Will partial privatization of social security pass the global test?
This post is primarily about the implications of partial privatization of social security on the budget, not about its overall merits.
For what it is worth, I personally think good old pay-as-you-go social security has an important role to play in a world where old fashioned welfare capitalism and defined benefit pension plans are giving way to 401 (k) type accounts, and more and more people’s retirement income already depends on what happens in the markets. There is something to be said for not putting all retirement eggs in the same 401(k) basket. The "insurance" aspect of social security is important -- insurance against bad financial investments, insurance against living longer than your savings, insurance against just earning less over a lifetime than you expected.
The most talked about partial privatization proposal apparently would divert 2% of taxable payroll (about 16% of social security tax revenue) into private accounts. That reform would get rid of social security’s cash flow surplus and would increase the consolidated budget deficit. Right now we depend on foreigners -- notably foreign central banks -- to finance our budget deficits, so it is worth asking if such a reform could pass the global test ... any debtor that is contemplating increasing its debt generally is well advised to ask its creditors first.
Indeed, one of the ironies of the current debate is that many on the American right now have an institutional position that a 1.5% of GDP cash flow deficit in social security after 2042 is a presssing national problem that needs to be solved, pronto, while the roughly 4% of GDP current fiscal deficit (a deficit that woudl be bigger if not for social security’s cash flow surplus) "doesn’t matter" and a 6% of GDP current account deficit TODAY is just a sign of America’s economic success.
So every time I see social security’s $10 trillion hole -- a hole that the Social Security trustees incidently have at only $3.7 trillion in present value terms over the next 75 years mentioned to support the argument that Social Security needs to be reformed now, I want to scream -- "the entire hole comes from cash flow deficits after 2042." 2042. The trade deficit will cause problems a lot faster. If current trends continue, the United States’ external debt will reach $10 trillion in 2010. 2010.
I suspect the United States’ foreign creditors will be less than impressed if we make our current fiscal problems worse in order to address a burning fiscal gap in 2042. The details of my argument follow:1) The key fact about social security today is that it is running a significant cash flow surplus -- $180 billion in 2005 according to the Trustees Report. About 1/2 the $180 billion comes from taking in more in taxes than it pays out in benefits, about 1/2 from interest on its bonds (sidenote -- social security is getting about 6% on its government bond portfolio, and got a bit more in 2001 and 2002 -- not a bad investment given what has happened to stocks over the past four years).
2) On a cash flow basis, the Social Security system looks pretty darn healthy for the next ten years. The trustees forecast a cash flow SURPLUS of $280 billion and assets of $3.6 trillion in 2012. Nouriel and I forecast -- on current trends -- a US trade deficit of $1,300 billion (7.5% of GDP) and US external debt of $14.6 trillion (84% of GDP). Interest on that external debt will cost the US $800 billion, and that interest plus some tranfer payments will produce a current account deficit of 2,200 billion. That sounds extreme, but the reasonable and meticulous Catherine Mann has trajectories that show even bigger deficits sooner if nothing changes. Again, tell me which is the bigger short-run problem, the growing trade deficit or the social security system’s growing cash flow surplus?
3) The Trust fund is meant to pre-fund some of the baby boom’s retirement. Around 2015, social security benefits will start exceeding revenues. But thanks to the Trust Fund, the system will have the funds to pay all programmed benefits until 2042, or it will if the rest of the government makes good on its promise to the Social Security trust fund. That explains some of the right’s desire to cut social security benefits. Without a benefit cut, the progressive income tax will have to go up over time to pay the social security system back -- the trust funds holdings of government bonds are effectively a way of shifting some of the costs of social security benefits off the payroll tax and onto the income tax from roughly 2015 on.
4) One important note -- partial privatization could easily make the social security system’s long-term solvency worse, not better. Private accounts reduce the system’s revenues, and if benefits are not cut by enough to offset the reduced revenue, social security will start running a cash flow deficit well before 2042 and the long-run actuarial hole will get bigger. That is why the honest "we are promising future retirees too much" crowd -- people like Pete Peterson -- seem far more interested in reforms that reduce benefits than they are in reforms that reduce revenues, like private accounts.
5) In the short-run, partial privatization of social security will significantly increase the reported budget deficit. The reform effectively will get rid of the current cash flow surplus in the only part of the government running a surplus, and thus add to the government’s consolidated deficit. Consider the following math.
In 2005, social security is forecast to have tax revenues of $605 billion (including revenues from taxation of benefits), and expenditures of $518 billion, for a surplus of around $87 billion. It is also expected to receive $96 billion in interest on its EXISTING holdings of bonds in the trust fund. The cash flow surplus and the interest would both be invested in government bonds, so the Trust Fund would add $183 billion to its assets. The $87 billion also reduces the reported on-budget fiscal deficit dollar for dollar. Without a similar contribution in 2004, for example, a $412 billion deficit would have been $490-500 billion.
Diverting 2% of the payroll tax from the system would reduce revenues by $114 billion in 2005 - money that would now go into private accounts. Rather than a cash flow surplus of $87 billion, the system would have a deficit of around $27 billion. That could be covered by using some of the interest on the trust funds existing assets, but that means that only $69 billion of interest would be added to the trust fund.
The net effect -- a reported budget deficit that rises by $114 billion, $114 billion in new private accounts and a trust fund that adds $69 billion rather than $183 billion to its assets, and thus a trust fund that will run out much, much faster ...
6) How will the folks currently financing our cash flow deficit respond to a $115 billion increase in cash flow deficit? If they are not willing to increase their financing at current interest rates, then interest rates will have to rise to attract the financing needed to cover the government’s current cash flow deficit. Some of the needed financing may come from the private accounts -- some folks may put their money back in government bonds -- but some will have to come from the "market" (i.e. foreign central banks, unless something changes).
7) The kicker, of course, is that if the higher cash flow deficits lead to higher interest rates, the higher interest rates will reduce the financial wealth of almost everyone by more than small private accounts (at least initially) "increase" in their wealth. Higher interest rates imply falling bond prices, falling agency bond prices, falling house prices ... even falling equity prices.
Those that think that there will be no effect on interest rates from a larger deficit today because the market already has priced in the costs of the unfunded social security liabilities after 2042 are smoking something -- today’s markets are all about predicting what will happen over the next month or maybe the next year, not looking 40 years ahead. And if a typical bond trader (falsely, in my view) believes social security won’t be around when he gets old, logically, he also wouldn’t worry about the costs that would arise from future social security payments ... This is all the more the case because the benefit cuts associated with the "private accounts" reform are unlikely to be large enough to cover the reduction in the system’s revenue, so there is likely to be an even bigger unfunded mandate after partial privatization than before.
8) The way to avoid a large deficit from producing higher interest rates and falling asset values, of course, is just to borrow more subsidized money from foreign central banks. avoiding any interest rate increase by borrowing from abroad lets the financial industry have its cake (private accounts) and eat it too (low interest rates/ high asset values). See the past three years -- financing from central banks has financed our budget deficit at a very low cost. But that implies a bigger trade deficit, faster foreign debt accumulation and a quicker plunge into external insolvency ...