Blog

Jane the Actuary Retirement

Forbes post, “Expand Medicare In The American Families Act? Not So Fast”

Originally published at Forbes.com on April 30, 2021.  While the legislation never passed, I think the background/context for “Medicare expansion” is still relevant.  

 

The American Families Act — Biden’s new spending proposal covering paid leave, child benefits, childcare subsidies, tuition-free community college, and more, but lacks one component progressives had been calling for: an expansion of Medicare, in terms of benefits provided and age eligibilities.

Here’s an excerpt from yesterday’s Washington Post:

“Congressional Democrats are planning to pursue a massive expansion of Medicare as part of President Biden’s new $1.8 trillion economic relief package, defying the White House after it opted against including a major health overhaul as part of its plan. . . .

“They specifically aim to lower the eligibility age for Medicare to either 55 or 60, expand the range of health services the entitlement covers and grant the government new powers to negotiate prescription drug prices. . . .

“Roughly 100 House and Senate Democrats led by Rep. Pramila Jayapal (Wash.) and Sen. Bernie Sanders (I-Vt.) publicly had encouraged Biden in recent days to include the overhaul as part of his latest package, known as the ‘American Families Plan,’ which proposes major investments in the country’s safety net programs. . . .

“Sanders said Wednesday he would ‘absolutely’ pursue a Medicare expansion as lawmakers begin to translate Biden’s economic vision into legislation. Sen. Ron Wyden (D-Ore.), the chairman of the tax-focused Finance Committee, similarly pledged that he would ‘look at every possible vehicle’ to lower drug costs.

“And Sen. Richard J. Durbin (Ill.), the Democrats’ vote-counter in the chamber, said he planned to push for Medicare reforms he saw as a ‘game changer.’”

Democrats and, to a lesser degree, Republicans have been on a quest to reduce prescription drug costs for years, and even in the waning months of the Trump administration, there had been efforts to implement a “most favored nation” pricing model, in which Medicare would pay drug prices equivalent to those paid by other developed countries; an executive order mandating this for Part B drugs was issued in September, published as a rule in November, and halted via injunction from a lawsuit in January. Proposals to expand this concept to Part D drugs are still pending, but have bipartisan approval; more extreme proposals such as the promise to seize patents and produce drugs via compulsory licensing, of course, do not.

But what exactly is the plan, when it comes to demands to expand Medicare eligibility? There have been two versions floated about: the first offers a “buy in” to otherwise-uninsured near-seniors, and the other simply offers the benefit under the same terms as for those already age 65. The letters from Sen. Sanders and Senate Democrats and from House Democrats do not spell out precisely their intention when they call for eligibility expansion alongside benefits expansion. Can it be inferred from their claim that Medicare would save $500 billion over 10 years with prescription drug price-control and the money could be used to “expand and improve Medicare,” that they intend for their pre-65 Medicare to be exactly as “free” as 65+ Medicare? Politico reports that merely to add dental, vision, and hearing benefits would cost $350 billion, but that Sanders, when asked, declined to discuss the costs of his proposals:

“He said he doesn’t want to just pick a target; he wants to check off as much as can be done to help people across the country and then figure out just how much that would cost.”

What is the actual cost of Medicare expansion? Even in a moderate form down to age 60, one estimate is between $40 billion to $100 billion per year, which, at the higher end, or with a further eligible drop down to age 55, would be far, far more than the projected savings from drug price controls, even without the addition of the benefit enhancements.

As to the various buy-in proposals, Kaiser evaluated several of the proposals being floated as of 2018. These proposals keep the Medicare benefit design, with its Part A, B, and D benefits, calculate the cost of benefits plus administrative expenses, and use the ACA structure to provide premium subsidies. The proposals would also require that all existing Medicare providers accept new patients at the same, low, Medicare reimbursement rates as for existing over-65s, and it is this low-reimbursement mandate, not superior management or reduced administrative costs, which would reduce the cost of this coverage to recipients. And for any iteration of “buy-in,” or any of the “Medicare for All” proposals from the various Democratic candidates, the ultimate impact of including far more medical care into the “Medicare rates” is something that generates a lot of worry, without much of a sense, even from experts, as to what would happen when providers are squeezed and cannot cost-shift from Medicare to private-insurance patients.

One analysis from 2019 attempted to model the effects of some sort of Medicare expansion, based on a hypothetical non-profit hospital system, in which private-sector insurance rates are double those of Medicare. This hypothetical did not include doctors’ practices, only hospitals. It found that its current 2.3% profit margin would drop to a 1.6% margin, in the case of a voluntary buy in for those age 50 and above, where employers could not shift coverage to Medicare; if employers did have the ability to move their older employees to Medicare, the hospitals would have a 5.3% loss; for a public option for all ages, that hospital system would have an 8.4% loss. (I was unable to find a similar analysis of doctors’ revenues.)

It stands to reason, then, that, however much the United States may be overpaying when it comes to drug costs, there is no such simple answer when it comes to costs for healthcare, in general, where expanding the number of recipients of care at Medicare rates (or, even more extreme, the Medicaid rates), would open up a can of worms. Proposals exist to reduce the cost of medical care, far more broadly speaking than simply forcing down reimbursement rates or expanding the number of people eligible for the lowest rates, but this is a far greater challenge than sloganeering, or simply declaring a new form of government benefit.

 

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.

Jane the Actuary General Interest, Retirement

Forbes post, “Why Biden’s American Family Plan’s Family Leave – Reportedly – Gets Social Insurance Completely Wrong”

Originally published at Forbes.com on April 22,2021.  The legislation has failed in the meantime, but the issue of Family Leave has not gone away.

Earlier this week, the Washington Post reported on some initial details on the American Families Plan, the third of the Biden administration’s massive spending bills, expected to be unveiled next week, and following on the American Rescue Plan already passed and the American Jobs Plan of infrastructure and social spending. The proposal is expected to include

  • $225 billion for child-care funding;
  • $225 billion for paid family and medical leave;
  • $200 billion for universal prekindergarten;
  • Hundreds of billions in education funding, including his “free community college” campaign promise; and
  • An extension of the expanded child tax credit/child allowance through 2025.

The plan is intended to be funded by tax increases on “wealthy Americans and investors, in addition to beefing up enforcement at the Internal Revenue Service.”

The Post did not provide details on the time frame over which these costs would be incurred or funded. (Recall that the “jobs plan” spends over 8 years and funds over 15.) But one might assume that his proposal for family and medical leave would be based on his campaign promise, that is,

“Biden will create a national paid family and medical leave program to give all workers up to 12 weeks of paid leave, based on the FAMILY Act. Workers can use this leave to care for newborns or newly adopted or fostered children, for their own or family member’s serious health conditions, or for chosen family; or to care for injured military service members or deal with “qualifying exigencies arising from the deployment” of a family member. During their time away from the job, workers will receive at least two-thirds of their paycheck up to $4,000 so they can better afford to take leave — with low- and middle-wage workers receiving larger shares of their paycheck.”

Now, the FAMILY Act is an existing legislative proposal which would provide 12 weeks of leave at 66% of pay, paid for by 0.2% of pay (employer/employee) contributions. And as it turns out, this proposal has been around since at least 2013, when on another platform I critiqued the bill not just for its egregious acronym (the bill title is the Family and Medical Insurance Leave Act, which ought to be the FAMIL Act, and even then “insurance leave” rather than “leave insurance” doesn’t even make sense, but it’s rearranged to get the acronym) but for the fact that the contribution level appear to have been chosen as a politically palatable tax rate rather than based on any actuarial analysis of the cost of running such a program. In fact, in January 2020, the Social Security Chief Actuary provided an analysis of the bill’s cost, and found that rather than the proposed payroll tax rate of 0.4% would be insufficient and instead 0.62% would be required to fund benefits — and that under a surprisingly low set of assumptions around use of the benefits, that only 35% of new parents would take advantage of the program, that 4% of workers would have medical conditions of their own and 0.4% of workers would need to care for a family member, and that, on average, they would receive benefits for only two rather than three months. Yet nowhere in the legislation is there any means of adjusting the payroll tax to meet actual financing needs, nor adjusting benefits to meet the revenue available.

Financing issues aside, however, this approach is a reasonable form of social insurance provision. Quite simply, this is how social insurance works. We, collectively, want a means of collectively providing funds to people as circumstances require — retirement, disability, unemployment, family leave needs — and the mechanisms of social insurance deliver: universal payroll taxes with rates set as needed to fund these payments.

Consider a few international examples, based on the data available at the International Social Security Association website (and highly simplified):

France provides sickness benefits and maternity/paternity/adoption leave, for up to 10 weeks after birth, funded with a 13.3% payroll tax, plus family allowances funded with a 3.45% payroll tax, which include both cash benefits, childcare subsidies, and benefits for reduced work hours, paid for 2 years, more or less.

Netherlands provides a maternity leave benefit for up to 16 weeks, funded through its unemployment insurance program (2.85% payroll tax).

Sweden’s benefits have a parental benefit-specific payroll tax of 2.6% of pay as well as a 4.35% payroll tax for sick leave.

Germany provides maternity and sick leave with a payroll tax that’s combined with the costs for medical benefits, at 7.3% for each of employees and employers, for up to 8 weeks. Additional benefits, including a child benefit and a one-year 67% parental leave benefit, are funded out of general revenues.

And — as a reminder — when other “western” countries do fund their benefits from general revenues, this means they fund them from income taxes in which there is no special effort to “soak the rich” but instead their top tax bracket applies, generally speaking, to everyone middle-class (or upper-middle-class) and above.

This means that Biden’s proposal to fund family leave through a tax hike on the wealthy is not just ill-conceived but far from the international norm. And it won’t get us where its supporters want us to be. If we define social insurance not as something we all pay for and benefit from, but as government benefits “paid for by other people,” it will create a dead-end that will impact the prospects for social insurance, generally speaking — including reform of Social Security itself.

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.

Jane the Actuary General Interest, Retirement

Forbes post, “Would Baby Bonds Save Retirement? Be Careful Of Quick Fixes And No-Cost Solutions”

Originally published at Forbes.com on February 24, 2021.

 

What’s a Baby Bond?

The name implies borrowing of some sort or another, as if a parent might borrow money and pay it back later. But the concept is really quite different, a sort of “universal personal account.” There are two different proposals circulating at the moment: one is meant to provide a sort of start-up fund for adult life, and the other, a retirement benefit. But they both are based on the idea of building up an account balance from government contributions while children are young.

The Booker Baby Bonds

Sen. Cory Booker of New Jersey introduced legislation earlier this month to establish what he calls “opportunity accounts.” The bill involves an initial “contribution” of $1,000 for each child at birth, and as much as $2,000 per year afterwards based on household income, with the full amount for families with income of less than 100% of the poverty line, phasing out to $0 at 500% of poverty rate; each of these dollar amounts would increase with the rate of inflation from year to year. These are described as “savings accounts” but there is one large fund, funded by the federal government, with individual account balances tracked nominally within that fund.

And — this makes my head hurt a bit — even though the money for the fund comes from the federal government, and even though, in the year 2021, like pretty much every other bit of government spending, the government would borrow the money to put in this fund (or money-print the money), the money is then “lent” back to the government, in the form of “special interest-bearing obligations of the United States,” with an interest rate equal to the average market yield of all US public debt with a maturity of greater than 10 years. According to Politico, this would be a 3% interest rate, but it’s not clear to me how Politico came up with that rate when a 20 year Treasury is currently 1.63% and even a 30 year Treasury is only 2.19% at the moment. At the same time, the bill implies that there are possibilities of investment losses, but perhaps there were drafting inconsistencies and versions of the proposal that involved equity investments?

So is it “real” money, or is it all a bit of a fiction? Like the Social Security Trust Fund, it’s both, or neither.

And this money could be spent in a limited number of ways: on higher education expenses (including trade school), on the purchase of a home, “any other investment in financial assets or personal capital that provides long-term gains to wages and wealth” (presumably this is meant to allow for investments in starting a small business), or any spending after age 59 1/2. In any case, this is all money that would be available on top of any financial aid rather than taken into account in calculating eligibility for financial aid.

Would this work?

Various proponents have claimed that this would have all manner of beneficial effects, and in particular would “narrow the racial wealth gap,” as the Urban Institute says, because black and Hispanic families are more likely to be poor, and white families are more likely to have inheritances from earlier generations. The cost would work out to $60 billion annually, according to Booker, or $650 billion over a decade according to the Committee for a Responsible Federal Budget, which reports on a prior iteration in which the plan would have been paid for by increasing capital gains taxes and estate taxes. At the time of their analysis, in 2019, 30 year Treasury rates had been level at 3% for two years, so they used this as their assumption to calculate that a child whose family had been in poverty for his/her entire lifetime would have an account of $46,215 at age 18; for a more middle-class family, with income of $56,000 for a family of four, the account balance would be $12,815. (My own math is somewhat different and arrives at an account balance of $38,000 adjusted to strip out the effect of inflation over time, but, eh, you get the picture.)

Would this achieve the goal of the program, helping young adults who had grown up poor get a good start in life? It seems obvious that having this money, over and above what financial aid programs would provide, would increase their financial well-being. But what is the purpose of building up an “account,” especially a notional account, over time? Why create a program which won’t provide any benefits for the next 18 years, when there is no particular rationale except to implement the spending gradually, so that only 18 years from now will the program be operating at full cost? Since all the money is borrowed anyway, what meaningful significance is there to the interest accrual and the lending-back of borrowed money for the purpose of “earning interest”?

And once a recipient leaves young adulthood, the notion of using these accounts as retirement income becomes shaky, because it makes so little sense to require that they continue accruing interest at the very low level of a government bond.

So what is the point of a “Baby Bond” rather than simply beefing up financial aid for education and grants for would-be entrepreneurs? One presumes that there’s a strategy in allocating the funds $2,000 at a time, and with $1,000 for all children regardless of income, to give it the “feel” of a broad-based program and a “savings account” just like the savings accounts we might encourage our children to start for themselves.

The RISE account – a Baby Bond for retirement

There is a second “Baby Bond” proposal which aims much more directly at retirement, because, in this version, the money isn’t accessible at all until retirement age. This proposal comes from financial advisor Ric Edelman, with a description at, among others, the Washington Post.

Here’s how his proposal would work:

The government would issue bonds and use the proceeds to fund accounts for children born in the US, just as with the Booker proposal. However, the bonds would be issued and the funds credited upon a child’s birth, be invested in diversified funds similar to a pension fund. Then, after 20 years, the bonds would be redeemed, but the investment earnings would be greater than the initial bond and interest, so this net earnings would remain in the fund and continue to grow for the next 50 years, at which point the individual would begin to receive retirement benefits in a spend-down using financial planners’ rules of thumb, and any funds remaining upon their death would be used to supplement funds for those who outlive their assets by living past 100.

Edelman does not provide the full details of his plan, and the biggest missing piece is his math around which children would get how much money. Similar to Booker’s plan, babies would be credited with different amounts based on parents’ income, with an average funding of $5,884, which he projects would create an annual payment stream of $26,810, in today’s dollars.

But his two key assumptions are these:

First, he assumes that the government would be able to issue bonds at the low interest rates earned by traditional “savings bonds” (EE bonds), which stands at 0.1%. (It is not clear if he builds into his calculation the one-time adjustment at 20 years which means that savings bonds double the saver’s money, for an annualized rate of 3.5%.) But the program would need to issue bonds of over $23 billion per year, by his own numbers ($5,884 per child and 4 million children born per year); however, the total series EE bond outstanding as of November 19 was $125.8 billion. He assumes that enough investors will choose these savings bonds for the “feel-good” element of the program, but that seems a bit of a stretch over the long term, especially when you’re no longer speaking of helping poor children but funding retirements 70 years in the future for people who may or may not be poor as adults.

Second, he assumes that the program will reliably earn the projected 7.27% in investment earnings. It’s important to note that this is the average assumed rate of return reported by public pension plans in the United States, which is, let’s face it, just about the most aggressive such assumption available — and this isn’t just a matter of the usual accusations of public pensions being to aggressive in order to report lower liabilities and make lower contributions. In the same manner as interest on government bonds is a fraction of what it had been a generation ago, so, too, are expectations for stock market returns declining as the population ages; even before the pandemic or the Trump administration, McKinsey was forecasting long-term declines in U.S. and European equity growth (see this December article on another similar proposal); more recently, Morningstar summarized long-term expectations coming in at anywhere between flat/negative to 5.7%, for U.S. equities, at best. Considering that it would be necessary to beat a 3.5% annualized return, on average, for 20 years, this is not actually a sure thing.

And Edelman acknowledges that the proposal is not truly risk-free, and suggests that “Congress could choose to give the Commission authority to use a surplus from one year’s program to compensate for a deficit incurred by another. Alternatively, Congress could choose to provide funding to eliminate a shortfall experienced by any birth year. Absent these actions, babies from each birth year will simply be limited to receiving only as much income as the program can provide.”

Finally, would it be fair to provide this retirement income only to those children who had been born into poor families? Edelman himself had made an earlier proposal with a flat benefit for all.

Would this proposal work?

My first reaction is that there’s simply no way to implement a plan which would first start to pay out benefits in 70 years, and that Americans further wouldn’t stand for there being differences in payouts by year of birth from what’s nominally a government program for all.

But looking at the core idea, rather than the specifics: this is a proposal for a sovereign wealth fund. Sure, it has a specific objective — providing retirement savings — and yes, unlike the sovereign wealth funds of nations with oil wealth, it would be funded by issuing debt, and its meant to be managed by private-sector investment managers, but in its core concept, functionally, the government would be acting as private-sector investor and earning money with the expectation that it investment revenue would exceed the interest on bonds.

And, in fact, there have been multiple proposals circulating recently. Most recently, in January, in “It’s Past Time for a U.S. Federal Sovereign Wealth Fund” at The National Interest, Richard Caroll proposed that the royalties the U.S. government receives from wealth-extraction on federal lands (with those royalties to be hiked substantially as well), as well as revenues from the Federal Reserve Open Market Operations’ bond purchasing program, should be directed into at U.S. Sovereign Wealth Fund rather than merely the U.S. Treasury, for the purpose of creating investment income as a future income stream given the growing indebtedness of the country. Back in May, Nir Kaissar at Bloomberg argued that instead of bailing out companies, the U.S. should have bought their equity or bonds:

“Like any private investor, the fund would buy the equity or debt of struggling companies at deeply discounted prices, handing the losses to owners or creditors of those companies and imposing the same onerous terms one would encounter in private-sector deals. That would remove questions about whether and which companies deserve to be bailed out because investment would be motivated by profit rather than charity. It would also allow the government, and all Americans by extension, to fully participate in the gains when companies turn around.”

The question of issuing debt to make investments would not be relevant, in this case, because the government would be borrowing to fund the bailouts, in any case, though Kaissar also suggests borrowing additional money to build up the fund even further.

As it happens, we have an example of a retirement-oriented Sovereign Wealth Fund just to the north of us: the Canada Pension Plan has a trust fund which invests in a wide variety of assets. As of 2018, they fully or partially owned Petco, Univision, and Neiman Marcus, for example. But they did not borrow money in order to make these investments; they used the surplus income coming into their Social Security system and are now using the funds generated by a tax hike to build up funds to fully pay for new benefits which will accrue over time. In other words, they are not promising that a plan will “pay for itself” but simply using tax revenues to build up a pension fund.

The long and short of it is this: people want a solution to retirement shortfalls (may I remind you of my Social Security reform proposal?) and to persistent demographic disparities, and this offers the promise of being something new and innovative. But just because something is “innovative” doesn’t necessarily guarantee it’s the right path forward.

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.