« Retirement Liberty | Main | Libertarian Rightwingery »

February 24, 2005

Transition Cost Confusion

For the past couple of weeks, Matthew Yglesias has been on a kick about Social Security transition costs. He elaborated on the subject here, but ironically, his critique is at least as sloppy as he accuses the privatizers' arguments of being.

Matt appears not to have fully grokked the distinction between cash and acrual accounting. Companies (and other organizations) typically publish two different statements of their financial condition. One, is a profit-and-loss statement, which shows the revenues and expenses of the firm over a specified period, and whether the net income is positive or negative. The other, is a balance sheet, which allows investors to evaluate the firm's net worth by showing assets and liabilities. Although these two statements are linked (since each year's P&L statement affects the next year's balance sheet) they are separate and distinct concepts. The profit-and-loss statement is important to make sure the company can pay its bills in the short term. The balance sheet shows whether the company will make a profit in the long run.

So Mike Tanner's claim that there are no transition costs is a statement about the nation's net worth. It says that in the long run, personal accounts will improve the nation's fiscal position. Despite Matt's implication to the contrary, that is not at all inconsistant with a discussion of how to finance those costs in the short run--a discussion about cash flow.

So our "story" hasn't changed. We argue simultaneously that in there are no transition costs in the long run, and that in the short term, funding them will be expensive. That is, of course, because the up-front costs will be offset by large savings down the road.

The confused, handwaving nature of his math is most obvious in last week's Prospect column:

Privatizers want you to believe that the costs of their plan merely reflect the cost of closing the long-term shortfall sooner rather than later. In fact, the two costs are unrelated, as you can see by contemplating what would happen if we privatized a system that didn't have a long-term financing gap. Hypothetically, one year's payroll-tax revenues would fully finance that year's benefits each and every year until the end of time. Then, one day, you start allowing workers under 55 to divert their payroll taxes into individual accounts; those workers accept an offsetting reduction in benefits when they retire. The problem is: The year after privatization is implemented, payroll taxes will plummet; but benefits won't fall at all. For 10 years, revenue will keep dropping, and expenditures will stay the same. By year 11, the first benefit reductions will kick in, but the overwhelming majority of retirees will still be drawing full benefits under the old system -- except without anyone paying the taxes to finance their benefits. Around 50 years later, all of the people who were over 55 when the new system was introduced should be dead, and the gap will go away. The debt, however, won't ever go away unless some huge some of money is found elsewhere. That, as you'll recall, used to be "where the budget surplus comes in" -- until the surplus went away and until privatizers first started talking about it.

"Around 50 years later," the cash-flow gap won't just "go away." It will turn into massive surpluses, since the system will still be taking in the same amount of revenues but will have very few people still dependent on traditional benefits.

Whether the long-term surpluses will be sufficient to fully pay off the bonds depends on the details of the plan and the performance of the stock market over time. But the way the math works out, as long as the accounts earn at least as much as treasury bills (and no one seriously argues they're going to underperform T-bills over long periods) personal accounts improve the fiscal position of the Social Security system by reducing the government's costs in giving workers a given level of benefits. But you can only see that if you look at the system's long-term balance sheet rather than focusing on cash flow during an arbitrarily short budget window. Over any arbitrarily short period, you're counting costs while ignoring offsetting benefits that occur at a later period of time.

I'm puzzled at the assertion that Cato and other reformers haven't thought this through. The math just doesn't work out the way Matt thinks it does, and it's bizarre to accuse us of "deep unseriousness" when his exposition of the subject is so sloppy.

There are legitimate (but completely unrelated) cash flow concerns with respect to personal accounts, but this post is already too long so I'll address those in a future post.

Posted by Tim Lee at February 24, 2005 5:50 PM

Comments

Good post. But first, I think you're giving the government too much credit w/ regards to their accounting as they don't keep a true income statement. In the private sector, the income statement isn't computed on a cash-in/cash-out basis for a given period. They make an attempt to match expenses with revenue received (hence the concept of depreciation). The statement of cash flows is then used to connect the income statement with the balance sheet. It's much more sophisticated than how the govt is keeping track of it's books.


Also, one question. In the example above I agree that the system in Matt's oversimplified example would run surpluses if the accounts grew enough so that people could forgo their traditional benefits. But if the accounts performed equally to govt bonds, then there wouldn't really be a savings to the government in my understanding. It would be a wash financially, yet still a benefit since you would now have property rights over a portion of your retirement funds.


Putting numbers that work behind Matt's La-La Land example (so that cash-in and cash-out match forever) let's say that everyone starts working at 20 and retires at 55 contributing $100 a month. People then live until they are 90 receiving $100 a month, and you have a constant number of workers to retires forever (lots of fake assumptions). If everyone below 55 can suddenly put 20% of their taxes in a private account, then the govt will need to borrow $20 a month to cover the difference for each person. Now 50 years down the road if the private accounts' performance only matches that of what you got in the previous system, then people would be getting $80 from taxes from the younger generation, while getting $20 from their personal accounts. But here the govt is collecting $80 a month and paying out $80 so they won't be running any surpluses. It only improves their financial position in the sense they are on the hook for less if something were to go wrong.


Of course this is way oversimplified but does illustrate the point that the real problem is the ponzi scheme aspect where you have a smaller base of workers who need to pay an ever increasing amount to retirees who are living longer and longer. No real saving is getting done. Getting people to rely on their own savings for retirement by putting it in real assets is the only long-term solution to such a problem.

Posted by: Peter at February 25, 2005 6:56 PM

Tim,


Why do you say that 50 years later that gap will turn into a surplus? In Matt's model at that time there are no more revenues.


Make it simple. Suppose people can be in one of two states, state 1 (S1) or state 2 (S2). In S1 they pay into SS. In S2 they receive SS benefits. At any given time they are perfectly matched so in every period those paying in pay the exact amount needed for those receiving benefits. Half the population is in S1 and the other half in S2. It is a perfectly sustainable system.


Then imagine all those in their S1 are allowed to fully divert their payments to private accounts. That means there isn't a dime left to pay people in S2. So the govt borrows the money. When the S1 group moves to S2 there is no need for any borrowing but the debt is still there. There is no more money coming in.


Let's finish this hypothetical. Imagine three periods. In the first period half the population is in S1 paying 100 to those in S2, perfectly balanced. In the second period those in S1 are allowed not to pay. The group in S2 (which was S1 in the first period) is now lacking 100. So the government boorows it and pays them. The S1 group invests in private accounts. In the third and final period S1 and S2 no longer exist. Those that would have been in S2 now get the money from their own accounts and those that would have been in S1 pay into their own accounts rather than pay the S2 group. But the government still has the 100 debt. There is no surplus to pay for that.

Posted by: GT at February 25, 2005 9:45 PM

GT: every privatization proposal I've seen diverts only a portion of payroll taxes to personal accounts. Under Cato's plan, for example, the "employee half" 6.2% of revenues goes to the personal account, but the "employer half" stays in the system to pay the benefits of current retirees (as well as survivors and disability benefits). That means that once all the retirees in the traditional system have died, that "employer half" becomes a surplus available for paying down bonds.

Now, perhaps Matt was imagining a system in which the full 12.4% payroll tax was diverted into personal accounts--which is a larger account than any of the proposals being considered in Congress. In that case, he'd be right that the proposal would worsen the government's finances. but that's just because you made the system dramatically more generous. Even under very conservative assumptions, a 12.4% personal account would offer much larger benefits than what Social Security promises. It's not surprising that a proposal that massively increases benefits would be more expensive than the status quo.

Posted by: Timothy at February 25, 2005 10:53 PM

How would the 6.2% become a surplus? It's still needed to pay benefits. See my #s above. Maybe I'm missing something. If so let me know.

Posted by: Peter at February 26, 2005 12:10 AM

Peter: I'm not sure I follow how your numbers relate to the 6.2%. There are two 6.2% taxes--one paid by the employer, the other paid by the employee. Under Cato's plan, the employee portion is diverted to a personal account, while the employer portion is used to pay beneficiaries.

To use a simplified example, if all workers under 40 switch to personal accounts in 2010, by 2060 (when those workers would be 90) virtually everyone who is retired will not need to draw benefits from the traditional system since they will have personal accounts to draw from. That means that the 6.2% employer share will be more than sufficient to pay the few over-90 folks still dependent on the traditional system.

Posted by: Timothy at February 26, 2005 12:38 AM

Ah, followed this link from Matt. Yes, this was not clear to me -- that you could pay comparable benefits without diverting the full measure of the tax payments. It's of course blindingly obvious when you think about it straight... the diverted payments to the "future self" appreciate, whereas the traditional system is an immediate payment so it doesn't appreciate. So a payment of X to the future self is really worth more than X, while a payment of X under this system is just worth X.



I know that's quite elementary but it had me confused so just in case anyone else is too.



This was definitely not made clear from Matt's example. but I still respect him as a thinker.

Posted by: mike at February 26, 2005 1:18 AM

To use a simplified example, if all workers under 40 switch to personal accounts in 2010, by 2060 (when those workers would be 90) virtually everyone who is retired will not need to draw benefits from the traditional system since they will have personal accounts to draw from. That means that the 6.2% employer share will be more than sufficient to pay the few over-90 folks still dependent on the traditional system.

I agree with this. But this assumes that the 6.2% you have invested in a personal account grows to a level that allows you to forgo traditional social security. In that case, then yes, the employers share would represent a surplus to the system.

But if you are diverting your 6.2% into a personal account that earns the same as a T-Bond, when you retire the amount in the account would provide half your traditional SS check. The other half of your retirement check is going to come from the traditional system. So in Matt's example where he assumes no financing gap (which I tried to replicate above), there is no surplus.

Yes individuals would now have property rights over half of their retirement check, and yes the government is responsible for less, but they are not running surpluses. You'd still need a way to finance the debt that Matt talks about because the "surpluses" in Matt's world would have been eaten up by the first generation of retirees who essentially got a free ride.

Now, I don't think it's a valid example that Matt uses since we are dealing with all the demographic issues and since we don't have a system where "one year's payroll-tax revenues would fully finance that year's benefits each and every year until the end of time," but he is right about the debt issue if you grant him a T-Bond return on the personal accounts in his hypothetical world.

Posted by: Peter at February 26, 2005 2:47 AM

Peter,

I think you've got a point. Here's the thing: if we view things from the perspective of just one generation and assume only treasury bond returns, personal accounts are a wash-- generation 1 borrows the money it needs to open the accounts, earns exactly as much interest as the bonds cost, and can pay off the bonds with the money in the personal account. But of course, as generation 1 is retiring, you've gotta borrow the money all over again for generation 2. So in this sense, you're right--if the personal accounts don't earn any more than treasury bonds and there are no benefit cuts, there will always be bonds on the government's books, because each generation has to pre-fund its own expenses and so can't afford to pay off the debt of the previous generation.

At the same time, there will always be money in the personal accounts that offset the value of the debt. So if you view the whole system--personal account value minus bonds--personal accounts don't change the net fiscal position of the system. It's just that at any given point in time, the debt will be on the books, while the reduced Social Security obligations will be in the future.

Of course, privatization would be almost pointless if we expected personal accounts to get the same rate of return as treasury bills. Fortunately, equities have out-performed bonds for more than a century, so in the real world, personal account proposals do improve Social Security's finances by allowing higher benefits with the same tax revenue.

Posted by: Timothy at February 26, 2005 3:30 AM

Tim,


As I read your posts here and at Matt's site I think I understand better what you were trying to say. But it is not exactly what you said.


You claimed that Matt was wrong, that his exposition was sloppy and that his math had a confused, handwaving nature to it. But in fact Matt was perfectly right.
Matt's model, as I showed above, leads to transition costs. In my example workers were allowed to set aside 100% of their SS payments. But the model reache trhe same result if they set aside only 50% (or any other number). If workers set aside X and their benefits are reduced by X there will alwys be a transition costs that has to be paid. There will be no surpluses because the drop in benefits is perfectly matched by the drop in revenues.


Your model is different than Matt's. You assume that workers reduce their payments by X and see their benefits cut by Y, where Y is some number bigger than X. Then you assume that the difference is made up by greater returns on the private accounts. Of course, under those assumptions, your model works.


Leave aside the issue of whether private accounts will perform well enough to cover the gap (there is a reason why you see those "past performance is no guarantee" lines). Even assuming you are correct on that this does not mean there is no transition cost. Your model has transition costs but claims to have found a way to pay for it. But that's very different than what you posted. You said there is no transition cost. Maybe you meant no net transtition costs once the greater private account returns are considered. But, again, that is not what you wrote.


In fact your model pays for the transiton cost by taxing the greater private account returns. This can be done by either directly taxing the returns, or indirectly by reducing the benefits. Economically it is exactly the same, in both cases the total return is reduced to pay for transition costs.


In the end Matt was perfectly right, there are transtion costs that have to be paid somehow. Your post could have made that clear. You were wrong to say Matt had his math wrong when in fact it was perfectly correct.

Posted by: GT at February 26, 2005 8:36 AM

GT: I think you're about right, and arguing with y'all about it has clarified my thinking somewhat, for which I'm grateful. Check out my latest for what I think is a clearer and more accurate statement of the point I was trying to make.

Thanks!

Posted by: Timothy at February 26, 2005 11:37 AM

Edmonton dominated the Carolina Hurricanes on Saturday night and the 4-0 margin in Game 6 makes it hard to imagine the Oilers not hoisting hockey's Holy Grail above their heads in less than 48 hours. And it would not come as any shock to see defenseman Chris Pronger, who had another 31-minute night, take the honors for the Conn Smythe Trophy as the playoff MVP.

Posted by: antoiantnette at June 19, 2006 2:51 AM

Edmonton dominated the Carolina Hurricanes on Saturday night and the 4-0 margin in Game 6 makes it hard to imagine the Oilers not hoisting hockey's Holy Grail above their heads in less than 48 hours. And it would not come as any shock to see defenseman Chris Pronger, who had another 31-minute night, take the honors for the Conn Smythe Trophy as the playoff MVP.

Posted by: amaleamaa at June 20, 2006 3:23 PM

quik cash quik cash

Posted by: quik cash at September 12, 2006 11:46 AM

www.chantixhome.com is a vast reservoir of chantix, the formidable foe of smoking addiction and you must browse through it if you earnestly desire to keep yourself at a safe distance from the numerous smoking induced harmful consequences.

Posted by: chantix at September 15, 2006 1:47 AM

www.levitrablisss.com is the suitable source to procure levitra, the anti-impotency medicine which even efficaciously treats spinal cord injury induced erectile dysfunction.

Posted by: levitra at September 15, 2006 1:52 AM

ionolsen17 Your home page its greathttp://www_3_2.gmail.com/
http://www_3_3.gmail.com/
http://www_3_4.gmail.com/
http://www_3_5.gmail.com/
http://www_3_6.gmail.com/
http://www_3_7.gmail.com/
http://www_3_8.gmail.com/
http://www_3_9.gmail.com/
http://www_3_10.gmail.com/
http://www_3_11.gmail.com/

Posted by: karel at October 13, 2006 7:34 PM

ionolsen22 May we exchange links with your site?http://www_3_2.gmail.com/
http://www_3_3.gmail.com/
http://www_3_4.gmail.com/
http://www_3_5.gmail.com/
http://www_3_6.gmail.com/
http://www_3_7.gmail.com/
http://www_3_8.gmail.com/
http://www_3_9.gmail.com/
http://www_3_10.gmail.com/
http://www_3_11.gmail.com/

Posted by: topicstarter at October 19, 2006 2:52 AM

Hello. Online information about insurance.

Posted by: protective life insurance at October 23, 2006 1:59 AM

Hello! Online miridia this is blog nice

Posted by: cheap meridia at October 24, 2006 11:55 AM

insurance cars insurance cars

Posted by: insurance cars at November 1, 2006 3:22 PM

In fact your model pays for the transiton cost by taxing the greater private account returns. This can be done by either directly taxing the returns, or indirectly by reducing the benefits. Economically it is exactly the same, in both cases the total return is reduced to pay for transition costs.

Posted by: Pozycjonowanie at November 1, 2006 3:37 PM

yea i agree

Posted by: Day Trading Tutorial at November 3, 2006 2:01 AM

cheap health insurance cheap health insurance

Posted by: cheap health insurance at November 3, 2006 11:20 AM

hmm yea

Posted by: Open Bank Account at November 3, 2006 12:01 PM

daytrading daytrading

Posted by: daytrading at November 5, 2006 10:17 AM

infermiera

Posted by: cuttier infermiera at December 7, 2006 1:35 PM

Electronic Goods for sales

Posted by: Electronic Goods at December 11, 2006 4:18 AM

Air Travel Tickets

Posted by: Air Travel Tickets at December 13, 2006 1:01 PM

your site design looks cool !!

Posted by: diamond rings buyout at December 24, 2006 3:25 AM

Huh...don't know what to say...i've had that feeling for about last 2 months, and deñided to put a cam in a bedroom. Shit, my hands are shaking...that dirty slut sucked my cock just twice a year, and look what she is doing with his cock:
blowjob video - çäåñü äîð
and she also gives in ass:
[url=http://www.wsboards.com/?mforum=shemalesex]free shemale sex thumb[/url]
Well, should i catch them and kill them, or invite him and show them tape?
or just go to court?
fuck, donno what to do...need your advices....

Posted by: DoctorBen at February 6, 2007 5:58 PM


I really enjoyed this page. I will be linking and I will be trying to read and research all that there is to offer from this site! Would you please also visit my site?
[url=][/url]

Posted by: Steve at February 9, 2007 7:29 AM

Basically nothing noteworthy happening right now, but eh. Today was a complete loss. I haven't been up to much recently. I've pretty much been doing nothing worth mentioning.

Posted by: Sten59645 at February 12, 2007 1:24 AM

Fantastic article. I like this. Keep up the good work. Greetings

Posted by: Aukcje at February 12, 2007 6:10 AM

cam sex cam sex

Posted by: cam sex at February 12, 2007 9:08 AM