tax. Thus, Romney missed out on some of the tax benefits available to Roth holders.
The Taxpayer Relief Act of 1997 first made such accounts possible. At about that time, I warned in Tax Notes Magazine about how some investors could use Roths to avoid almost all tax on significant amounts of income. But I didn’t have a smoking gun, since I was anticipating future accruals on this then-new option. Tax professionals who put high-return hedge fund or private equity return assets into a Roth for wealthy clients have no incentive to disclose the details of these deals.
But a recent article by Richard Rubin and Margaret Collins for Bloomberg describes a redesigned retirement plan of Renaissance Technologies that allows employees to put retirement monies into Medallion, a hedge fund that historically has produced extraordinary high returns. My colleague, Steven Rosenthal, examines the tax implications and questions the future role of the Labor and Treasury Departments in these developments.
But you don’t need to work for a private equity firm or hedge fund to benefit. Let’s start with the tax sheltering opportunity available to almost anyone willing to invest far into the future and who earns no more than the normal long-term rate of return on stocks.
Suppose a 25 year-old taxpayer puts $10,000 in Roth accounts, perhaps helped out by parents. Once he pays a small upfront tax on the contributions, the tax on these accounts is now totally prepaid. He will never owe another dime. If stocks provide a traditional historic return of, say, 9 percent per year, the portfolio would grow to $483,000 by age 70, the age at which people must withdraw from traditional retirement accounts but not Roths. Make a few such deposits, and soon millions of dollars of income can escape tax.
But 9 percent represents an average return, chicken feed compared to the successful private equity investor, inventor, entrepreneur, or small business person. If they are insiders or somehow know, they might well know that they are likely to generate 15 percent per year over time. Others might also generate this higher return because of invention, luck, or simply leveraging up their returns. If the investment returns 15 percent annually, a $10,000 deposit held for 45 years grows to $5.4 million by age 70.
Perhaps you think the example of a 25-year old depositing money until age 70 is exaggerated. But it understates the more extreme cases. A Roth IRA can be held much longer than 45 years, e.g., a 35 year old living to 90 can leave the money to his kids who must withdraw the money only gradually after the parent’s death, so in this case complete tax exemption would last 55 years, and partial exemption would extend outward toward a century.
Think what this does for the fairness of a tax system. People who turn out to make millions of dollars of income on their investments will often pay no more tax (up front or down the road) than people who make little or nothing on their saving. To be clear, I’m saying dollars of tax, not rate of tax, would often be the same. The unsuccessful will be taxed the same as the successful, violating the whole notion of proportionality, much less progressivity, in taxation. I can think of almost no other tax provision, whether in an income or consumption tax, that goes so far in violating the notion that those who either make more money or consume more can afford to pay more tax.
This post originally appeared on TaxVox.
That is how Senate Majority Leader Mitch McConnell explains his support for granting the president expedited authority to negotiate trade deals and fast-track through Congress a vote those deals. The authority would apply not only to an agreement negotiated by President Obama with several Pacific partners, but to agreements made through mid-2018 and potentially through mid-2021.
McConnell recognizes, at least in this case, that a president must have enough room to perform his executive duties. The Constitution provides the president with particular powers, including the negotiation of treaties. It makes no more sense for 535 members of Congress to negotiate treaties than it does for them to micromanage other tasks that should be left to the country’s chief executive.
I would like to extend McConnell’s thought beyond treaty making. Now is the ideal time to empower both the president and Congress to better perform their assigned functions: the president to execute and Congress to legislate.
Why? First—and obviously to even the casual observer—both branches of government have been weakened extraordinarily over recent decades by the politicization of every action and the growing influence of interest groups. Many impasses in both legislation and administration arise when one party thinks that it can diminish the other by imposing roadblocks. This thought is neither new nor unique to government; when an organization’s decisionmaking boundaries are ill defined and its members cross those boundaries to seek additional power, the organization often becomes dysfunctional.
Second, less than two years away from a presidential election with an uncertain outcome, elected officials more likely recognize, as Senator McConnell does, that maintaining roadblocks deters one’s own party’s likelihood of future success as much as that of one’s opponents.
Third, the last years of a presidency seldom focus on the changes that new power brings a political party. Yet it need not be a lame-duck period. It offers the opportunity to turn to those process reforms usually neglected when the political debate centers on bigger or smaller, rather than more effective, government. President, cabinet secretary, congressional leader, and committee chair alike should be examining their power and reorganizing both internally and across jurisdictions—or, quite bluntly, they are not doing their jobs.
Behind closed doors, almost every elected official will admit that many government systems are broken. Examples abound: Medicare continually out of balance, infrastructure and highways unfunded while bridges fall apart and trains crash, the inability to pass budgets or appropriations bills, a sequester requirement that must be overridden, and much more. These are process failures, not policy failures.
How might strengthening the hands of both the president and Congress improve the likelihood of solving such problems?
Medicare provides a good example. In the last presidential election, the candidates attacked each other for trying to constrain cost growth in what both knew was an unsustainable system. Governor Romney castigated the president for cutbacks that helped expand health insurance for the nonelderly, and the president attacked the governor for favoring a voucher-like approach put forward by then–House Budget Chair Paul Ryan.
In truth, all answers to the Medicare problem involve payment constraints. The program simply has to operate within a budget. Congress won’t create a budget for Medicare, but it won’t allow the president to use his executive power to do it either. Instead Congress passes the power of appropriating money onto our doctors and us as beneficiaries.
The fix is simpler than it seems. If the Democrats favor price controls for Medicare and the Republicans favor voucher-like approaches, then set up the general rules but let whoever attains executive power use it whenever spending starts to exceed a congressionally approved limit.
This method can work well in other arenas too. Congress can set guidelines for what it wants accomplished—certainly within a budget—but then it should allow the executive branch to fulfill those guidelines. If Congress over-constrains any particular function by demanding that more be done than allowed by the budget it sets, then the executive should be empowered to make changes necessary to restore balance. This is not rocket science, it is basic management theory.
In his acclaimed book The Rule of Nobody, Philip K. Howard similarly argues that the president must have executive powers restored, to be able to avoid wasteful duplication and unnecessary bureaucracy, to expedite important public works, to refuse to spend allocated funds when circumstances change and the expenditure becomes wasteful, and to reorganize executive agencies.
When Congress limits the president on executive matters, no matter how small, it isn’t empowering itself. Instead, it entangles itself in complex and contradictory legislation, attempting to appease every interest (no matter how small), while weakening itself as it spends less and less time tackling the big issues that it is elected to address.
All this does not let recent presidents off the hook. The constant expansion in political appointees and the centralization of power in the White House over several decades has led to even more roadblocks to progress. When every decision must go through several political layers, almost no good idea can filter through to the president. When so many public statements and decisions on millions of government actions must be fed through the White House, civil servants and even top political appointees can’t function well, and they often retreat to doing nothing risky and seldom attacking limitations or failures in their own programs. Among the further consequences, many excellent government officials retreat to the private sector. Who wants to work where you are not allowed to do your job?
Whether one agrees with the examples presented here matters less than recognizing the ripeness of this time for procedural reform. Senator McConnell is right. Let’s empower the next president—and the next Congress as well.
The ultimate goal of educational policy must be progress for every child. Not standards. Not attainment of grade level proficiency. Not college readiness. But progress toward developing each young person’s potential to the fullest extent possible every year. Not only is this the right educational goal, but it is the only one that pulls parents, teachers, and administrators together politically in a shared vision of helping every child, disadvantaged and advantaged alike, to grow into smarter and more capable citizens. With rare exception, each student’s progress, no matter how high or low the base level of attainment, eventually benefits others in society.
While most teachers and parents adhere to that goal, they also care deeply about their own children and protégés, and many top-down requirements ignore that legitimate and natural impulse. When my children were in school, governments and school boards devoted additional resources to the “gifted and talented.” Others felt left out. Later, the focus shifted: the schools needed to do more for “special education” students, then those for whom English was a second language (ESL). Next efforts were made to pull more students into advanced placement AP courses, then to mainstream a larger share of students with different abilities. Later, standards became the focus du jour, culminating partly in 2001 in the No Child Left Behind (NCLB) law, and more recently in new fights over the “common core” level of achievement agreed to by state officials.
The reaction? I think it can be summed up well by the story of the school superintendent who in response to No Child Left Behind essentially told his principals to focus on that group, let’s say 20% of students, with the highest probability of being brought up to the standard. As for the other 80 percent of students, those above the standard and those more likely to never achieve it, well, they implicitly had to accept relatively fewer resources if relatively more were devoted to the 20%. A more extreme reaction came to light with the “racketeering” conviction of several Atlanta school teachers and officials for feeding students answers to standardized tests and changing test sheets.
These various reform efforts didn’t necessarily fail. They responded partly to past areas of neglect. But one can also see that if each program’s success depends mainly upon shifting attention and resources, and if there are narrowly circumscribed measures of success, then it’s quite possible to come full circle on these efforts, succeed modestly with the targeted population each time, and then at the end of the day advance not at all with any one of them as the targets rotate into and out of view. To me that partly explains why our school systems have fallen behind those of many other countries.
The literature on performance measures, successful organization, and statistics gives us many lessons that need to be heeded. Among those most relevant here: no organization succeeds unless it engages in a process of continual improvement. Those who are on the ground must buy into and have ownership of the process. Tests should occasionally shift to areas that haven’t been checked. For instance, if success in math comes about because extra time to it came out of fewer recesses, one had better check on whether active students become more out of control and if obesity is increasing. Finally, it’s all right to teach to a test if we know that the test incorporates much of what we want to succeed. A great example in the school systems are the advanced placement tests, where teachers often follow a fairly rigorous but also standard way to advance a selected group of students on a subject matter eventually to be tested.
Teaching to the test is bad enough when it discourages educators from teaching necessary skills that those tests neglect. But it is even worse when state governments, school boards, or school superintendents grade a school, or principal, or teacher primarily on the percentage of students reaching some minimum standard. Such judgments ignore both other sources of knowledge and the potential progress of many students above or below the standard.
Though No Child Left Behind may have failed on some fronts, it did help set in motion more rigorous efforts to track students over many years. These tracking systems often provide the types of data by which progress can be measured along several (but, of course, far from all) useful dimensions. The potential of these data for first empowering teachers and parents with multiple measures of each student’s progress, not just attainment, has yet to be realized.
Senators Lamar Alexander and Patty Murray, chair and ranking member of the Senate Committee on Health, Education, Labor, and Pensions, have recently released a bill known as the Every Child Achieves Act. The label at least suggests a focus on every child, rather than NCLB’s focus on a subset of students who are “behind.” Many experts call the bill a step in the right direction because it tries to maintain some accountability even while allowing states much greater flexibility in setting standards.
Still, whether the states advance our still-mediocre educational system—either on their own or with the help of federal incentives—remains to be seen. A crucial telling point will be whether enough schools and jurisdictions start to recognize how to better use measures of progress, not just attainment, and then aim to develop each and every student’s potential regardless of whether they fall well below, near to, or well above any particular attainment standard.
There are many ways to restructure the tax code. Elected officials often fail to detect opportunity when they adhere in a path-dependent way to one past model of success, such the 1986 tax reform. An alternative approach where even the acclaimed 1986 effort made at best modest progress would focus on making tax code simpler for taxpayers and improving compliance without adding to IRS costs. This reframing of reform stands a chance of stepping around the partisan wrangling that deters progress on so many other policy fronts.
In recent weeks, IRS Commissioner John Koskinen has been lamenting the Service’s inability to perform its functions well, mostly recently at today’s Tax Policy Center event on the impact of IRS budget cuts. It not only continues to lose staff, but less experienced staff quickly must replace a large fraction of senior employees who are leaving.
At the same time, congressional Republicans have been finding fault with much of what the IRS does, from its inability to regulate the activities of social welfare 501(c)(4) organizations to its implementation of the Affordable Care Act. Democrats say the solution is a bigger IRS budget. Republicans say it is a smaller one. But there is an alternative choice.
Focus tax reform on an objective almost never before given priority: developing a system that is less costly for taxpayers to comply with, mostly through simplification of existing laws. Simultaneously modify laws where IRS enforcement is difficult or impossible or too costly ever to do well. Such an effort would not necessarily require base broadening, though that often would be consistent. For instance, rather than engaging in a non-productive debate over whether to eliminate or expand refundable tax credits for low-income households, Congress could simply replace the existing credits with versions that are easier for taxpayers to understand and for the IRS to administer.
Put everything on the table, ranging from business deductions to multiple capital gains tax rates to complexity of retirement plan options to refundable credits that go mainly to the poor. Include many of the small items like mileage deductions that can be vastly simplified.
The first step in such an effort would be for the IRS to better understand known error rates for each preference and “program” based partly upon the audit data it keeps. Believe it or not, the agency has never done any good comprehensive study on this, though on occasion it has tried to measure the extent of noncompliance in rough terms.
And the IRS has never conducted this type of research to inform policy, as opposed to allocating internal audit functions. IRS should study taxpayer time costs, error rates, causes of error, costs of auditing, and administrative aspects of every program put into the tax code. For instance, although relatively easy to do, it has never really distinguished in its studies what it can’t know or can’t enforce well from the error rates that it can find under some audit technique.
The second step would be for the IRS and its colleagues at the Treasury’s Office of Tax Policy to be prepared, at least in a tax reform setting, to acknowledge what provisions it cannot monitor at a reasonable cost. For instance, it can estimate compliance with wages reported to the IRS on W-2 forms, but it seems fairly certain that it has little idea of how much charitable contributions of clothing are overvalued. Even if it did, it has no way to efficiently allocate resources to monitor such deductions.
I recognize that a balancing act is required, since IRS doesn’t want to encourage even more noncompliance. But an agency can’t engage in continuous improvement if it doesn’t study and acknowledge its problems in the first place. Indeed, some of IRS’s recent public conflicts, such as over the tax-exempt status of Tea Party and other social welfare organizations, stemmed from what every tax policy analyst has always known: that limited resources devoted to vague or hard-to-enforce law eventually can lead to disaster.
With attention to individual simplification, along with the improved compliance that comes from removal of laws that can’t be enforced, both parties could claim a victory. A simpler and more efficient tax system conforms to both progressive and conservative principles.
Unlike more money for audit, moreover, making the system more administrable and simpler modestly increases national well-being.
In a fairly comprehensive study I undertook for the American Bar Association and the American Institute of Certified Accountants many years ago, I found that expanding IRS audit resources was likely to raise more than was spent, ranging from $2 to $7, depending upon type of return, for every dollar spent. I doubt the conclusion has changed greatly since then. But while more auditors would help redistribute the tax burden in favor of compliant taxpayers, there’s no gain and a likely loss in output by taking workers away from more productive activities.
Administrative reform and simplification, on the other hand, can reduce noncompliance while increasing the time that taxpayers can devote to more productive pursuits. Of course, to achieve this goal, leaders of both parties have to cooperate to make changes that benefit the public, even when it gives them no partisan advantage.
This column originally appeared on TaxVox.