Social Security Crisis in 2018? Ridiculous

Take a look at this Washington Post article by Jonathan Weisman, which the big blogs have justly pounced on.

The initial framing of Weisman’s article is just way, way off. Social security in no way faces a day of reckoning in 14 years, as Weisman implies. Rather than taking the Administration’s talking points seriously, the press should simply highlight how ridiculous they are.

Consider the following example: Suppose I bought a 15 year treasury bond that matures in 2020. The Treasury does not sell 15 year bonds, so this is a bit of a thought experiment. Do I have a problem in 2020? Do I face a day or reckoning in 2020?

Of course not. I have a financial asset. I am better off in 2020 than I would have been had I not bought the bond.

Social security expenditures will exceed its payroll tax revenues in 2018. So what? Social security has built up assets over time, since from the 80s on payroll taxes have exceeded social security benefits. It bought a lot of US government bonds with its surplus. Social security can sell some of its assets to cover the gap between its expenditures and its revenues. That is the point of saving -- it lets you spend in excess of your income in the future. The Trustees forecast those assets will last til 2042, using VERY conservative assumptions. The CBO forecasts those assets will last until 2052 [Corrected, Brad -- I initially eyeballed the graph, instead of looking at the text, my apologies]. That’s pretty good.

Unless the US treasury cannot make good on its promises -- something that truly would change international financial system -- social security does not face a day of reckoning in 2018. When a bond you own comes due, you have the right to redeem it -- or in the case of social security, at least to collect the interest on your bonds (remember that right now social security is both taking in $68 billion more from the payroll tax than it spends on benefits and is lending the $85 billion interest it recieves on its existing holdings of bonds back to the rest of the government ... in total, social security is providing over $153 billion in financing to the rest of the government)

Don’t forget that the government -- the non social security part -- has expenditures well in excess of revenues RIGHT NOW. Dick Cheney apparently thinks cash flow deficits that have to be financed by issuing tons of debt don’t matter, but cash flow deficits than can be financed by drawing on the interest from your stock of existing assets are a real problem ... interesting financial logic.

An aside: Anyone writing about the social security ought to read both the Trustees and the CBO report. I do have one quibble with the CBO though: its graphics subtly suggest that social security is in much worse shape than it really is. The main graphs focus on payroll taxes v. expenditures, and the graph has been chopped off to magnify the apparent size of the shortfall (the graph starts at 3% of GDP, not 0%). The CBO also buries the key graph showing when the Trust Fund runs out -- i.e. when the gap between payroll taxes and promised benefits can no longer be financed social security’s assets. Looking just at the graphs, it is not obvious that the CBO thinks the system is in better shape than the Trustees ...

In 2018, social security won’t be able to lend its surplus to the rest of the government, and the rest of the government will have to adjust. That is a problem if you don’t like the income tax, because the rest of the government is financed largely by the income tax. Income taxes have to rise, or non-social security spending will have to fall. But so long as the US government is not planning on defaulting, it is not a problem for social security. The payroll tax does not need to increase in 2018, retirement benefits do not need to be cut, there is no problem, let alone a day of reckoning. The Treasury just has to start paying interest on all the bonds social security has bought ...

I think part of the problem the press has is that it seems like the retirement of the baby boom should cause problems for social security. The number of retirees will grow, the numbers of workers per retirees will shrink, and social security benefits will increase from around 4.4% of GDP to about 6% of GDP. How could a constant payroll tax be able to finance the retirement of the baby boom? On the surface, it does not make sense.

The funny thing is that US government was actually responsible in the 1980s, and planned ahead for the impact of the baby boom on social security. The retirement age was increased (a de facto benefit cut), and payroll taxes were raised. Right now the payroll tax takes in more than is needed to pay for current benefits -- revenues are around 5% of GDP and expenditures are around 4.4% of GDP -- and the surplus is lent to the rest of the government. It is a loan, not a grant. The government has to pay it back.

Rather than debating the "problem" created when social security starts to redeem its bonds, we should be debating how to fix our real problem -- the fiscal deficit. Social security now takes in more than it spends. The rest of the government now spends way more than it takes in. On realistic assumptions, it will run a 3.5% of GDP deficit from now til eternity unless something changes -- and even bigger deficits from now til 2014 if you took out the social security surplus.

These ongoing fiscal deficits are the real financial problem facing the US -- not the projected gap after 2042 (or 2055) of 1.5% of GDP or so between payroll taxes and trust fund assets and social security benefits that underlies concern about social security’s long-term solvency. I’ll put it differently: unless something changes, the rest of the government will go broke long before social security has any problem paying all its projected benefits.

(Update: Paul Krugman covers much of the same ground, but does so in far fewer words)

There is an honest case for getting rid of social security, though not a honest case that social security faces a day of reckoning in 2018. People like Pete Peterson think we should not be promisng to transfer 6% of GDP to the baby boom, and want to cut benefits -- even if though we can pay for them with the payroll tax and the Trust Fund’s bonds. Cutting social security benefits would make it easier to pay the rising costs of Medicare; payroll taxes now going to social security benefits could go for health benefits. Fair point. But there is also an honest case for keeping social security as it is -- perhaps with a few tweaks -- as a way to protect against income volatility in retirement.

Take a look at the series of excellent articles on the growing income volatility by Peter Gosselin of the Los Angeles Times that Brad Delong has justly lauded. They key point -- the average American, even average Americans at the upper end of the income distribution -- have a much greater risk of seeing a sharp fall in their income than in the past. Take a look at the article in the Economist lamenting limited social mobility in the US -- there is a significant risk that any individual will not rise out of the socio-economic group he or she is born into. Put the two together and you have a picture of a society with lots of volatility, but not one where -- in aggregate - income volatility is matched by lots of social mobility.

As Gosselin notes, lots of the formal and informal protections that used to limit income volatility have withered away; working americans are shouldering more and more risk.

Throughout this series, The Times has sought to make sense of an American paradox: why so many people report being less financially secure even as the nation, by many measures, has grown far more prosperous.

The answer, the newspaper has found, lies in the shifting of economic risks from the broad shoulders of business and government to the backs of working families.

Over the last quarter of a century, many safeguards that people once counted on to shield them from financial harm have been weakened or completely lost. These include formal protections such as guaranteed corporate pensions and state and federal unemployment benefits. And they include informal ones, like the loyalty that employers once showed their workers by offering secure jobs with relatively little prospect of long-term layoff.

The fact that social security -- unlike many private pensions -- is fully portable limits the risk that lots of "job churn" and income volatility during an individual’s working years will translate into limited retirement income. Social security’s guaranteed income also protects against too much income volatility after an individual retires. Private pensions once were structured like social security -- they assured a guaranteed income. Not any more. You get just what you stuffed away in your defined contribution/ 401 (k). Some folks may do well (if you bought stocks in 1980s and sold them in 2000). Some folks not so well (Say if you bought the NASDAQ at 5000 ... ).

The fact that social security benefits are not correlated with the stock market -- or the bond/ housing markets for that matter -- is another one of the social security system’s key virtues. Most retirees already have plenty of market risk from their private pensions. Social security both diversifies the sources of an individual’s retirement income, and insures against a range of other risks: the risk an individual might make less than he or she hoped during their working life, the risk an individual may be disabled and unable to work, the risk an individual may invest their savings poorly (remember Enron employees with all their retirement assets in Enron stock?), the risk an individual might outlive their savings, etc.