by Bruce Webb
In the past I have travelled or lived in most parts of the U.S. and in visiting the local tavern, bar or lounge am offered some choices in types and serving sizes of beer. Most places by answering ‘bottle’ to the question ‘bottle or draft’ you are locked into a measured 12″ serving, no more no less, which is why in most places outside Texas and maybe Colorado you can’t go too wrong ordering ‘a bottle of Bud’ (though you might get some dirty looks by asking for a cold glass). But if you chose ‘draft’ you can be in uncertain territory. Most places will give you some choice between small, large and pitcher where the small is somewhat smaller than the 12 ounces you get in a bottle and the large is modeled on the English pint. But as experienced travelers/drinkers understand all too well ‘pint’ in too many places has little relation to its standard definition of ’16 ounces’, in fact in a lot of places if you order a 12 ounce bottle and a cold ‘pint’ glass the resultant pour gets you suspiciously close to the rim. Which is why in my own State of Washington you never encounter ‘pints’ outside an Irish Bar or brew pub, instead we drink ‘pounders’. Now since a standard pint of water weighs exactly a pound the implication is that your ‘pounder’ actually contains sixteen ounces, but it is not like this is an actual violation of truth in advertising laws, most people know the score, you are lucky to get 14 ounces in a pounder and 42 ounces in a pitcher.
What does this have to do with Social Security? Well it offers a different way of examining what ‘solvency’ means in relation to SS: should we be worried how close the pour is to the rim? Or to the actual contents of the serving? And the answer to that question ends up having some surprising constraints on your policy options. Actual Social Security discussion under the fold.
Social Security solvency is mostly reckoned in terms of income to cost ratio and presented as either a percentage of projected payroll or GDP over the given projection period. But neither that projected income or projected cost is fixed, each instead is the product of a formula or formulae whose factors are determined by actual economic and demographic outcomes compared to the baseline assumptions. And it is in my view the failure among most commentators to fully grasp that variability that distorts the policy decision field.
Starting at the cost end we have this CBO study from 2003: The Future Growth of Social Security: It’s Not Just Society’s Aging. If we examine its figures and tables we can see that fully 45% of the projected growth in Social Security over the 75 year projection period is due to improvements in real benefits delivered compared to today, in terms of our metaphor the Trustees are promising us that if all things go right our ‘pint’ will actually have more content ounces going forward than today and cumulatively a lot more, at the end of the day we are drinking something closer to those beer steins you see at Munich’s Oktoberfest. Prost! But a look at the revenue end gives us some more sobering news, while Social Security left unchanged can deliver 100% of scheduled benefits right up to 2037, in our metaphor a pour right up to the rim, after that projected revenues only deliver a 78% pour. Hmm, downer. But how down should we actually be? This graph from Dean Baker at CEPR (though designed to illustrate a different point) is instructive: Social Security and the Washington Post: Who Is Going Down First?. What the graph shows is a fairly dramatic cutback in benefits at Trust Fund depletion, in numeric terms 22%. But note that the resulting beer stein STILL holds more ounces of beer than the one retirees get today, indeed about 20% more ounces. Moreover the pour starts improving in short order and by mid-century is as big as it ever was and then stein and pour continue to grow throughout the projection period. Now if we take the 2010 to 2037 line and extend it we see that our retiree/drinker of 2085 coulda/woulda been drinking from a full stein some 25% bigger than the one he ends up with. But still his daily beer ration remains more than double of what it is today.
Most policy discussions today hyperfocus on ways to get the stein and the pour into perfect alignment. Social Security ‘reformers’ tend to focus on ways to shrink the stein to meet the projected pour, preferably without a sudden transition that will upset the customers. Veteran beer drinkers will recognize this approach, typically bartenders wanting to skimp on beer serving sizes simply swap out their ‘pints’ with ones with gradually thicker sides and bottoms, the result looks the same and feels the same in your hand, but over time just doesn’t deliver the same zip. On the other hand supporters of traditional Social Security like Angry Bear’s Dale Coberly focus on ways to keep the pour right up to the rim of the growing stein, in Dale’s case by pointing out that the cost to the customer for that brimful of beer is just pennies per week. But I suggest what is important is not the alignment of brim and pour but the actual amount of the pour.
This post was inspired by an exchange between Dean and Henry Aaron of Brookings that I was privy to in which Aaron made the following observation in relation to the effect of higher real wages or an adjustment in the payroll cap adding to the total Social Security wage base:
A larger tax base has some effect on benefits. Since 1972, the system gains only very little, because a higher wage base raises taxes today and benefits later on; the system makes money on the time-value of money, but nothing else (in the pre-indexing days, the story is more complex).
And if you define a balanced ‘system’ with a perfect alignment between brim and pour this is quite true. But if you focus on actual pour size that ‘nothing else’ becomes something pretty significant. If we turn again to Dean’s figure and plug in higher real wages and a larger wage base oveall the effect is to drive the dark blue benefit line into a steeper rise. This does not necessarily delay the time of Trust Fund Depletion or the percentage of reset but it does or should change the resultant pour, from the perspective of the worker the same percentage cut from a higher baseline means more beer today and more beer than projected in the long term.
Some beer drinkers feel cheated if their ‘pint’ is not filled to the point that it overflows onto the serving tray as it is being brought to the table. Me I’ll just take that honest pint. Now looking forwards I would want as big a pour as I could get in my growing stein but my concern would be measuring from the bottom of the stein to the top of the pour and not the top of the pour to the rim of the stein, in the end I am a beer drinker wanting my fair serving whatever the variations in the glassware.