Wage replacement is an important measurement, whether or not it's good policy

Become a Patron!I've written before about the FAMILY Act, the leading Democratic initiative in the House and Senate to create a nationwide system of paid family and medical leave like that enjoyed by citizens of virtually every other country in the world.One of the things advocates say when describing the FAMILY Act is that it's based on a "social insurance model." This is a meaningless phrase if you're not a policy advocate, but in its simplest form, it means that, just like you can only insure a house for the house's value, or take out a life insurance policy up to some multiple of your earnings, the taxes (or "premiums") you pay while working are based on your income, as is the benefit you receive while caring for a new child, family member, or yourself. The direct parallel is to Social Security's old age benefit, where higher-income taxpayers pay more in Social Security taxes, and receive correspondingly higher benefits when they retire.The social insurance model is therefore based on the idea of "wage replacement:" what percentage of your wages should the program replace during qualifying life events? Social Security, for example, uses an extremely complicated weighted average of your highest-earning 35 years, then applies different replacement rates to income in different brackets up to the maximum. The FAMILY Act relies on its own formula, slightly more similar to unemployment insurance benefits, based on your highest earning year in the previous 3 years, up to an arbitrary cap.

There are serious arguments for a wage replacement model

Since at the end of the day I disagree with the wage replacement model, I want to give the absolutely fairest, most compelling arguments for it — and they are compelling!

• Wage replacement encourages higher-income workers to take leave. This is plainly true. If the goal of a paid family and medical leave program is to reduce the cost to workers of using the leave they're entitled to, then the benefit should scale up to the income they sacrifice when taking leave. A new parent with a high income is by definition sacrificing more money when they take leave than a new parent with a low income; if their benefit doesn't reflect that, they'll be somewhat less likely to take leave, or take as much leave as they're entitled to.
• Wage replacement reflects regional living expenses. If differences in wages do not reflect different levels of education or experience, but rather different regional costs of living, then giving higher benefits to higher earners simply reflects those regional differences. A flat benefit would punish leave-takers in high-wage, high-cost-of-living areas, and be a windfall to leave-takers in areas where wages and prices are lower.
• Wage replacement promotes gender equity. There's both a crude and sophisticated version of this argument, so I want to offer both. The crude version is that, at an economy-wide level, male employees earn more money than female employees, and a flat benefit will perpetuate those wage disparities: if either parent can take leave and receive the same benefit, a given family unit will be financially better off if the lower-earner takes the leave and the higher-earner continues to work. In many relationships, that means a female caretaker and a male worker, which is not a pattern the federal government should be explicitly subsidizing. The sophisticated version is that within any individual relationship, whenever there's a higher-earner and a lower-earner they face the exact same choice. Mitigating, although not eliminating, that power dynamic is an important argument in favor of wage replacement regardless of the gender of either worker.

These are serious arguments, and they have convinced the paid leave advocacy community of the advantages of a wage replacement model.

Wage replacement is a complicated, intrusive policy design

I do not think wage replacement is a good policy design, but I understand I have lost this fight, so I will only briefly summarize the problems I see in all such policies:

• It rewards accurate income reporting, and magnifies the existing pain to the low-income workers most vulnerable to underreporting. A tipped worker whose tips go unreported or underreported now loses out on both Social Security old age benefits and the paid family and medical leave benefit they're entitled to.
• It rewards high-income workers in low-cost states. This is the inverse of the second point made above: while some amount of wage differentials are accounted for by regional differences in costs of living, others are not. The more convincing you find the "regional wage-and-cost differences" theory, the less willing you should be to offer wage replacement to a high-income worker in a low-cost state. If $60,000 per year represents poverty wages in Manhattan, but a middle class income in Nebraska, then an identical wage replacement rate will impoverish the New Yorker while the Nebraskan barely notices!
• The inscrutable means-testing formulas make it impossible to know what your benefit will actually be. A flat benefit allows people to plan their expenses around their income. A wage replacement model based on a rolling 3-year period simply makes it impossible to know what your benefit will be. Not because it's impossible to calculate (the Social Security Administration has a nice website showing what your disability and old age benefits will be, based on all the information they have available), but because it's constantly changing. If you get pregnant in July and plan to give birth in March, will your benefit be based on that really good sales year you had 3 years ago or will that year already have rolled off your benefit schedule by the time you take maternity leave?

These are my basic arguments against a wage replacement model, and all forms of means-testing, but as I say, I've lost that fight and have come to peace with it.

Whether or not wage replacement is good policy, it's a good measurement

No one has to agree with me on the downsides of the wage replacement model. I don't always agree with myself! But whether or not wage replacement is good policy design, it's an essential lens when looking at the bevy of alternatives to the FAMILY Act bubbling up in Congress. Precisely because those policy alternatives are so different, you need a set of uniform tools to judge them by.For example, the "bipartisan" Cassidy-Sinema bill offers new parents the ability to claim a present-year $5,000 additional child tax credit, repaid by reducing the child tax credit available in the next 10 years by $500 per year. While this is idiotic program design, the idiocy is not the point of this post. The wage replacement rate is.In the ideal case, a two-parent family claiming the Cassidy-Sinema tax credit would receive a total of $5,000 in the year of their child's birth or adoption to assist them with their expenses during their 12 weeks of federally-guaranteed FMLA leave. The median annual household income in the United States is approximately $63,179, 23% of which (reflecting 12 weeks of leave) works out to $14,579. A $5,000 benefit — even ignoring the fact that it has to be repaid — represents a wage replacement rate of about 34% for the median household.Whether or not you think a paid family and medical leave benefit should be linked to a given individual's or household's income, or be offered as a flat benefit based on qualifying events, you cannot believe that a 34% wage replacement rate is anywhere close to adequate.

Conclusion

The point of this post is not that a social insurance or wage replacement model is the right or wrong way to design a paid family and medical leave policy. Rather, it is to give you an easy, back-of-the-envelope calculation you can perform whenever you hear someone promoting an alternative to the FAMILY Act, which is the only serious attempt to enact such a policy in the United States:

• divide the length of the benefit, in weeks, by 52
• multiple the result by $63,179
• divide the benefit by the result.

That's the wage replacement rate for the median household in the United States. The more valuable you think paid family and medical leave is, the higher you should want that replacement rate to be, and the less valuable, the lower.Likewise, the 66% wage replacement rate in the FAMILY Act is the only viable solution to the cancer of unpaid family and medical leave in the United States. The time to fight about how it can and will be improved is after we've passed it.Become a Patron!

Biden is wrong about Social Security, but so is everybody else

Become a Patron!In a bit of a surprising turn in the lead-up to the Iowa caucuses, former Vice President and Delaware Senator Joe Biden has finally come under sustained criticism for his career-long advocacy for cuts to Social Security.I'm not sure why his rivals didn't elevate this criticism earlier, but it's obviously correct: he has advocated and voted for cuts to Social Security. During his second term in the US Senate, he voted for the Social Security Amendments of 1983, which increased the full retirement age gradually from 65 to 67, a cut in benefits for the majority of workers who claim their old age benefit at age 62 (by increasing the reduction in their old age benefit), and also cutting benefits for those who delay their retirement until age 70 (by reducing the number of delayed claiming credits they earn).I know my readers are more sophisticated than average, but I want to review this because it's important to understand the mechanism at work here. Workers with 40 Social Security credits (roughly 10 years of reported work history) are:

• eligible to claim a reduced old age benefit at age 62;
• eligible to claim the "primary insurance amount" of their old age benefit between age 65 and 67, depending on their year of birth;
• and eligible to claim an increased insurance amount between their full retirement age and age 70.

Those with a full retirement age of 65 filing at age 62 received a maximum reduction of 3 years (if they filed at age 62) and a maximum increase of 5 years (if they filed at age 70). By raising the full retirement age to 67 for my generation, the Social Security Amendments of 1983 increased the maximum reduction to 5 years, while reducing the maximum increase to 3 years.But this has no connection whatsoever with the age at which people stop working, or increased life expectancy, or improved health, or anything of the kind. It was purely a cut to Social Security old age benefits: those who claim their old age benefit at age 62 receive less money, and those who claim their old age benefit at age 70 receive less money. Everybody receives less money, because it's a benefit cut, which Biden advocated for and voted for.

Social Security is already "means-tested" at the top

One of the strange defenses that has been floated for Biden's career-long support for cuts to Social Security is that wealthy, high-income people don't depend on Social Security in retirement, so "means-testing" Social Security old age benefits is a common-sense way to shore up the program's finances. Here, we run into a problem, since Social Security old age benefits are already means-tested for high earners.Old age benefits are subject to a flat and arbitrary cap. This is the purely mechanical result of the limit on earnings subject to the Social Security portion of the payroll tax. The Social Security Administration makes it deliberately difficult to calculate this, but if you fiddle around with the "Quick Calculator" long enough you can get it to spit out the facts: someone reaching age 62 in 2020 with a 35-year history of maxing out their Social Security earnings is eligible for the following old-age benefits.That's because the annual amount of earned income subject to the Social Security portion of FICA, and included in the old age benefit calculation, is capped at the amounts shown here (35 years):So the maximum old age benefit to the highest-earning American, in 2027, will be roughly $47,364 per year. That's more than I make, but it's hardly a princely sum.

Social Security is also "means-and-compliance-tested" at the bottom

In the Year of our Lord 2020, we all know what means-testing really means: cutting off benefits for those who can't or won't comply with increasingly onerous reporting requirements. Fortunately, if you're a sadist, Social Security old age benefits are also means-tested at the bottom. To earn the maximum of 4 annual Social Security credits in 2020, you need to earn about $5,640.Or, more precisely, you or your employer needs to report about $5,640 in earnings. If your employer fails to report your earnings? No credits. If your employer misclassifies you and you fail to report your earnings? No credits. Whatever you think about the quality of wage-and-hour compliance in the United States, no one thinks that it comes close to 100%.Even worse, the compliance rate is lowest for the lowest-income earners, which is especially problematic for a system like Social Security in which old age benefits can only increase until 35 years of earning history are accumulated (at which point the lowest-earning years start to roll off). As I've described in the past, since reported earnings are wage-inflation-adjusted, the earlier you earn them the more they contribute to your earnings history as they have more time to appreciate along with wage growth.

Social Security is also "means-tested" in the middle

So high-earners already have their old age benefit capped by the earnings limit, and low-earners have their old age benefit restricted by their employers' compliance with wage reporting. What about those in the middle?They get to enjoy the phased-in taxation of Social Security! It turns out, 0%, 50%, or 85% of your old age benefit may be taxable, depending on your adjusted gross income. As I've explained before, just because your old age benefit is taxable doesn't mean you owe any tax on it — it may be taxable but the tax owed is $0.For those with some earned income, some taxable income, and some required minimum distributions from IRA's and 401(k)'s, their old age benefit is already "means-tested:" the more they earn from those sources, the more they owe in taxes on their old age benefit.

This is, obviously, madness

There's no secret if anybody came along and felt like "fixing" Social Security:

• Remove the earnings cap on the Social Security portion of FICA;
• Make capital gains and other business income subject to FICA;
• Improve enforcement of and compliance with wage and hour laws;
• and make all Social Security benefits subject to ordinary income taxes.

But nobody wants to "fix" Social Security. The entire project of Social Security "reform," from Joe Biden's vote in 1983 to the present day, is driven by the desire to cut benefits, because the people pushing reform don't believe in Social Security.But if you do believe in Social Security, all you have to do is keep fighting for it, now and forever.Become a Patron!

What the media can't get right about retirement, and why

Become a Patron!A long-time reader passed along an article from the World Economic Forum's blog (yes, for some reason the World Economic Forum has a blog) titled "To retire at 65, American millennials need to save almost half their paycheck." While the article is absolute drivel, it's a good starting point to understand some things the business and financial press gets consistently wrong about savings.

The article itself makes no sense

The extremely thin premise of the article is easily dispatched with: relatively high stock market prices today suggest that future returns will be relatively low. That means projecting today's return on savings forward means today's workers will have to fund a relatively large portion of their retirement spending out of savings and a relatively small portion of their retirement spending out of dividends and price appreciation.I will make a simple observation about this thesis: retirement savings are accumulated over time, not with lump-sum contributions. The idea that "high prices today predict low returns tomorrow" only applies to this year's contributions. Next year's contributions will earn returns based on next year's prices, and contributions made 10 years from now will earn returns based on prices 10 years from now. In other words, assuming steady or rising contributions, today's prices tell you nothing about the lifetime savings rate required to achieve a given level of income replacement in retirement.To give a simple example, the 30-year Treasury bond rate reached a high of about 14.8% in 1981. Since $1 invested at 14.8% for 30 years yields $62.84, using the World Economic Forum's logic, a 35-year-old in 1981 could have invested just 0.8% of their salary each year in order to replace 50% of their pre-retirement income, since 30 years later, that 0.8% would have swollen to 50%.The problem is that 10 years later, the 30-year Treasury rate had fallen to below 8%, and the same 0.8% salary savings would replace just 8% of their salary 30 years later. The World Economic Forum is making the primitive mistake of projecting this year's expected returns forward to all future years.With the Social Security full retirement age already raised to 67 for my generation, even the oldest millennials have 25 or more earning years ahead of them; of course it's possible today's high stock market prices will continue for the next 25 years, but let's just say I have my doubts. And if we do have 2, 3, or 4 crises, recessions, or depressions between now and then, there will be plenty of opportunities to buy assets with expected returns just as high as they are low today.

The financial press doesn't get it: no one saves anything

In some ways it's unfortunate that we use the same verb in two radically different, almost opposite meanings: to "save," and to "save up for." To "save up for" is the most common and natural thing in the world, and virtually everybody does it. Whether you're saving up for Christmas presents, a new console game, a down payment on a house, or an anniversary dinner, the idea is the same: you set money aside every week or month, for a known or unknown period, and then you spend it, and it's gone. This is not fundamentally different from buying something on credit, it's just cheaper: with credit you get the object of your desire up front, then pay for it over time, while with savings you pay for it over time, then get it at the end.But "to save," let alone "to invest," means something totally different. It means setting money aside, for an indefinite period of time and for no definite purpose. This behavior is vanishingly rare because upon a moment's reflection, it makes almost no sense.First, setting money aside today comes with an obvious cost: that money could be spent instead. It's one thing to save up $300 for a new gaming console; it's another thing entirely to have saved $300 and then, instead of buying a new gaming console, keep saving instead!Second, the indefinite period creates obvious risks. I'm not talking about investment risks, since even most people who do save don't invest their savings. I just mean the risks of doing anything over time. Maybe you'll die before you ever spend any of your savings — you can't take it with you. Maybe you'll get divorced and "lose" half your savings to your former spouse. Maybe word will get around that you have savings and your friends and relatives will come around asking for a "loan."To put it as simply as possible, if saving were easy, the government wouldn't have to go to outlandish lengths to encourage people to do it. And despite those outlandish efforts, the only people who save are still the same people who would do it anyway: high-income professionals. Don't get me wrong, they're happy to take the government handouts, but the handouts are pure gravy.

The fantasy of the retirement "number"

Let's turn back to the underlying premise of the World Economic Forum post: the goal of retirement saving is to replace a certain percentage of your pre-retirement income by setting aside money during your working years. If you can project forward both your earnings trajectory and your investment returns (whether on a year-by-year basis or using the "starting year" fallacy described above), then you can determine the total balance you need to have in your accounts on the day you stop collecting a paycheck.But who could this model possibly describe with any degree of accuracy? The majority of people start collecting Social Security old age benefits on the day they become eligible, and have no savings or an amount that is simply inconsequential to their long-time retirement spending needs. Using the 4% rule, someone who has managed to accumulate the fantastic sum of $100,000 at retirement can only safely withdraw $333 per month to supplement their Social Security benefits, and if the money is kept in a risk-free account, the safe withdrawal rate must by definition be even lower.On the other hand, savvy high-income professionals pour vast sums into their pre-tax retirement accounts, 529 accounts, Roth accounts, HSA's, after-tax 401(k) accounts, and these sums accumulate so quickly they dwarf any conceivable financial needs in retirement.What the "retirement number" presupposes is that there's some group of people in-between, who can afford to save more than nothing, but less than the absolute maxima allowed in tax-advantaged accounts. And this group no doubt exists: there are a lot of numbers between $0, the amount saved by most people, and $19,000, the amount contributed by high-income professionals to their 401(k) plans in 2019, so in a country of 350 million souls there are certainly people contributing $1,000 and people contributing $18,000.Unfortunately, the first $1,000 you save does nothing for your retirement security (you'll still rely on Social Security old age benefits), and the last $1,000 you save does nothing for your retirement security (you'll retire wealthy no matter what).

Finance journalism is an ideological project

What all the above should make clear is that the relentless focus on this sliver of the population, the "in-betweeners," is not about personal finance, it is about ideology. Tax-advantaged investment accounts obviously provide the overwhelming majority of their tax benefits to the people who can afford to contribute the most money to them, and no benefit at all to people who can't afford to contribute anything.If retirement income security, or affordable higher education, were really public policy goals, then the money currently spend shielding the investments of the wealthy from taxation would go a lot further in boosts to Social Security benefits and public subsidies for higher education.But if you can create a mile-wide, inch-thick class of people who have contributed "something" to an IRA, 401(k), 403(b), or 529 plan, then you suddenly have tens of millions of people mobilized to protect those programs for the wealthy despite the trivial benefits they receive from them, and the ideological basis for cutting Social Security benefits and public higher education funding even further: anyone without sufficient retirement income should have raised their retirement contributions faster; anyone who can't afford to send their kids to college should have opened that 529 plan earlier.After all, the financial press was there all along telling you to, so you have no one but yourself to blame.Become a Patron!

California's AB 5 is about benefits, not wages or unions

Become a Patron!I've been following with interest the development and passage of AB 5, California's attempt to end the misclassification of employees, especially at app-based and platform companies including household names like Uber and Lyft.What I found is that there's a high degree of confusion surrounding the need for the bill, whom it will affect and how, and why the implicated companies have been so strongly opposed to it.The key to understanding the issue is that it's not about wages or unions; it's about benefits.

Uber isn't afraid of a union

One common idea people have about the employee classification debate is that it has "something to do with unions." And indeed, it does have something to do with unions! Union leaders in California have spearheaded the drive to end misclassification of platform employees, and the final passage of AB 5 will be a triumphant display of their ability to turn out activists and corral legislators.But it doesn't have anything to do with forming unions. The reason is simple: it's virtually impossible to form a union in the United States, and the distributed nature of platform employees is guaranteed to make it impossible in fact.That's because US labor law is based on the organization of similar workers in specific workplaces. In the mid-20th century that might be all the assembly line workers in a particular GM plant (but not the bookkeepers at the same plant), or all the electricians on a particular construction site (but not the plumbers).The recognition of workers as employees of platform companies will give them the theoretical right to collectively bargain — but only after their bargaining unit has been recognized, and Uber has unlimited tools at its disposal to make that process as onerous as possible:

• First, it will certainly insist that employees who provide rides belong in a different bargaining unit than employees who deliver food, who are different from the employees who repair scooters, who are different from the employees who charge scooters.
• Second, it will certainly insist that the relevant "workplace" is a single city, at the very largest. If you've ever taken a look at a map of "the San Francisco Bay Area" you quickly realize it's not one city, but dozens. Uber will insist its employees organize separately in each city.
• Next, they'll insist that workers will only be included in a bargaining unit if they work at least 20, or perhaps more, hours per week in that workplace, and they'll almost certainly insist that the only hours that count "in the workplace" are hours spent picking up in the relevant workplace. If you drive a passenger from San Francisco to Oakland, and another passenger back, only the outbound ride will count towards your membership in the bargaining unit.

Note that I came up with these obstacles in 10 minutes. Uber will employ the best union-busting law firms in the country to come up with many, many, many more obstacles.It is, of course, possible that there are cities where particularly industrious union organizers will meet all these tests. Unified metropolitan areas (without multiple divisions into separate municipalities) like Detroit would be good candidates.But I want to be very clear: there is no chance more than a trifling number of Uber drivers will ever belong to a collective bargaining unit, and I'd be extremely surprised if the number is ever more than 0, if for no other reason that if a union is ever formed, Uber always has the option to simply leave the market, which is entirely legal under our labor laws (there are a few trivial exceptions but, obviously, Uber will get around them).

Uber isn't afraid of higher wages

Another thing people are understandably bewildered by is that, where Uber drivers have been able to organize into local advocacy groups, Uber has not hesitated to meet with them and has made all sorts of concessions. In August Uber and Lyft even offered drivers in California a $21-per-hour minimum just-don't-call-it-a-wage, far above California's state $12 minimum wage, if they would oppose AB 5.If you thought AB 5 was primarily about wages, this would be extremely confusing behavior. Why would a company offer its workers a wage higher than the minimum wage in order to prevent them from being subject to the minimum wage?

It's all about the benefits

It shouldn't be much of a surprise at this point, since it's right there in the headline, but the reason these so-called "platform" companies oppose properly classifying their employees is that US labor law, for all its shortcomings compared to the rest of the developed world, does offer one nearly-airtight protection: non-discrimination in the provision of tax-deductible employee benefits.In short, this means that with a few exceptions, for an employee benefit to be tax-deductible, it has to be provided on more-or-less equal terms to more-or-less all full-time employees. And the problem with running a hot startup tech company is that if you want to attract talent, you need to offer pretty generous benefits, among the most important of which is health insurance.Under their current "fissured workplace" model, platform companies are able to provide tax-deductible health insurance exclusively to their white collar workforce. Let us very delicately and preliminarily guess that workforce is, on average, younger, healthier, and more educated than their driver and delivery employees.Of course, their driver and delivery employees mostly don't go uninsured. Thanks to the Affordable Care Act, in California the uninsured rate was just 7.2% in 2018. How did California achieve this? Through massive federal expenditures, of course: to expand Medicaid to cover those below 135% of the federal poverty line, and through refundable federal tax credits to cover the cost of insurance purchased through California's ACA exchange.So you can see, it's irrelevant whether Uber does or doesn't "care" whether its driver employees have health insurance or not: rather, it's that Uber wants to offer a generous employer-based health insurance plan to attract the white-collar employees it believes it needs to succeed, while federally-financed ACA exchange plans are "good enough" for its drivers.The misclassification of employees as independent contractors, in other words, is targeted directly at the only real legal protection American workers have: the principle of non-discrimination. You don't have to offer generous health insurance or retirement benefits, but if you do, you have to offer them to all your employees on equal terms — and Uber doesn't want to.

Conclusion: who is flexibility for?

There's a final point I want to make that's somewhat abstracted away from the nitty-gritty of labor and employment law. In all these debates over misclassification the platform company lobbyists and PR goons are always quick to point out that "drivers have the flexibility of deciding when to work." And of course, in a hyper-literal sense this is factually true. Uber cannot force drivers to open their phones and log into the app, so drivers have a choice of whether and when to do so.But while drivers may have control over selecting the hours they're logged into the app, they don't have any control over user demand for rides, or the process of pricing and assigning those rides, and so they don't have any control over the amount they're paid. In this way, they resemble nothing so much as an Applebee's waitress told to clock out and wait in her car outside when business is slow, then clock back in for the dinner rush.Platform companies rely on "flexibility" in order to spin up the availability of workers during periods of high demand. If they couldn't do so, wait times would stretch out of control and users would migrate to other services (or, God forbid, traditional taxis and delivery services). In other words, companies get paid for their worker's flexibility; they rely on it for their very survival, such as it is.But worker's don't get a symmetrical payoff. If a driver's "flexibility" doesn't match up with the company's needs, the driver just doesn't get paid. Again, this isn't unusual across the employment landscape: if you're only available to work at a grocery store between midnight and 5 am, and the grocery store closes at 10 pm and opens at 8 am, you're not going to get hired. Your availability doesn't match up with the store's hours!The platform companies, on the other hand, would appreciate it if you stood outside the grocery store anyway, on the off chance that today is the day they decide to open at midnight instead. There's nothing wrong with asking people to show up even when they're not needed and there's nothing for them to do. But we already have a perfectly good word for those people: employees. And unlike independent contractors, employees still have a few rights left.No wonder Uber is scared.Become a Patron!

Job protection, wage insurance, and universal benefits

Become a Patron!As readers may have observed, I've been doing a lot of research lately on paid family and medical leave policies around the country, and I've found myself frustrated by the way several different ideas are confusingly and unnecessarily combined. Today I want to spell out the relevant issues, and try to explain how they do and don't interact with each other.

Job protection

The most basic protection workers can be provided for family and medical leave is job protection. That's because even if someone can afford to go without pay while recovering from childbirth, bonding with a child, or caring for themself or a relative, no one can afford to do so without knowing they have a job to return to.Currently, the only form of nationwide family and medical job protection is the Family and Medical Leave Act of 1993. For qualifying events, the law provides:

• 12 weeks of leave in a 12-month period;
• to employees who have worked for at least 12 months, and at least 1,250 hours in the previous 12 months, and 
• at private-sector employers who employed 50 or more employees in 20 or more workweeks inthe current or preceding calendar year.

Note that the qualifications are non-transferrable. In order to qualify for job-protected leave, it's not enough to be continually employed for 12 months and work 1,250 hours in the previous 12 months: you have to be employed for 12 months at the same employer. That means workers with working two jobs with 20 hours per week at each will never be eligible for job-protected leave from either employer, since their 1,040 hours of work each 12 months leaves them short of the 1,250 minimum.When we isolate job protection in this way, we can imagine all sorts of possible improvements:

• Increase the quantity of job-protected leave. If we think 12 weeks of leave is inadequate, we might increase the amount of time workers have to return to their jobs after taking family and medical leave. The public health consensus seems to be that about 6 months of leave after having a child is optimal for child and maternal health, so we could increase our job-protected leave from 12 to 26 weeks.
• Make more workers eligible for job-protected leave. The obvious way to do this is linking eligibility to the individual's work history instead of their employment at a specific firm. Instead of requiring 1,250 hours and 12 months of employment at a specific firm, eligibility could be based on total hours worked at all employers over the preceding 12 months. Likewise, the number of hours and required length of employment could be reduced.
• Require more firms to provide job-protected leave. Like many social phenomena, employer size has the curious characteristic that while most workers are employed by large employers, most employers are small employers. Reducing or eliminating the number of employees before a firm is required to provide job-protection is an obvious way of expanding access to job-protected leave.

The reason it's worthwhile to isolate job protection from other features of a family and medical leave policy is that job protection is valuable whether a worker's leave is paid or unpaid. That is to say, there is a difference between going 12 weeks without pay after giving birth knowing you'll have a job to return to, and going 12 weeks without pay after giving birth knowing that you'll be unemployed at the end of the 12 weeks and need to seek out a new employer.

Wage insurance

The second piece of a family and medical leave policy we can isolate is wage insurance, also sometimes called wage replacement, during a period of leave. The argument for wage insurance is that whether or not a worker is entitled to job-protected leave, they may not be able to afford to go weeks or months without a paycheck, and so return to work earlier than would be ideal for their own or their child or dependent's health.The United States has no national system of wage insurance, and consequently a substantial number of new parents return to work without using their full 12 weeks of FMLA leave, even when eligible, because they can't afford to go without a paycheck any longer. Fortunately, as demonstrated in the states operating their own paid family and medical leave systems, wage insurance is extremely cheap to provide. From a recent National Partnership for Women and Families fact sheet, the total employer and employee cost, wage replacement rate, and maximum benefit of each state's program is:

• California: 1% of employee's first $118,371 in annual wages, replaces up to 70% of average weekly wages, up to $1,252.
• New Jersey: up to 1% of employee's first $34,400 in annual wages, replaces up to 85% of average weekly wages.
• Rhode Island: 1.1% of employee's first $71,000 in wages, replaces 60% of average quarterly wages up to $852 per week.
• New York: up to roughly $139.17 per year, replaces 55% of average weekly wage, up to 55% of state average weekly wage (rising to 67% in 2021).
• District of Columbia: 0.62% of wages, replaces 90% of average weekly wage up to $1,000.
• Washington: 0.4% of employee's first $132,900 in wages, replaces 90% of average weekly wage, up to $1,000.
• Massachusetts: 0.63% of wages, replaces up to 80% of weekly wages, up to $850.
• Connecticut: up to 0.5% of wages, replaces up to 95% of average weekly wages, up to 60 times the Connecticut minimum wage.
• Oregon: up to 1% of employee's first $132,900 in wages, replaces up to 100% of average weekly wages, up to 120% of the state's average weekly wages.

Note here again that job protection and wage insurance are conceptually totally unrelated. You can provide wage insurance without providing job protection, and you can provide job protection without wage insurance. Indeed, that's precisely the situation in the 42 states that don't provide paid family and medical leave.Once you've conceptually isolated them, you can suddenly imagine all sorts of combinations: you could leave job protection at 12 weeks and expand wage insurance to 26 weeks. You could expand job protection to 26 weeks and wage insurance to 40 weeks for folks willing to forego job protection.

Universal benefits

There's a final set of benefits that are sometimes pulled into discussions of paid family and medical leave but that are rightly considered separately, and that is universal or near-universal benefits. Universal benefits in this context are those you're eligible for regardless of work or earnings history. The great advantage of universal benefits is they allow us to put the material welfare of people above incoherent attempts at social engineering.Job protection protects only those with eligible jobs and work histories. Wage insurance supplements the income of workers whether or not they are eligible for job protection. Only universal benefits are aimed at ensuring the material well-being of people regardless of their employment status or wage record.The closest thing we have in the United States to a universal benefit is the refundable Child Tax Credit, which increased to $1,400 per year in 2018, or roughly $116 per month per child.The problem with the Child Tax Credit, of course, is that it's only claimed once a year; a parent who gives birth or adopts a child in January won't see any benefit until as late as April the following year, and has to wait a full additional year to receive their next cash infusion. This is an absurd system and, oddly, another better system is already in place: Social Security Child’s Insurance Benefits. These cash payments are available to the minor children of disabled and retired workers, and are received monthly, either by paper check or direct deposit.Besides providing benefits to parents, patients, and caregivers with insufficient work and earnings histories, universal benefits are also capable of reducing the stakes involved in job protection and wage insurance. As indicated above, the maximum weekly wage insurance benefit in most states is around $1,000. Mechanically, a $1,000 monthly universal basic income would allow that maximum wage insurance benefit to fall to just $750, leaving the maximum income of parents with work histories unchanged (at $4,000 per month) but increasing the income of non-working parents from $0 to $1,000.

Conclusion

This post isn't intended to convince you of any one particular policy solution. Personally, I think we need more job protection, more universal wage insurance, and more universal benefits, but I don't know for sure whether we need 26 weeks or 40 weeks of job protection, whether we need 66% or 100% wage insurance, and whether we need fully universal benefits or a mixture of universal benefits and wage insurance and job protection.Rather, this post is meant to help you ask the right questions when your state, your congressperson, or your senator makes a paid family and medical leave proposal: does it extend job protection to additional workers, or beyond 12 weeks? Does it provide wage insurance, and if so, for whom and at what rate? Is it universal or is it means-tested?If you ask the right questions, you at least have a chance at making the policy better. And if you don't know what questions to ask, we'll be stuck with the stingiest welfare state and unhealthiest population in the developed world.Become a Patron!

What will the Social Security funding "crisis" look like?

Become a Patron!When ignorant people want to sound serious, they start solemnly intoning about the Social Security funding crisis and the need for a "long-term" "fix" to the "problem." This serves three useful purposes: it allows the speaker to change the subject from any actual problems existing today, it provides the superficial cover of concern for vulnerable populations, while also giving a sheen of non-partisanship to plans that will decimate the working class.This makes it extremely important to understand what the Social Security funding "crisis" will look like in the real world.

Nothing at all will happen until 2034

The most important chart to understand in the Social Security Trustees' report is Figure II.D2, which I'll reproduce in its entirety:What this chart says is that until 2034, under current law, with no additional benefit cuts or additional taxes or funding streams, the Social Security Administration is projected to be able to pay 100% of scheduled benefits, including old age, disability, and survivor benefits. So whether or not any changes are made to Social Security, nothing will change for any new or existing beneficiaries for the next 15 years.If you are not a current beneficiary, or planning to begin receiving benefits soon, you may find this pedantic, but if you're already receiving benefits you should know: under current law your benefits will not change in any way for the next 15 years.

What happens after 2035?

After 2035, the Social Security trust funds are projected to begin exhausting the Treasury bonds they accumulated during the years the funds took in more money than they paid out in benefits. Unlike a private company that does not receive enough income to repay its debts, the Social Security Administration cannot, and does not have any need to, "declare bankruptcy" or "go bankrupt."Instead, as Figure II.D2 shows, benefits will be reduced to the level payable through annual payroll contributions. In 2035 the actuaries predict there will be a 20% reduction in payable benefits, which will increase excruciatingly slowly to a 25% reduction by 2093.This is an important moment to remind you that the exhaustion of the trust fund cannot "reduce your promised benefits," because these are the benefits promised by the Social Security Act, as amended. If you're happy with 20-25% lower benefits in old age, disability, or widowhood after 2035, you don't need to do anything and you don't have a dog in this fight.

Should we allow this benefit cut to take place in 2035? Probably not!

If, like me, you think the American welfare state is not generous enough then, like me, you probably don't think we should suddenly cut Social Security benefits in 2035.But if you think the American welfare state is not generous enough, then you have the luxury of other, lower-hanging fruit. The Trump administration is planning to reimpose a cumbersome asset test on SNAP beneficiaries nationwide, denying 3.1 million people access to nutritional assistance. Another 500,000 elementary and secondary students are expected to lose access to free school lunches.Meanwhile, Republican officials are using a laughable legal argument to pursue the invalidation of the Affordable Care Act, including its protections for pre-existing conditions, subsidies for low-income workers who don't receive health insurance through their employers, and Medicaid expansion. And thanks to the Republican effort to amateurize the judiciary, there's no reason to believe they won't win in front of the Gorsuch-Kavanaugh Supreme Court.If you're elevating a fantasy problem in 2035 over the real problems we're facing right now, your priorities say a lot about you and nothing about Social Security's funding mechanism.

My boring solution to the imaginary Social Security funding crisis: magic it away so we can address real problems

The total funding shortfall between now and 2093, in the Social Security Trustees' actuarial report, is a bit under $14 trillion, or $186 billion per year for the next 75 years.So here's my boring solution: let's round up, and deposit $14 trillion in special issue Social Security bonds into the relevant trust funds, out of nowhere. This will, of course, increase the debt service costs of the US government as it pays interest on those bonds, and that increased cost will need to be financed through reduced spending, increased taxes, or seignorage.But this is, of course, the exact issue the Social Security funding "crisis" is supposed to present: should we reduce spending, increase taxes, or print money? It may be that we should reduce Social Security benefits, especially for higher earners. It may be that we should increase them, especially for low-income workers and those with a limited official work history.But since there's obviously no reason Social Security benefits should be tied to the year-to-year revenue produced by FICA taxes, we can easily make sure they aren't.

Conclusion

In reality, no one in politics wants to solve the imagined Social Security funding "crisis." When the Greenspan amendments to the Social Security Act were made under Reagan, they were explicitly designed to create another funding crisis further down the line, so that benefits could be cut further.Whether we want benefits to be cut or not is up to us. We can eliminate the cap on Social Security taxes and treat capital gains and dividends as ordinary income and eliminate the "funding gap" tomorrow. Some of us support those policies, others oppose them, but no one should support or oppose those policies because of their downstream effect on Social Security benefits: whether or not Social Security benefits are suddenly cut 15 years from now doesn't have anything to do with the program's funding mechanism.Rather, it's a question about who we are and what kind of country we want to live in. That's the question Roosevelt had to answer in 1935 when Social Security was created, and it's the question we have to answer today. Nothing an actuary says is going to get you out of having to answer it for yourself.Become a Patron!

The SECURE Act and the defects of the centrist mind

Become a Patron!I've written previously about the SECURE Act, the House version of a measure designed to encourage employers to allow employees to gamble their retirement savings on the long-term financial stability of private insurance companies. The measure has since cleared the House but is currently being held up in the Senate for now by Ted Cruz who is trying to turn the 529 loophole into a piggy bank for the wealthiest Americans. With this terrible measure one step closer to passage, I want to use it to illustrate a particular problem raised by the cult of moderate centrism.

401(k) plans have enough bad investment options without adding annuities

Like many pieces of the American welfare state, 401(k) plans became a retirement savings tools for private sector workers more or less by accident. Consequently, they are governed by a mishmash of rules and regulations of different vintages. For example, while the investment options within a 401(k) plan have to be administered exclusively for the benefit of the contributing employee, the design of the plan and selection of investment options does not — employers can and do receive kickbacks from 401(k) custodians for filling their plans with high-cost investment options, for instance. Employee lawsuits typically revolve around this distinction: is the kickback an employer receives based on employees' investment choices or based on plan design?Every 401(k) plan I've seen has included 10-15 high-cost, actively managed mutual funds alongside 2-4 low-cost index funds. That's not the exception: that's the rule. When deciding on an asset allocation within a 401(k), there's simply no alternative but to look up the ticker for each available investment option, record its benchmark and expense ratio, and try to cobble together an appropriate low-cost asset allocation from the options available. This is not terribly difficult for someone who knows what they're doing and has the time and patience to do it, but it's naturally overwhelming for the vast majority of people who have neither aptitude not interest in making these kinds of investment decisions.

Insurance companies want to add their annuities to your 401(k)

That brings us to the core policy goal of the SECURE Act: give employers a "safe harbor" from employee lawsuits when they include annuities offered by insurance companies that meet a rudimentary test of financial stability.You might observe this is a very counter-intuitive way to frame the policy change, and you would be right. The SECURE Act is supposed to improve employee retirement security by preventing employees from suing when their employers offer them crappy retirement investment options? Surely it would make more sense to improve employee retirement security by increasing their legal recourse against employers who do not offer them appropriate and appropriately-priced retirement investment options!But of course the circle is easy to square when you remember that annuities are sold, not bought. The issue is not that employers, let alone employees, are demanding access to annuities in their 401(k) plans. The issue, rather, is that insurance companies are clamoring to have their annuities included in 401(k) plans but are being stymied by the unwillingness of employers to take on the legal risk of vetting them. With that barrier removed through the "safe harbor," insurance companies will be free to offer the same kickbacks investment companies do today to sell their confusing, expensive annuity contracts to 401(k) participants.

Annuities don't eliminate risk, they transform and hide it

In theory I don't have anything against single-premium immediate annuities as a method of converting a lump sum into a predictable stream of income upon retirement (or any other time), but it's important to understand what is and is not happening when you do so.When you hold an FDIC-insured bank deposit, or a SIPC-insured security in a brokerage account, you are roughly speaking entitled to some stream of income (interest, dividends, capital gains distributions), plus the value of the underlying asset. Both components of the asset will generally fluctuate: the interest you earn on your bonds will change as you reinvest coupons over time; dividends will rise and fall along with the profitability and capital allocation decisions of the underlying companies; the asset's resale value will change along with the winds of capitalism. Even cash grows more or less valuable as inflation rises and falls.When you exchange those real assets for an annuity contract, you receive a different kind of asset in exchange: the promise of an insurance company to pay you a fixed or variable sum described in the contract over some time period. You've now converted the variable stream of income and variable asset value of your stocks and bonds into a fixed income stream.But there's no sense in which you have eliminated your risk by doing so. Instead, you've simply transformed the risk you're taking. Instead of being subject to the whims of inflation, interest rates, and the stock market, your income now depends on your insurance company being able to pay the promised stream of income over the promised time horizon.To be clear: I'm not an insurance company analyst and I don't have the tools to perform a comprehensive assessment of the creditworthiness of American insurance companies over the next 70+ years (a 40-year career followed by a 30-year retirement, for instance). Of course, your employer probably isn't either, and that's why they don't offer annuities in your 401(k) plan. If passed, what the SECURE Act will do is relieve employers of that responsibility, so you won't have any recourse if the annuity you select flounders and your contributions are lost or deeply discounted in your insurer's bankruptcy.

Social Security has always been the answer

Social Security's old age benefit is the only source of income security for the overwhelming majority of older Americans. It's important to understand exactly what this means. Social Security is not the only source of income for older Americans. About 27% of Americans continue to participate in the labor force (i.e. work or look for work) between the ages of 65 and 74. Others receive passive income from rental real estate, farm, mineral, and gas leases, etc.What distinguishes those sources of income from Social Security is that Social Security old age benefits are paid by the federal government and subject to annual cost of living adjustments, and they're guaranteed to continue for as long as you live. Employment income in old age lasts as long as you're employed, rental income fluctuates over time (just ask Detroit if you don't believe me), and commodity prices go through long cycles of rise and decline. Social Security isn't like that.That means the first place you should look to improve income security is the only source of income security most people have, and the obvious place to start is allowing people to make additional, voluntary Social Security contributions. Since the only input into the Social Security benefit calculation is the average wage-inflation adjusted income reported in each year of a worker's earnings record, a natural approach is to treat voluntary contributions as "increased income" for the year the contribution is made.For example, a worker earning $50,000 in 2019 would ordinarily pay $3,100 in OASDI (the Social Security component of FICA), matched by their employer. An additional, voluntarily payment of $6,200 (conveniently close to the 2019 IRA contribution limit) could raise their recorded OASDI income for that year to $100,000. Note that this would not double their Social Security old age benefit, since each year of earnings only contributes 1/35 to a worker's average earnings, and old age benefits increase at a graduated rate.

Centrists say personal responsibility when they mean risk

I'm all for "personal responsibility," defined properly. I simply don't know how a person can take personal responsibility for the trajectory of interest rates, or the performance of the S&P 500, or the rise and fall of US auto manufacturing, or the financial stability of America's insurance companies. If you believe an important problem facing America is the problem of retirement security, by all means let us allow workers to reduce their present consumption in exchange for higher income in retirement.But having decided to do so, why on earth would we then subject them to the cost, complexity, and vulnerability of private insurance companies?Because when a centrist talks about personal responsibility, what they really mean is risk. Personal responsibility for the decision of whether to go to college, and what to study, means the risk of poverty. Personal responsibility for an unplanned pregnancy means the risk of homelessness. Personal responsibility for filing your SNAP application on time means the risk of hunger. Personal responsibility for your income in retirement means the risk of being taken advantage of by unscrupulous employers and insurers.But there's no way to take personal responsibility for what happens to us under a system that mechanically produces pain and trauma. Our personal responsibility is to fix the system.Become a Patron!

Roth 401(k)'s are different, but they're not special

Become a Patron!The other day I got into a very frustrating argument with a popular anonymous Finance Twitter account about the epistemology of 401(k) savings vehicles. The frustration arose from a very specific phrase he used to describe one of the benefits of a Roth 401(k) over a traditional 401(k): "you’re able to contribute more with a Roth 401k." This is such a strange claim I asked him to clarify his argument, but he was never really able to do so.That being the case, I want to see if I can try to make the strong form of the argument that he wasn't able to make.

It's legal to save money in taxable accounts

While most financial planners will readily tell you that most people, under most circumstances, should maximize their workplace retirement savings contributions before investing in a taxable brokerage account, it's important to remember that it's perfectly legal to invest in taxable accounts at any time for any reason. Not only that, there are some advantages to doing so:

• Funds can be contributed in unlimited amounts, at any time, for any reason. For example, the S&P 500 hit its financial crisis low in March, 2009. If you were somehow aware of this fact at the time, making contributions through a workplace 401(k) would leave your hands somewhat tied: first of all, two months of payroll contributions had already been made for 2009, leaving you just 83% of your maximum $16,500 contribution left. Deferrals are limited to the amount of your paycheck, and sometimes to just a fraction of your paycheck, so it might take months to max out your contributions for the year, dramatically reducing the value of your clairvoyance.
• Dividends and capital gains are taxed at preferential rates, as low as 0%. Moreover, over time reinvesting dividends and capital gains allows you to establish a range of different cost bases for different lots, allowing you to right-size your capital gains depending on your tax rate year-to-year.
• Capital losses can be used to offset gains and ordinary income. A diversified taxable portfolio will experience gains and losses in different positions over time, giving taxable investors additional tools to manage their tax liability.
• Assets can be withdrawn at any time for any reason. There are no restrictions, no holding requirements, no caps, and no repayment requirements when withdrawing assets from a taxable account. It's your money from the day you deposit it to the day you withdraw it.

Of course, there are disadvantages as well, first and foremost among them that taxable investments are made with after-tax money. Thus, the price you pay for all the advantages of taxable investments is the income taxes owed on the invested amount in the year it's earned. To be consistent throughout, let's use the 22% marginal tax rate on single filers making between $51,475 and $96,200 in earned income per year. In order to make $19,000 in taxable investments, this investor must first pay $5,359 in federal income tax on $24,359 in earned income.Splitting it up in this way allows us to think about the $19,000 deposit as a "bundle" of two different assets: $19,000 in investible funds, plus the combination of rights and privileges the investor paid $5,359 for. This doesn't mean preferential taxation and ease of access are worth $5,359, rather, it means they cost $5,359.Knowing these facts allows us to perform all sorts of fancy calculations. For example, at a 0% discount rate (you value a dollar today exactly the same as a dollar in the future), it would take a little over 9 years for the benefits of maximal tax-loss harvesting ($3,000 in losses per year at a 22% income tax rate saves the taxpayer $660 per year) to recoup the amount paid for the right to tax-loss harvest. At a discount rate of 6%, it takes over 11 years to break even.It also allows us to adjust the value of our asset in the face of changing circumstances. Raising the maximum amount of deductible losses from $3,000 to $3,500 raises the value of the asset by reducing the amount of time needed to break even. Lowering or eliminating deductible losses reduces the value of the asset by extending the breakeven period.This sounds straightforward, because it is, but it's also a useful tool to use when thinking about the difference between traditional and Roth 401(k) contributions.

Traditional 401(k)'s combine a discounted investment asset and a tax liability

Traditional 401(k) contributions are made from payroll and removed from your taxable earned income before it's reported to the IRS. The same $24,359 in earned income as above can be split into a $19,000 traditional 401(k) deferral and, after paying federal income taxes at 22% on the remainder, a $4,180 taxable investment.It's essential to understand that your current-year cash flow situation is identical whether you make the traditional 401(k) deferral or whether you direct the income entirely into a taxable account: in both cases you earned $24,359 in income, in both cases you spent $0 on current-year consumption. The fact that in the first case you paid $1,179 in taxes while in the second you paid $5,359 in federal income taxes is entirely irrelevant to your current-year cash flow, which in each case is $24,359-in and $24,359-out.The difference is that in the second case you have created, alongside $23,180 in total investible assets, a new bundle of liabilities. Just as a taxable account represents "investments plus rights," a traditional 401(k) represents "investments plus duties."But this is a very peculiar liability. Most importantly, it is:

• the duty to pay taxes at your ordinary income tax rate, plus a 10% penalty, on withdrawals made before age 59 1/2.
• the duty to pay taxes at your ordinary income tax rate on withdrawals made between ages 59 1/2 and 70 1/2.
• the duty to make withdrawals (and pay ordinary income tax) starting at age 70 1/2.

Moreover, we know precisely what the federal government is willing to pay our investor for this liability: $4,180, the taxes foregone on his income when he made the traditional 401(k) deferral. The terms of this deal have some obvious advantages to the investor:

• playing with the house's money. Traditional 401(k) deferrals offer a kind of heads-I-win-tails-you-lose dynamic, since only 78 cents of each dollar in the account was contributed by the investor. This might allow the investor to take on more risk: if they hit a home run, there will be plenty left over to pay back their liability with, while if the account goes to zero, they get to walk away from their liability with no consequences.
• control over timing. While the $4,180 liability was created at a 22% marginal tax rate, it's paid back at the investor's marginal tax rate at the time of retirement. Planning on an early retirement and taking advantage of the Social Security magic trick might allow the investor to repay their liability at a 12% tax rate or below.
• pre-retirement rollovers and conversions. During a break in service it may be possible to move assets from traditional 401(k)'s into Roth IRA's, exchanging a liability for an asset at a deep discount, depending on your marginal tax rate.

Just as we can adjust the value of our taxable rights up and down as the tax code changes, we can adjust the value of our traditional 401(k) liability up and down. With the end of the "marriage penalty" and cuts in personal tax rates at least through 2028, tax brackets are historically wide and marginal tax rates are historically low. If those brackets narrow, or are tied to a slower measure of inflation, then the value of the traditional 401(k) liability might even grow faster than the value of the assets in the account. If you work longer or earn more money than expected, the value of the liability will likewise grow.Meanwhile, if you retire earlier than expected, your income is lower than expected or tax-exempt for whatever reason, tax rates are cut further, inflation adjustments are made even larger, or the required minimum distribution age is increased to 72 or higher, then the value of the liability on your books will fall.

Roth 401(k)'s are just another bundle of investments and rights

All of this brings me back to Jake's bizarre claim that Roth 401(k)'s allow you to "save more money." As should be clear at this point, nothing could be further from the truth. With $24,359 in earned income, you can:

1. invest $19,000 in a taxable account;
2. invest $19,000 in a traditional 401(k) account and $4,180 in a taxable account;
3. or invest $19,000 in a Roth 401(k) account.

In each case you have "saved" the same $24,359, in the concrete material sense that you earned the money but did not spend it. In the first case you pay $5,359 in federal income taxes for the right to preferential dividend and capital gains taxation and the ability to offset ordinary income with capital losses. In the second case you receive $4,180 in investible assets in exchange for a liability to pay your marginal ordinary income tax rate, plus penalties under certain circumstances, on any amount withdrawn from the account. And in the third case, you receive $19,000 in investible assets that can be withdrawn tax- and penalty-free in retirement.All of these assets and liabilities can be analyzed, assigned a value, and recorded on your personal ledger based on a sober analysis of present and future conditions. Moreover, the value can be adjusted over time as the future becomes clearer. It can be worth incurring liabilities to invest more money (traditional IRA's) if the investment will yield a higher return than the liability, and it can be worth paying for an asset up front (taxable and Roth investments) if you believe the value of the asset will appreciate at a faster rate than the consideration paid.And, of course, you can diversify your investments, assets, and liabilities across all three, and change that allocation over time, as public policy and your life circumstances change.That's the argument I hoped Jake was making in his initial tweet, but since he wasn't able to make it, I figured I would instead.Become a Patron!

Bill Cassidy has the worst paid parental leave plan yet

Become a Patron!This morning I hauled myself out of bed early to go down to the American Enterprise Institute and listen to Bill Cassidy discuss the outlines of his plan for paid parental leave. The dervishes at AEI already uploaded the video, so you can take a look for yourself, if you can stand to listen to politicians speak for more than 30 seconds at a time. Bonus: if you hear someone guffawing off-screen a few minutes in, that's me laughing at the idea anyone is going to fall for his schtick.

The problem with existing Republican paid leave plans

I've written at length about this before, so won't belabor the point here, but the existing Republican proposals for a national paid leave program are based on forcing new parents to choose between the financial support they need to take time off work and the financial support they need to retire with dignity. It's a cruel choice to force anyone to make: do you turn down paid parental leave and return to work a few days or weeks after giving birth or adopting a child, or do you accept paid parental leave and suffer from a permanently lower income in retirement?

Cassidy's plan is designed to become as unpopular as possible as quickly as possible

If you watch past Cassidy's Hee Haw routine with Aparna Mathur (who has also graced these digital pages before), you can enjoy the much more interesting follow-up discussion between several panelists from several points on the political spectrum. The panelists over-generously described Cassidy's plan as a work in progress, the specifics of which they didn't dare speculate about. This is incorrect: a carefully trained ear allows you to clearly identify the core tenants of Cassidy's plan. Since I have such an ear, I can tell you: folks, it's not great.The first question any paid leave plan has to answer is, who is the beneficiary of the program? This might sound like a weird question, but the answer matters a great deal. In traditional paid leave programs, the beneficiaries are thought to be the workers. This allows you to apply the same logic across the board: workers are entitled to a certain number of days or weeks of paid (or in the United States, unpaid) leave to deal with a serious illness, to care for a relative, or to bond with a new child. Cassidy's plan answers the question differently: the baby is the beneficiary of the program. Why does this matter? Because it means his plan provides paid parental leave exclusively to one parent. Cassidy is a homophobe, but not an idiot, so he allows that "under some circumstances" a child's father could take the leave instead, but it's clear despite his hemming and hawing that his plan will only cover one parent. This is not parental leave, it's maternity leave, and there's a reason virtually every other social democracy has abandoned it in favor of parental leave: it reinforces gender stereotypes in the home, it penalizes the employers of women and subsidizes the employers of men, and at the most basic level it keeps fathers apart from their children during a critical early bonding period.The second question is, who is eligible for the program? Is the program universal, or is the program exclusively for those with earned income? How recent must the earned income be? Is the triggering event the birth of a child, or is the triggering event leaving work to care for a newborn child? Are you still eligible if you continue to work while caring for your child? This question has been answered in a variety of ways in a variety of contexts, and I have made clear to the relevant organs that it's one of the objections I have to the most important Democratic proposal for paid family leave: I would prefer a universal allowance detached from work history, rather than a benefit exclusively for those with work histories and tied to their earnings record. But Cassidy's plan goes the other direction: only mothers making less than $70,000 a year will be eligible for his benefit. My readers are no dummies, so they're already asking: how will income be measured? Is it based on an individual earning record, a joint earning record, or on tax returns? Can you game the system by having children at particular times of year before taxes are filed? Can you game the system by getting married — or by getting divorced? The issue is not that these are unanswerable questions. The issue is that people should not have to ask them!Having spent half an hour in a room with him, I can confirm that Bill Cassidy is a dumbass. But however little firepower he's working with under the hood, I absolutely do not believe any of this is accidental. This plan is designed, top to bottom, to generate as much animosity and antagonism towards the welfare state as possible, at the lowest possible cost. If a paid maternity leave plan like this were passed, it would be used as a cudgel by employers against workers ("if I give you a raise you won't be eligible for maternity leave"), workers against workers ("can you believe our janitors get paid maternity leave and we don't?"), and fathers against mothers ("really you can't afford to keep working since you're entitled to maternity leave").

Conclusion: this plan won't pass, but that doesn't let Republicans off the hook

There is one realistic plan for comprehensive paid family and medical leave on offer: the FAMILY Act. That's not to say it's perfect. I think the benefit should be fixed, universal, and unlinked from earnings history. But I've lost that fight, and the fact is, if you want a near-universal benefit that includes paid medical leave, caregiving leave, and parental leave, financed with a small payroll tax, you have to support the FAMILY Act, because it's the only game in town.None of the Republican efforts to destroy retirement security and worker solidarity will pass, but they weren't written to pass. They were written to serve as vehicles for Republicans to say that they, too, care about providing the paid time off American workers desperately need. But it's just not true. If they did, there would be 100 Senate votes for the FAMILY Act. Instead, there are 47 votes in favor of paid family and medical leave, and 53 votes against it.You know what to do.Become a Patron!

Beware Republicans bearing robust debate and compromise

Aparna Mathur is a Resident Scholar, Economic Policy, at the American Enterprise Institute, a libertarian think tank with an excellent kitchen that I periodically visit when I want to save money on lunch or cocktails. If you're ever in DC I strongly recommend visiting their events page to see if you can enjoy one of their catered meals or open bars.Scholar Mathur has recently focused her residency on the question of paid parental leave in the United States, as in a recent blog post, "The Birth of a Compromise on Paid Parental Leave" (get it, "birth?").As she and her co-authors write, "history shows that when both sides express a willingness to compromise, great policies can emerge. Our elected officials are now facing one such historic opportunity. It is time for them to pass legislation that creates a national paid parental leave program."The question, as Napoleon famously asked about the pope, is "how many divisions does the AEI-Brookings Paid Leave Working group have?" More to the point, how many votes do they have?

Ted Cruz and the amazing vanishing immigration vote

The year was 2013, and after years of gridlock, Democratic and Republican moderates in the Senate had hammered out an immigration compromise that included border enforcement, changes to future immigrant flows, and legal status for long-time US residents. The bill quickly ran into a problem: the Texas Senator, and Canada native, Rafael Edward "Ted" Cruz.Cruz, the son of Cuban revolutionary Rafael Bienvenido Cruz y Díaz, was concerned that the bill as drafted would allow long-time undocumented residents of the United States to receive permanent resident status and, eventually, US citizenship like his and his father's (n.b.: his mother was born in Delaware, so while he immigrated to the US from Canada at a young age he enjoyed US citizenship from birth).What followed was a grueling rearguard action as Cruz moved heaven and earth to try to block newly normalized permanent residents from any so-called "path to citizenship." Unable to sway his colleagues, he was forced to vote, more in sadness than in anger, against final passage of the "Border Security, Economic Opportunity, and Immigration Modernization Act."At least until the 2016 primaries came around, when Ted 2.0 was launched, and much to the surprise of his colleagues and constituents, it turned out he had been opposed to the law all along, whether or not it included a path to citizenship. Incredulous reporters went back and checked the tapes, and found out Cruz had masterfully outplayed them in 2013: it turned out Cruz had never made his vote conditional on removing a path to citizenship from the bill. Indeed, it soon became clear his vote was never in play.To take one of literally hundreds of examples, on May 21 Cruz introduced an amendment to remove the path to citizenship, saying, "I don’t want immigration reform to fail. I want immigration reform to pass. And so I would urge people of good faith on both sides of the aisle, if the objective is to pass common-sense immigration reform that secures the borders, that improves legal immigration, and that allows those who are here illegally to come in out of the shadows, then we should look for areas of bipartisan agreement and compromise to come together. And this amendment—I believe if this amendment were to pass, the chances of this bill passing into law would increase dramatically" (emphasis mine).Now, you or I might look at this statement and conclude that Ted Cruz, someone who says they want immigration reform to pass, but who can't vote for a bill that includes a path to citizenship, and is proposing an amendment to remove a path to citizenship, is saying that he will vote for the bill if the amendment passes.But Ted Cruz never said that. In 2016, the trap was sprung: Cruz revealed that he had been opposed to immigration reform all along, and that his amendments were intended solely to reduce the bill's chance of passing, and to weaken it in case it did.

Olympia Snowe and the amazing vanishing healthcare vote

If you'll step into my time machine yet again, let's return to 2009, when Democrats held a filibuster-proof majority in the United States Senate. Max Baucus, Democrat of Montana and chair of the Senate Finance Committee, had been holding months of hearings and closed-door sessions in an effort to get three Finance Republicans on board with healthcare reform: Mike Enzi, Chuck Grassley, and Olympia Snowe.After months of wrangling, the death of Senator Ted Kennedy, the seating of Al Franken, and Arlen Specter's frantic last-minute party switch, all three Republicans voted against cloture and against final passage of the Affordable Care Act.

Paul Ryan and the amazing vanishing Earned Income Credit expansion

Before he decided to retire to spend more time with the Koch brothers' money, Paul Ryan reinvented himself in office one last time: as an anti-poverty champion. The so-called "Better Way" Republican agenda, launched in the summer of 2016, included an anti-poverty program (available for now at the Internet Archive). That document contains the following enigmatic paragraph:

"The Earned Income Tax Credit is another potential solution. The EITC is a refundable credit available to low-income workers with dependent children as well as certain low-income workers without children. It can help with the transition because it increases the financial rewards of work. Increasing the EITC would help smooth the glide path from welfare to work."

Fortunately, on other occasions Paul Ryan has written more extensively about his support for Earned Income Credit expansion. In a 2014 "discussion draft" for the House Budget Committee, he wrote:

"there’s a growing consensus to expand the EITC for childless workers...Given the EITC’s success in boosting work among families with children, a larger EITC should have a similar effect on childless workers. Given these troubling trends for young workers, there is a real need to consider lowering the age of eligibility for the EITC, which currently does not serve this population...Because the EITC helps low-income households while encouraging work, this proposal would expand the credit for childless workers. Specifically, it would double the maximum credit, phase-in, and phase-out rates for childless adults, and it would lower the eligibility age for workers from 25 to 21, assuming they are not a dependent or qualifying child for another taxpayer."

If you knew nothing else about Paul Ryan, you might conclude from this evidence that he supports an expansion of the Earned Income Credit for childless workers.But since you are reading this today, you know that would be wrong. When offered the opportunity to pass changes to the tax code that could increase the deficit by a total of up to $1.5 trillion over the 10-year budget window, i.e., the changes did not have to be paid for as long as they were not scored as increasing the deficit by more than that amount in that period, Paul Ryan did not lower the eligibility age for the Earned Income Credit. He did not double the maximum credit, nor the phase-in, nor the phase-out rates for childless adults. He did not expand the Earned Income Credit at all.That is because Paul Ryan does not, and never did, in fact support an expansion of the Earned Income Credit. Paul Ryan was lying.In a roundabout way, this brings us all the way back to Aparna Mathur, whom I asked on Twitter earlier this year, "why do you think Paul Ryan refused to expand the EIC after making it the cornerstone of his woke Republican anti-poverty agenda?." Aparna Mathur didn't have me muted back then, so she saw my question and even replied to it:"I wish we could move forward with an EITC expansion. I have no idea why a policy that has so much support doesn't make it through Congress..it would be so helpful."She. Has. No. Idea. Which tells you almost everything you need to know about Aparna Mathur.

The Republican Party and the amazing vanishing national paid family leave program

Aparna Mathur thinks that the time has come for a national paid parental leave program. I also think the time has come for a national paid parental leave program. The proposal with the most widespread support in Congress today is the FAMILY Act (don't ask what it stands for), which would create a small additional payroll tax and use those funds to pay for wage replacement for workers who need to take time off to care for a new child or in case of serious illness or injury. It currently has 35 Democratic co-sponsors in the Senate and 178 Democratic co-sponsors in the House of Representatives, and widespread support among Democratic members of Congress.Unfortunately, Aparna Mathur doesn't much care for the FAMILY Act. Without meticulous documentation, I can boil down her objections as follows:

• It's too long. While the FAMILY Act entitles workers to 12 weeks of paid leave, that's a little bit excessive. Isn't 8 weeks of leave a bit more realistic?
• It's too generous. While the FAMILY Act entitles workers to 66% of their weekly pay up to $1,000 per week, that seems like an awful lot of money to give to new parents. Why not limit it to $600?
• It covers too many life events. While new parents surely need some time to bond with their children, there's no need to pile family and medical leave into the same law. Why not restrict the paid leave benefit to birth and adoption events only?

I think these objections are ridiculous, but this post isn't about what I do or don't consider ridiculous. That's a judgment you have to make for yourself.This post is about the fact that Aparna Mathur, AEI, and the Brookings Institution don't have the votes. If Aparna Mathur could come up with 13 Republican Senators willing to vote for a motion to bypass Senate Majority Leader Mitch McConnell and bring this pared-down version of paid parental leave to the floor of the Senate, it would get 47 Democratic votes, and it would pass the House with a comfortable Democratic majority.But there aren't 13 Republican Senate votes for a pared-down paid family leave law. There isn't one Republican vote for a pared-down paid family leave law. There are, currently, maybe 3 votes for a Rubio-style "mortgage your retirement to spend a few weeks with your kids" bill, but there are zero Democratic votes for that idea because it's terrible.The point is, there's no secret backdoor workaround to find Republican votes for compromises, hacks, or kludges. If, like Aparna Mathur, you think "It is time for them to pass legislation that creates a national paid parental leave program," then you have to support the FAMILY Act. You have to contact your representatives in Congress and demand they support the FAMILY Act. If they refuse, you have to vote against them in your congressional primary, and vote for Democrats in general elections, until you are represented by someone who does support the FAMILY Act.That's the entire show. While Aparna Mathur and her colleagues at AEI and Brookings furiously workshop compromises, the one thing they can't do is provide the votes necessary to pass national paid family leave into law. For that, you need Democrats — and you need a lot of them. Get cracking!

There's no such thing as an "American retirement model"

There's no cliche more beloved by the financial press than the "changing American retirement model."The beauty of the changing American retirement model is that you can use it to prove any point you want:

• if you want to cut Social Security old age benefits by raising the full retirement age, the changing American retirement model is that people are living and working longer;
• if, on the contrary, you want to increase Social Security benefits, the changing American retirement model is that fewer employers offer defined benefit pension plans so the government needs to step in to supplement retirees' income;
• if you want to increase the contribution limits to tax-sheltered retirement accounts, the changing American retirement model is that retiree health care and housing expenses are increasing faster than consumer price inflation.

The problem with these arguments, some of which I'm sympathetic to and some of which I detest with every fiber of my being, is that there is no American retirement model, and there never has been. There are at least 3 different ways to illustrate this, and I think each one has value.

People who die young don't get to retire

In the 2015 United States life table, you can see that of 99,065 infants surviving to age 18, only 86,915 survive to age 62, the earliest age they'd be eligible for Social Security old age benefits, a survival rate of 87.7%. The American retirement model for the other 12.3% of the population is to die.The reason I bring this up is that arguments based on life expectancy only apply to people who survive to retirement. There are two moving pieces here. First, the number of mortalities per age rises steadily, which means for each year you raise the retirement age by, you exclude an increasingly number of people from ever being able to retire. Second, increasing longevity is concentrated in high-income, low-impact white collar professions — people who have other resources to draw on in order to meet their retirement needs in the case of a higher retirement age.No wonder higher retirement ages are so popular among the wealthy.

No one saves anything for retirement, and no one ever has

Vanguard reported in 2017 (page 45) that their average 2016 401(k) account balance was $178,963 for participants between the ages of 55 and 64, while the median balance for those participants was $66,643. I don't want to argue about whether Vanguard assets are "representative" or not, so apply as wide a margin of error to that figure as you like; maybe Vanguard participants are particularly savvy, and maybe they're particular slouches.The point is, using a primitive 4% withdrawal rule, $67,000 can produce about $2,680 in safe withdrawals per year in retirement. While the average is pulled up by what Vanguard calls "a small number of very large accounts that significantly raises the average above the median," even the average account balance can only produce $7,160 in "safe" withdrawals per year.

Defined benefit pension coverage has not changed

The third piece of the "American retirement model" puzzle is defined benefit pension plans, which have been said to be steadily shrinking for at least as long as I've been alive.In a 2018 Bureau of Labor Statistics report, 26% of civilian workers had access to defined benefit pension plans. Sounds low, right? The third leg of the American retirement model has been plucked from its fitting!The trouble is, most workers never had access to defined benefit pension plans. There are two reasons this is so confusing to people. First, defined benefit pensions are typically reported as a percentage of total retirement plans, as in this Vox.com article: "as of 1983, defined-benefit plans were the majority of retirement plans in the US. By 2004, most were defined-contribution."You can find an even more extreme claim in this report: "At one time, 88 percent of private sector workers, who had a workplace retirement plan, had a pension" (emphasis mine).But that doesn't tell you anything about the number of defined benefit pensions in existence in 1983 or at any other time — the entire effect could have been (and indeed was) created by increasing the number of defined contribution plans at employers that never offered defined benefit plans!Second, the decline in defined benefit plans is concentrated almost entirely in the private sector. According to a 2017 Congressional Research Service report, 85% of state and local government workers still have access to defined-benefit pensions today.Have I kept you in suspense long enough? Today's 26% access rate to defined benefits pension plans has fallen from a 1989 access rate of...32%. That's it. 30 years — almost my entire lifetime! — of being lectured by neoliberal think tanks about the collapse of the American retirement model, about the need to work longer, harder, re-skill, up-skill, trans-skill, and the defined benefit pension access rate has fallen, in total, by a little under 19%. Another way of thinking about it is that the total number of workers covered by defined benefit plans has risen, from 32% of the 1989 workforce to 26% of the much larger 2018 workforce.

Stop over-thinking: the suffering of the elderly is the problem and Social Security is the solution

There is no problem under our young sun that the neoliberal mainstream of our politics doesn't think can be solved with three-legged stools, with success sequences, with means-testing, and with personal responsibility.But it's not true. Elder poverty is a problem with a single cause: under our version of market capitalism, people who work get paid, and people who don't work don't get paid. If people insist on living longer than they're able to work, then they either starve or receive money from those who are still able to work. This has nothing to do with life expectancy, it has nothing to do with retirement savings, and it has nothing to do with defined benefit pension coverage.Either we transfer money from workers to retirees, or retirees die hungry and helpless. Fortunately, we already have a solution in place: monthly cash transfers to retirees through Social Security's old age benefit. Unfortunately, rather than finding a consensus to expand that system to ensure none of our elders are left in poverty, that system is under attack by people who want to reduce those benefits.We have an economy that's more than capable of providing for children, workers, retirees, and those unable to work. All we need is a politics that's up to the task.

Everyone should have access to the federal Thrift Savings Plan, but not for you reason you think

Shortly before his humiliation and expulsion from public life, Florida Senator Marco Rubio resurrected the idea of allowing some non-government employees to invest in the Thrift Savings Plan, which is the federal equivalent of private-sector 401(k) and non-profit 403(b) plans. The proposal got some sympathetic attention from lazy personal finance journalists, and some critical responses from bloggers serving the federal workforce.

The lazy reason to support TSP-for-all: lower costs

The narrow reason financial journalists latch onto in support of TSP-for-all is administrative costs, and it's true that TSP charges less in overhead and fund management fees than virtually all private retirement schemes, between 0.032% and 0.033%, depending on the fund.

Let me be clear: costs matter, and that's a lower cost than, for example, a Vanguard mutual fund held in a Vanguard 401(k).

But, as opponents of TSP-for-all are quick to point out, those costs are subsidized in part by fees paid by employing agencies. Universal enrollment in TSP might provide some economies of scale, but would also introduce a large number of non-subsidized workers into the system, dragging up average costs even if the final cost is lower than most workers pay today.

For that reason many TSP-for-all proposals involve replicating a second pseudo-TSP for non-federal-employees, so federal employees continue to enjoy their agency subsidies and lower costs, while non-federal-employees are shunted into a parallel system with higher, unsubsidized costs.

With costs as low as they already are in the private mutual fund industry, this strikes me as the least convincing argument for TSP-for-all.

A decent reason to support TSP-for-all: simplicity and universality

An argument for TSP-for-all that I'm more sympathetic to is that decisions over the availability of tax-advantaged retirement savings shouldn't be made by employers at all. If the federal government decides it's willing to let people shield $25,000 in earnings from income tax in 2019 ($6,000 in IRA contributions and $19,000 in workplace contributions, plus catchup contributions in both), why should your boss get to decide whether you are able to take advantage of that benefit or not? It's as nuts as companies running their own private health insurance offices.

Providing everyone, with or without their employer's permission, with access to the total retirement savings benefit they're entitled to, is a commonsense measure, although as I've said in the past, my preference would be to simply eliminate workplace retirement savings plans entirely and increase and universalize the contribution limits for individual retirement accounts (which isn't mutually exclusive with TSP-for-all).

TSP-for-all is also sometimes praised on the grounds of simplicity, and here the argument is fair enough as far as it goes. It's somewhat shocking even to me, but the "best retirement plan in the country" has just 5 investment options (plus "Lifecycle" funds that blend the 5 in different proportions): US large-cap, US small-cap, US bonds, international developed stocks, and the TSP's proprietary "G fund."

The first four are fairly standard low-cost market-capitalization-weighted index funds, while the G fund is unique in that it offers interest income based on intermediate-term US treasury rates, but with the principle guaranteed so that the fund can't lose money when interest rates rise. A neat trick!

I think between them these are excellent components in a retirement savings portfolio, and wouldn't have any serious objection if they were the only options available in a TSP-for-all plan. But there are perfectly reasonable objections to this fund lineup as well: emerging markets are completely absent, as are international bonds. Do those asset classes have a role in every portfolio? Of course not. Should they be categorically excluded from investments? That depends on the trade-off you want to make between simplicity, versatility, and diversification.

The best reason to support TSP-for-all: a simple national supplemental annuity

The primary source of retirement income for Americans is not traditional defined-benefit pensions, and it's not defined-contribution retirement plans like 401(k)'s and 403(b)'s. It's Social Security's old-age benefit, a federal program that guarantees retirement income based on your recorded earnings paid into the program during your working lifetime.

One problem with our current jerry-rigged retirement system is that under our laws, annuities are not investments at all: they're insurance contracts, sold on a commissioned basis by state-licensed agents and state-regulated for-profit companies.

If, instead, we want voluntary savings to accumulate and supplement workers' Social Security income in retirement, then there's no better solution than a federally-guaranteed annuity, available nationwide, based on a worker's balance in the federal Thrift Savings Plan.

The Thrift Savings Plan already offers a free calculator showing how your retirement savings can be converted into a single lifetime, joint lifetime, or inflation-indexed annuity (and even a few more exotic options related to non-spouse beneficiaries), and I haven't been able to find another company offering better terms. However, those annuities are still technically issued by the private MetLife corporation, and subject to state law (and bankruptcy court).

But since the federal government already runs a guaranteed income program for retirees, the greatest promise of TSP-for-all is the ability to convert accumulated savings into a supplemental Social Security benefit, guaranteed by the federal government to last as long as you do.

What do we want from our retirement system?

Nothing could be simpler than our retirement system as it exists today: the rich do what they can, while the poor suffer what they must. But this rule is not written in the stars, it's the product of very specific decisions made in the development of our retirement policy:

• Employers are allowed to decide the retirement savings options available to their employees, so low-wage employees have less access to tax-advantaged savings vehicles than high-wage employees;
• Social Security contributions are capped so high-income workers have more disposable income to contribute to tax-advantaged plans than low-income workers do;
• Low-income workers who are able to save in individual retirement accounts can be lied to by financial advisors who are not required to act in their best interests.

A simple, cheap, universal, federal Thrift Savings Plan won't solve every problem in the world. But no one defending the current system is even pretending that it's doing an adequate job providing for income security in retirement.

So how about we try something different?

The risks of specialized knowledge

The other day, I received an invitation to an event at the Brookings Institution called "The new American dream: Retirement security." This seemed right up my alley, so I clicked through to see the event details. The description starts off with some generic language:

"The American dream has drawn millions to the 'land of opportunity' and long encapsulated the idea that every citizen has the right to improve their lives. Yet, the current state of the U.S. retirement system may threaten the ability of some to fully achieve the American dream by ensuring their health and quality of life in retirement."

I naturally found myself nodding along, since the inadequacy of our old age insurance programs is a subject near to my heart. The description continued:

"The traditional three-legged stool of retirement—social security, pension, and retirement savings—is transforming into a wobbly one-legged stool,"

so far, so good,

"with personal savings and investment providing the only retirement security."

Wait, what?

Social Security is the only source of income security in retirement

It's been a while since I've written about Social Security, so let's do a quick refresher on how the program works:

• any time between age 62 and age 70, you can claim an old age benefit based on your income in the highest 35 wage-inflation-adjusted years for which you reported earnings;
• the longer you wait after turning 62, the higher your benefit is;
• your benefit will never fall;
• and your benefit is adjusted upward for inflation each year.

Because the benefit is paid in US dollars by the United States federal government, the system can never go bankrupt. This is retirement security.

Personal savings and investment provide income, not security

What is the difference between income and security?First, income fluctuates. If your savings are in a savings account then the interest rate might fluctuate monthly or quarterly. If they're in a CD ladder, then each time a CD matures you're forced to reinvest the principle at the currently prevailing interest rate. If they're in stocks, bonds, or mutual funds, then the dividends and coupon payments you receive will likewise fluctuate along with interest rates and economic conditions.Second, income is risky. If you don't have control over when you sell your investments, you risk selling them at depressed prices, permanently impairing your ability to generate additional income in the future.Finally, income is vulnerable to inflation. The very safest federally insured deposits may pay little or nothing in excess of inflation, meaning to generate real income you need to draw down your principal or invest in riskier assets.No one in their right mind should confuse personal savings and investment for retirement security. So why did the Brookings Institution?

The risks of specialized knowledge

If you said to me, "Social Security succeeded in lowering the elder poverty rate from 35% in 1965 to 10% in 1995, but some elderly people are still in poverty," I would say, "that's because Social Security benefits are too low and the minimum benefit needs to be raised so no seniors live in poverty."If you said to me, "many children in the United States live in poverty," I would say, "that's because children don't earn income, while requiring adult supervision, and we need a universal child allowance that reflects that fact."Specialized knowledge, the kind of knowledge possessed by scholars at the Brookings Institution, makes it very difficult for people to identify problems and propose solutions that address them directly. Once you know that 401(k) accounts exist, but that most people don't have access to them, and most people who do have access to them don't participate in them, then it's the most natural thing in the world to find yourself talking about how to expand access to 401(k) plans, how to increase participation, how to increase the quality of the investment options, how to ensure people are getting unbiased investment advice, etc, and calling that a set of solutions to "retirement security."But those questions are all downstream from, "how do we provide retirement security to elderly Americans?" That's a question we already know the answer to: bigger Social Security checks.Likewise, once you know the Social Security Administration collects more money than it spends and saves that money in a "trust fund" that will be used to pay benefits once outlays begin to exceed FICA tax revenue, and that the "trust fund" will be "exhausted" in 2034, then it's natural to start frantically wondering what combination of benefit cuts and payroll tax increases will be necessary to make the program "solvent."But if all you want to know is "how will the federal government pay for Social Security benefits in the future?" then the answer is obvious: the same way it pays all its other bills, a combination of corporate and individual taxes, estate taxes, licensing fees, seignorage, and debt. My preference would be fewer wars, higher taxes, and less debt, while if you're a Republican your preference might be for more wars, lower taxes, and more debt, but there's no use pretending Social Security benefits pose some unique threat to the Republic. That is the risk of specialized knowledge.

Marco Rubio wants to end retirement security for you and everyone you know and love

It's often pointed out that Americans are the only people in the developed world who don't have access to guaranteed paid leave from their jobs to care for themselves and their families.But that's not exactly right: while we don't have a nationwide paid family leave program, lots of people in America do have access to paid family leave. According to the National Partnership, California, New Jersey, Rhode Island, and New York have statewide plans covering many private sector employees, and many cities and counties extend paid leave benefits to their own employees.The success of those programs raises the obvious question: why don't we have a nationwide paid leave program?

Marco Rubio has a commonsense solution to make your retirement more precarious

That brings me to Marco Rubio's "Economic Security for New Parents Act," which, to be clear, does not provide economic security for new parents. Instead, it allows new parents to claim a cash benefit after giving birth, and then permanently reduces their Social Security old age benefit in retirement.It took me quite a bit of searching to find the actual text of the bill, so I'll save you the trouble and link to it here.To understand how the bill works, you need to keep three numbers in mind:

• everyone is eligible for "early retirement" benefits at age 62;
• people are eligible for "retirement" at an age determined by their year of birth (67 for most people living today, slightly earlier for some Boomers);
• and "delayed retirement credits" if they delay claiming benefits past their "retirement" age.

If you begin to claim old age benefits between your "early retirement" and your "retirement" age, your primary insurance amount is reduced (by 6.67% per year for the first 36 months and 5% per year after that).If you claim old age benefits at your retirement age, you get your primary insurance amount.And if you delay claiming old age benefits past your retirement age, your primary insurance amount increases by 8% percent per year.

This bill would destroy millions of lives

Once you know how Social Security old age benefits work, you can see why Rubio's bill is so devastating:

• people who take parental leave have their early retirement age deferred: "the early retirement age with respect to such individual shall be deemed to be the early retirement age determined with respect to such individual plus the parental leave benefit adjustment with respect to such individual."
• people who take parental leave have their full retirement age deferred: "the retirement age with respect to such individual shall be deemed to be the retirement age determined with respect to such individual plus the parental leave benefit adjustment with respect to such individual."
• people who take parental leave will lose deferred retirement credits: "the Social Security Act is amended by inserting after 'age 70' each place it appears the following: '(or, in the case of an individual described, age 70 plus the parental leave benefit adjustment).'

While these penalties sound similar, it's important to differentiate them. Between 42% and 48% of workers start claiming their old age benefit at age 62. For these workers, the changes to Social Security would mean for every child they have, they're forced to delay claiming their already-reduced old age benefit.Workers who claim their old age benefit at their full retirement age will also be penalized: a millennial eligible to claim their full benefit at age 67 who made the mistake of spending time with their child will instead receive 1-2% less per year, per child, permanently.And a worker who wants to wait to claim their old age benefit until they've maximized their deferred retirement credits will have to wait, not until age 70, but months or years after that depending on the number of children they have.

Social Security is a guarantee of income during old age and disability

Social Security is not a savings account. It's not a trust fund. It's not an investment. You cannot "draw" on it, you cannot "make contributions" to it, and needless to say you cannot "fund YOUR paid leave with a portion of YOUR social security benefits which YOU paid for with YOUR taxes."That is not, and has never been, and never will be how the Social Security Administration works. You cannot "borrow against" your future Social Security benefits. You cannot "withdraw" your Social Security benefits. You cannot "invest" your Social Security benefits.Social Security is a guarantee of income during old age and disability. You can sabotage it if you want to — that's the nature of politics, there will always be people trying to sabotage the income security of the old, the sick, and the poor.But if you sabotage Social Security, the outcome will be lower incomes, for more people, during old age and disability. There's no hack, or workaround, or trick to keep Social Security benefit cuts from making the oldest, sickest, and poorest people in our society worse off.And shame on Marco Rubio for trying.

The Social Security magic trick

I often say I like writing about money because it's an area of human activity that naturally lends itself to my unfortunate literal tendency. Today's post is a good example of what I mean by that.

Everybody knows Social Security retirement benefits rise the longer you delay claiming them

Usually this is framed as an incentive to "delay retirement" or "keep working" past age 62. I've even written before about the effects of late-career Social Security contributions. But today I'm not interested in your career plans, I'm interested in your Social Security claiming strategy.While early retirement benefits are decreased on a month-by-month basis and late retirement benefits are increased on a month-by-month basis, let's dispense with that calculation and use just three values: claiming as soon as you're eligible at age 62, claiming at full retirement age (67 for workers born in 1960 or later), and claiming at age 70:

• at age 62, your benefit is 70% of your full retirement age benefit;
• at age 67, your benefit is 100% of your full retirement age benefit;
• at age 70, your benefit is 124% of your full retirement age benefit.

As an aside, no one uses these values because of the 30-year-old fad for panicking over the solvency of Social Security. But defending Social Security is a political choice, not a matter of accounting, so I am using the figures found in current law.There is one additional nuance that is not especially relevant for reasons that will become clear but I want to mention anyway for the sake of completeness. The dollar amount you will receive at age 70 is not 77% larger than the dollar amount you would have received at age 62, even though 124 is 77% larger than 70: the dollar amount you'll receive is substantially larger, since your "primary insurance amount" is adjusted for inflation in addition to being adjusted for delaying retirement. Note that this is not the wage-inflation adjustment that's applied to your earnings in order to calculate your primary insurance amount. It's a separate, consumer-price adjustment (technically "Consumer Price Index for Urban Wage Earners and Clerical Workers") applied to your primary insurance amount every year whether or not you are collecting a retirement benefit. If you want way, way too much information about this process see this Bogleheads thread.

Nothing up my sleeves

Ready to be dazzled? Let's say you plan to have no earned income beginning at age 62, and you need exactly $14,880 in annual income (mechanically scale this amount according to your own needs).This is, conveniently, your old age benefit if you delay claiming Social Security until age 70 if your primary insurance amount is exactly $1,000 (again, scale according to your own situation). Now, you have three options:

• Claim your old age benefit of $8,400 at age 62 and supplement it with $6,480 in income from your accumulated savings. Using a 4% "safe withdrawal rate," you'll need $162,000 in savings to comfortably retire.
• Claim your old age benefit of $12,000 at age 67 and supplement it with $2,880 in savings at a 4% safe withdrawal rate, which requires $72,000 in assets at age 67. However, you'll need to supply 5 years of risk-free, inflation-adjusted income between age 62 and age 67 (remember, you won't have any earned income after age 62). That will require $74,400 more housed in a nice inflation-adjusted vehicle like Vanguard's short-term TIPS fund, for total assets at age 62 of $146,400.
• Claim your old age benefit of $14,880 at age 70. You won't require any ongoing income supplement from your accumulated assets, but will need to cover $119,040 in expenses for the 8 years between age 62 and age 70.

The longer you wait to start claiming your old age benefit the less savings you need when you stop working.How's that for a magic trick?

Do we have a retirement savings crisis or a retirement income crisis (or neither)?

I'd like to set some facts up on the board:

• In 1966, prior to the Johnson Administration's "War on Poverty," 28.5% of Americans over the age of 65 were living below the poverty line. In 2012, just 9.1% were;
• Since 1978, the number of workers covered by defined benefit pension plans has been in steady decline.
• in 2013, 39.5% of new female filers and 35.6% of new male filers claimed Social Security's old age benefit at age 62, in the year they first became eligible.

If you knew nothing else about the world, you could arrange these facts in different ways in order to draw different conclusions. For example, it might be that early Social Security filers are disproportionately high-income individuals who accumulated a lot of private savings during their working lives, and so don't need to rely on Social Security in retirement, while low-income individuals without private savings wait to claim their higher old age benefit at full retirement age or later, making up for their lack of private savings and lower lifetime earnings.Alternatively, it may be that while defined benefit pension plans have been in overall decline, they remain concentrated in low-income sectors, so early Social Security filers have their retirement incomes "topped up" by defined benefit pension plans.Of course, in reality low-income workers are much less likely to have defined benefit pension plans and much more likely to claim Social Security old age benefits as soon as they're eligible. But elder poverty has still been in decline for the last 50 years.

Social Security replaces a large share of low-paid workers' income

Unpacking this mystery requires a little knowledge of Social Security's old age benefit calculation. Consider a worker making the federal minimum wage of $7.25 an hour at age 62, and assume their annual income has only kept up with wage inflation for the last 35 years. That lets us use the worker's current monthly income of $1,256 as their average indexed monthly income, which produces a primary insurance amount of $915 (90% of $885 plus 32% of $371). That's their monthly benefit at their full retirement age of 66 (assuming they were born in 1954 and turned 62 in 2016), and it replaces about 73% of their gross income. Note that since they won't be paying FICA taxes on their old age benefit, it replaces closer to 79% of their net income.If the worker starts collecting at age 62 instead, their benefit will be reduced by 25%, to $686, a 55% gross or 59% net replacement rate. On its own, a $7,872 annual benefit is not enough to put our new retiree above the poverty line (neither would their full retirement age benefit of $10,980).

Social Security benefits are only taxable for relatively high earners

High-income workers are often advised to delay claiming their old age benefit, for three good reasons: the longer you continue working the more high-income years you can add to your Social Security work record, the longer you delay claiming the fewer penalty months or more bonus months you'll receive on your ultimate benefit, and some Social Security benefits are taxable for high-income households.How do these considerations apply to low-income workers?With respect to the first, we know that wages have been stagnant for lower-income Americans for decades. If you're low-income at age 62, there's no reason to believe your next year's income is going to "roll off" a low-income year in your youth. In fact, thanks to the way the wage-inflation-adjusted federal minimum wage has bounced around in the last 35 years, you may be earning less in wage-inflation-adjusted terms than you were in your youth!As shown above, the second consideration still applies, with your benefit increasing each month you delay filing for your old age benefit.But the third consideration is irrelevant: Social Security old age benefits are only taxable to the extent that your adjusted gross income, plus half your old age benefit, exceeds $25,000. In the case of our minimum wage earner, that sum only comes to $19,188, well below the taxable threshold.That means at age 62 a low-income worker can claim their old age benefit completely tax-free while continuing to work. The old age benefit doesn't replace their income, it supplements their income. You can see this clearly in Chart 5 of this report, showing that among Social Security recipients aged 62-64 (by definition "early" filers), in 2009 Social Security made up about 31% of income, while earnings represented about 38% (older age groups also show significant earnings but the report doesn't separate early filers from full retirement age filers in the older age groups).

Raising the full retirement age is the worst way to "save" Social Security

I'm not a fan of the "Social Security in crisis" narrative ginned up periodically to justify attacks on the welfare state (today's entry in the genre). The United States is a woefully undertaxed country and modest tax increases would solve a slew of problems, including the financing of Social Security. But maybe you're an enthusiast for this crisis narrative! If you think the solvency of the Social Security trust fund in 2033 is a pressing national issue, you might be familiar with some of the options people float to "save" Social Security:

• raise or eliminate the cap on taxable earnings, increasing the amount of money flowing into the Social Security trust fund;
• investing the Social Security trust fund in riskier assets, hopefully improving its long-term returns;
• use a chained inflation measure for cost-of-living adjustments instead of the current unchained CPI, which would represent a modest cut in benefits over time;
• raising the full retirement age.

Of these options, raising the full retirement age is the one that targets the income of the elderly poor most directly. Increasing the full retirement age from 67 to 68 would represent a cut of 9.3% to the old age benefit of someone filing at age 62 (reducing their benefit from 75% to 70% of their primary insurance amount). This would be a permanent reduction in the retirement income of the lowest-income elderly, who are by definition the marginal elderly who have been pulled out of poverty by the program in the last 50 years.It would be hard to come up with a plan more narrowly targeted at the people who need Social Security the most.

What is a retirement savings crisis?

The finance industry has gone to great lengths to convince American policymakers that the country is undergoing a "retirement savings crisis." There are two reasons they've done this:

• the finance industry is in the business of managing money, so the more urgency they can gin up about the savings rate, the more money they can convince people to save;
• the finance industry is populated by the kinds of wealthy individuals who receive the majority of the benefit of the tax-advantaged savings vehicles they've convinced policymakers to create.

Now to be clear, I don't have anything against saving a high percentage of your income; I save a high percentage of my income. But there are other things you can do with money as well, like using it to pay down debts. But using your savings to pay down debts instead of invest is a double-whammy to the finance industry: fewer assets to manage and less debt to charge interest on!In fact, it does not seem to me that we have a retirement savings crisis. We have a glut of tax-advantaged savings vehicles, and consequently a shortage of collected tax, which is certainly a problem, if not a crisis. We also have a fairly extensive scam economy featuring things like reverse mortgages, deeply conflicted financial advice, and variable indexed annuities. Those are real problems we could come up with policy solutions to, but they're primarily problems of elder abuse, not elder poverty (although elderly people can certainly be impoverished by abuse).

A retirement income crisis demands retirement income solutions

What is true is that we have a retirement income problem, which is real but manageable. As mentioned above, 9.1% of those over the age of 65 were living below the poverty line in 2012. Obviously age 65 is too late to build a Social Security earnings history or save enough money to live on in retirement. I'm not familiar with the exact composition of this group, but there's no difficulty imagining how someone would fall into this category:

• Workers whose earnings weren't reported because they worked off the books, or were improperly reported due to employer fraud. They may have a spotty official work history and underreported wages that produce a lower old age benefit than they'd otherwise be eligible for.
• People who didn't work and don't have a spouse's work history to rely on. This may be as complex as not wanting to attract attention to an undocumented spouse or child, or as simple as being in a gay relationship and losing your partner before gay marriage was legalized.
• The long term disabled elderly who lost the ability to work before accumulating a sufficient work history and don't have family to rely on.
• Low-income workers like the one described above who file for an old age benefit and leave the workforce at age 62.

This is a disparate group with different needs, none of which are tax-advantaged savings vehicles:

• Workers whose income isn't being properly reported need vigorous enforcement of our labor laws (and not to fear deportation). We also need a streamlined system for reporting household and casual workers' income, so employers have less incentive to hire people off the books.
• Developing a way for unmarried couples to claim their partner's work record would need to be carefully thought out to prevent potential fraud.
• The long-term disabled elderly are a unique problem that can probably best be solved with simple cash payments.
• Low-income early retirees can be easily helped by lowering the full retirement age back to 65 (thereby reducing the penalty for early retirement) or following Canada's example and supplementing the earnings-based old age benefit with a fixed payment to all retirees, taxable at your marginal income tax rate (Canada's is a bit more complex than that).

What all these solutions have in common is that they address the real problem of insufficient retirement income, rather than the imaginary problem of insufficient retirement assets.