Arnold Kling  

Adequacy of Personal Savings

PRINT
Friedman on the Battle of Idea... Dollars and Deficits...

In a research paper, John Karl Scholz says that over 80 percent of households are saving at least as much as is optimal. However, much of their result is based on counting Social Security wealth in their analysis.


Private net worth significantly exceeds the value of social security only in the top two deciles of the lifetime income distribution. The metaphor of the “three-legged stool,” in which retirement income security is supported by the three legs of social security, employer-provided pensions, and private wealth accumulation, appears to apply only to households in the top 70 percent of the lifetime income distribution because low-income workers lack employer-provided
pension coverage.

I found the following statement difficult to fathom:

Optimal wealth targets are $69,777 for the median household and are $253,631 for the median household in the highest decile of the lifetime income distribution.

Those figures represent the simulation of a life-cycle consumption model for wealth accumulation in addition to Social Security and pension wealth. The figures strike me as ludicrously low. Perhaps part of the answer is that "the mean age of households in our sample is 55.7, so the average household will work many additional years before retiring."

One exercise I wish that the author had conducted is to look exclusively at households aged 60 and over. It strikes me that in the simulation model, the bar for optimal saving for someone below age 40 may be so low that just about every household steps over it. It's probably the case that only above, say, 50 years of age, are the model's predictions of adequate saving high enough to be interesting.

If 100 percent of households under the age of 56 have met the model's modest targets for wealth accumulation, while only 60 percent of households over the age of 56 have met their targets, that would say suggest a rather different interpretation of the conclusion that 80 percent of all households have met the targets.

UPDATE: For a discussion the Scholz study, go here and click on the event materials in the upper right of the page.

For Discussion. Scholz' data are from 1992. How would changes in asset prices since then affect optimal saving?


Comments and Sharing


CATEGORIES: Social Security



COMMENTS (18 to date)
spencer writes:

The Fed just published a study where they attribute virtually all of the drop in savings over the the last quarter century to the bull market in stocks reducing investors need to save.

Boonton writes:

I know Arnold likes to hark on the idea that American's are not saving as much as they should but he doesn't seem to often address why this should be. After all, the market more or less sets the return on savings & the returns are only all the more augmented by super favorable tax policy. Favorable not only in the obvious sense of 401K's and various types of IRAs but also favorable treatment of capital gains income as compared to wage income.

If you believe the market works & is usually smarter than its participants what makes you think your feelings about optimal savings are more valid than the actual market result?

But while we are on the subject, real estate is probably the largest asset that many people own and it has been going through the roof recently. Many Americans own stocks but often indirectly and often in very small amounts except for those near the top of the income bracket, but many Americans own real estate. How much appreciation in real estate is causing people to offset their savings?

Mark Horn writes:

Quoting Boonton:

If you believe the market works & is usually smarter than its participants what makes you think your feelings about optimal savings are more valid than the actual market result?

Boonton, I don't mean to speak for Arnold, but I want to respond on my understanding of what Arnold has written in the past on the subject of savings. A market, all by itself, isn't the good thing. A free market is the good thing. I think Arnold's position is that the market for savings isn't free. The government keeps meddling with it through the style of taxation that we have. That style of taxation causes the market to react. What I think Arnold is complaining about is the style of taxation - the cause of the market's reactions, not the market reaction.

The government taxes production (income, interest, capital gains) instead of taxing consumption. The market's reaction is to favor consumption over production. The relative comparison of tax between different forms of production (say income vs capital gains) is not nearly as important as the comparison of tax between production and consumption. The end result of this is that the non-free market for savings does not result in the most efficient savings mechanisms. This isn't the market's fault. It's the fault of those who interfere in the market. The fix isn't to complain about the "failure of the market". It's to complain about those who are meddling with the market.

But I'm an econ newbie. I could be wrong and I don't want to try and represent Arnold here. This is just the way that I interpret what he's written. Constructive criticism of my interpretation is encouraged.

Boonton writes:
comparison of tax between production and consumption. The end result of this is that the non-free market for savings does not result in the most efficient savings mechanisms. This isn't the market's fault. It's the fault of those who interfere in the market. The fix isn't to complain about the "failure of the market". It's to complain about those who are meddling with the market. But I'm an econ newbie. I could be wrong and I don't want to try and represent Arnold here. This is just the way that I interpret what he's written. Constructive criticism of

That's a good answer Mark but I think there's still several problems with it. First the 'not really a free market argument' sounds a bit like the old fashioned arguments over communism. Defenders of communism said you couldn't judge it by looking at Stalin or Cuba because they didn't have 'true communism'. There will never be a truely free market anymore than there will ever be a true Ecludian circle or perfectly straight line.

Second, the difference between the tax system today and the system in place in 1980, 1970, and 1960 is that today there is MORE incentive to save! If you make your income from saving (i.e. capital gains) it is taxed at a lower rate than income from working. In fact, capital gains are not taxed at all until they are realized. In the past people didn't have the ability to shelter their saving in 401K's, IRA's etc. In the 80's and before credit card interest was deductible, today it is not (that was a subsidy to consumption and debt that no longer exists!).

So maybe there isn't a perfectly free market in regards to the choice between consumption and saving but it's a lot less tilted against savings than it used to be. How come people are bemoaning that we aren't saving as much as we used to? If anything there should have been a crises of low savings 20, 30, 40 years ago.

Finally you should consider that rewards for savings produced a mixed result. If I'm saving to get a return then yes lower taxes, better returns etc. will motivate me to increase my saving. However, if I'm saving towards a specific goal (like having $400,000K in my 401K at 65 yrs old or to have $10K for a down payment on a house in three years) then improved returns actually will cause me to cut back on savings! Why? Because I can accomplish my goal by setting aside less money.

Lawrance George Lux writes:

I am presenting a Posting of Retirement savings at my blog sometime today, based on a current Study etc. at

My basic thesis is that Economic modeling does not reflect modern mechanisms for Savings, and bring zero or low ratios of Savings when actual ratios are much higher. lgl

Arnold Kling writes:

There are two factors that I believe would cause under-saving.

1. We now live longer and use health care much more intensively than in the past, and so behavior that made sense and presumably was ingrained historically is now inappropriate. I think that people behave rationally when they have plenty of opportunities to learn from their mistakes. With lifetime savings decisions, you can't learn from experience so much.

2. Social Security and Medicare take away a lot of the means and incentive to save. A lot of "middle-class squeeze" is simply the huge chunk of the median paycheck that goes to fund SS and Medicare. And we tell people that they don't have to save for their old age--that government will take care of them.

Boonton writes:

Arnold,

I can agree with #1. But if Social Securities return is really negative & many doubt it will be there 20, 30 years from now then why wouldn't that result in an increase in savings? If SS drives out savings then shouldn't we have seen less saving in the 50's-70's when SS was paying superior market returns and more people had employer provided pensions?

How about the following proposal as a comprimise that should be acceptable to any serious advocate of privitization:

Have a 2% additional payroll tax that will be diverted directly to a IRA of the person's choosing. You can set up an IRA easily at any bank. If SS's problems are not solved and in 2030 we start cutting benefits people will have their IRA balances to fall back on. IF SS's problems do not materialize then people will have both SS benefits and their IRA balance.

This would directly increase the savings rate in any case.

Dewey Munson writes:

Arnold - Each of following is in fact a question.

Saving is the holding of Money received for current Production.

Saving therefore is refusal to Consume that which was Produced by the Consumer of that which you Produced.

Therefore the Production/Consumption process wiil cease

So for the Production/Consumption process to continue the Consumption shortage must be replaced.

Retirement consumption of previously saved Production should equal the current Consumption deferral.

(Skip the complexity of debt etc for the moment)

Now Time becomes a problem and with it the stability of the monetary system in its function as a store of value.

The fiat monetary system(s)of printed Money are in trouble so that the $0.05 I saved from my first job in 1939 to buy a loaf of bread is far short of the $1.40 I will need tomorrow.

Mark Horn writes:
So maybe there isn't a perfectly free market in regards to the choice between consumption and saving but it's a lot less tilted against savings than it used to be. How come people are bemoaning that we aren't saving as much as we used to?
I don't know. Maybe we haven't reached the tipping point yet. Maybe the changes that have occurred take a lot longer to show up in society than we know?
If anything there should have been a crises of low savings 20, 30, 40 years ago.
Unless, of course, 20-40 years ago was too soon to see the consequences of the "New Deal", and we're only now starting to feel those consequences. Maybe it'll take that long before the more recently enacted changes make there way into society. But I don't know. Just guessing.
Mark Horn writes:
Therefore the Production/Consumption process wiil cease
I think that the production/consumption process will go slower than it would without savings, but in times of non-production (job loss or retirement) the production/consumption process will go faster than it would without savings. In other words with savings, the production/consumption process doesn't cease, it evens out. Lower peaks, but also shallower valleys. IMHO, that's a good thing.

Or have I made a critical mistake that I don't see?

Boonton writes:
Unless, of course, 20-40 years ago was too soon to see the consequences of the "New Deal", and we're only now starting to feel those consequences. Maybe it'll take that long before the more recently enacted changes make there way into society. But I don't know. Just guessing.

Mark, I won't do it but Keynes had a very fitting quote about theories that depend too much on 'the long run' :)

In all fairness, it seems really strange to blame the lack of savings on an unfriendly tax environment. If you have a savings fetish you are living at probably the happiest moment of American history since the days when the primary form of taxation was tariffs.

Arnold makes a good point about lifecycles. People can learn lessons that take a day or month or year to learn but lessons that require 30, 40 years cannot be learned by trial and error. If you screw up once that's more or less it.

That's a good argument for paternalistic gov't intervention. The 'New Deal' (the original plus the additions from the Great Society) basically did just that. What is missing from Arnold's explanation is the demand side....those that demand savings for investments.

If ignorance is causing American's to not save as much as they should, shouldn't this result in a shortage of savings available for those who use it? Shouldn't banks & other financial institutions be rewarding savers with huge returns in order to pry American's loose from their ignorance?

Mark Horn writes:
Arnold makes a good point about lifecycles. People can learn lessons that take a day or month or year to learn but lessons that require 30, 40 years cannot be learned by trial and error. If you screw up once that's more or less it.

That's a good argument for paternalistic gov't intervention.


Ok. So summing up, it sounds like you're saying the following: most people don't save because they can't learn the value of savings. They can't learn this value, because it takes too long to experience the negative consequence that comes from making a mistake. And by the time they learn, it's too late to do anything about it. This is why social security and medicare (et al) are good things. Because the government (which presumably can learn from these types of mistakes) can react to them and proactively respond, where as individuals can't (for the most part).

Did I get that right? If so, then how would you address the inefficiency that comes from just about every government operation? What I like about markets is that I don't have to wait for 2, 4, or 6 years before I vote out someone who's underperforming. I simply switch vendors - whenever I want. This results in efficiency.

Meanwhile, I've been contributing to social security for my entire working life, and it's only recently that we've had any serious discussions of allowing me to opt out from the current vendor. The potential for inefficiency is staggaring. And when I examine my SSA annual benefits summary, I find the returns from that particular vendor to be very poor.

So if I've got it right there's a tradeoff. What we get is the ability to ensure that those who didn't save aren't completely screwed when they realize their mistake. And for that we give up an efficient market for savings through forced savings from a single (presumably inefficient) vendor. In other words, we prevent anyone from making the savings mistake, and we give up efficient markets for savings for those who would save without being forced. So the people who make mistakes are subsidized by those who don't.

I'm not ready to claim that this is necessarily bad - it may just be the way we do things. I subsidize my children, and I don't consider that bad. I just want to see if I understand correctly.

If ignorance is causing American's to not save as much as they should, shouldn't this result in a shortage of savings available for those who use it? Shouldn't banks & other financial institutions be rewarding savers with huge returns in order to pry American's loose from their ignorance?

When I look at TV, I see ads for Janus, Schwab, Vanguard, T. Rowe Price, Bank of America, Wachovia, etc. They aren't promising returns, but they're certainly out there and they're advertising. There seems to be quite a bit of competition for getting the dollars of the savers.

That being said, I wonder how much bigger the market for these types of services would be if there wasn't a single vendor with a forced savings program (i.e. Social Security). The market certainly seems to have grown. And from my own (non-scientific) observation, that growth seems to be correlated with the "friendlier" tax environment. Would the market be bigger if the tax system were even friendlier? Would it operate more efficiently than the SSA? Would it provide better returns?

Would the increased efficiency and better returns offset the negative from people getting screwed who didn't save?

Boonton writes:
Meanwhile, I've been contributing to social security for my entire working life, and it's only recently that we've had any serious discussions of allowing me to opt out from the current vendor. The potential for inefficiency is staggaring. And when I examine my SSA annual benefits summary, I find the returns from that particular vendor to be very poor.

In terms of efficiency social security is pretty good. SS is a system that taxes some people (workers) & gives benefits to others (retired people). The only real objective measure of this efficiency is how much money gets lost inbetween. This amount is less than a penny on the dollar. Not surprising because economies of scale make it easier to lower transaction costs. A computer system that can track accounts and cut checks for 20 million can probably do so for 200 million with much less than ten times the cost.

Adding multiple 'vendors' will increase transaction costs and the vendors will compete with each other to capture as many of those costs as they can, in the long run they may indeed come down. But transaction costs add zero economic value, why should we aspire to have more of them even if they are delievered cheaply?

So if I've got it right there's a tradeoff. What we get is the ability to ensure that those who didn't save aren't completely screwed when they realize their mistake. And for that we give up an efficient market for savings through forced savings from a single (presumably inefficient) vendor. In other words, we prevent anyone from making the savings mistake, and we give up efficient markets for savings for those who would save without being forced. So the people who make mistakes are subsidized by those who don't.

I'm not really sure that social security has anything to do with the market for savings directly. It is true that on one hand SS may discourage some savings because some people will be content with just their SS check when they get older (or think they will be content). On the other hand I suspect that social security has added a measure of stability to our economic system, letting people take healthy risks because they know there's that safety net if things go really bad. Recent research has shown that people tend to be a bit irrational about risk. They fear flying more than they do driving even though the odds of being in a plane crash are much lower. If people over-compensate for 'irrational' worst case scenero fears then I suspect a program that allieviates such fears might be quite beneficial for the economy. Saving is a two way street. In order for your saving to generate economic growth there has to be someone on the other end who is willing to borrow your money and take a shot at some economic endeavor. What many seem to ignore (probably because it is hard to measure) is that the incentive to borrow is as important as the incentive to save.

Captain Arbyte writes:

Boonton,

if Social Securities return is really negative & many doubt it will be there 20, 30 years from now then why wouldn't that result in an increase in savings? If SS drives out savings then shouldn't we have seen less saving in the 50's-70's when SS was paying superior market returns and more people had employer provided pensions?

Speaking for myself only, it has increased savings. I'm in my 20s and I'm convinced I won't ever see one dime of my SS taxes back. I'm planning my finances without any benefit from SS.

The impact to the incentive to save varies with age. Those presently near retirement have a disincentive to save because they know their SS benefits aren't in jeopardy.

However, I wouldn't know how to spot this influence in a data set because young people tend to have less available to save in the first place. Particularly after payroll taxes!

mcwop writes:

Who cares what SS costs to administer compared to private accounts? The admin cost could be zero, and it is still be a bad deal for a lot of SS contributors. The returns for the vast majority of recipients born after 1964 will be negative or low (less than 2%). If you raise payroll taxes or the retirement age, then those returns get even worse. This is the fundamental complaint with the system as it stands.

The "cost more to administer" sound bite is outright FUD.

The federal thrift savings plan is every bit as cheap to administer as the SS system. I work with large 401k’s, and these plans are every bit as cheap to administer as the SS system. Vanguard can run an IRA just as cheaply.

The "three legged stool" model is bad, in America.

What is missing is the HUGE housing cost component, and house-buying consumption/saving. Most middle class folk invest in a series of houses -- more expensive as they reach their income peaks; and they "save" the equity build-up. In fact, if they do have any non-equity based savings, comparing their after-tax savings returns to after-tax investment returns of buying a bigger house leads most to go for a bigger house.

Skimming the start of the 64 p paper:
Households in the model form realistic
expectations about social security, which depends on lifetime income; pension benefits, which
depend on income in the final year of work; and earnings.

It prolly does not adequately incorporate house-equity and the house asset as a saving component.

For daily living needs, SS income is enough for most modest folk.


Perhaps I learned too much from my grandparents retirement decisions: they sold their house and bought a retirement mobile-home, with the rest of their equity into savings. And they left little for inheritance after they died. Seems good to me.

++The paper seems to include reasonable housing issue handling. [This is a blog, darn it -- if I wanted to WORK I wouldn't be here.]

could be accumulated in real and financial assets, the current value of defined
contribution pensions, including 401(k)s, and housing net worth (for now, we assume households
are willing to reduce housing in retirement to maintain consumption standards).

The authors don't highlight it, but I'm comfy with the idea that house prices reflect a lot of the "savings deficit".

But, uh, this could be really bad news if there's another bout of inflation and rapid house price devaluation. Like, if the USD goes into foreign currency free fall. But it seems the Fed is already worried enough about that possibility, watching the long term T rates should be the key indicator.

Lancelot Finn writes:

If Scholz et al. are right, this is good news for the rationality assumption that characterizes the models conservative economists prefer.

Arnold's argument that the rationality assumption depends on people making and learning from mistakes, and that people are therefore unlikely to satisfy rationality conditions when it comes to saving for retirement, is plausible. But not necessarily right. It's equally plausible that rationality is more about calculation, and that people plan out their futures with certain rough and regularly updated statistics in mind, with a dose of pessimism. People may make rough calculations of how much they need to maintain their standard of living after retirement. In that case, we should expect people to save adequately, and Scholz's result is welcome.

If some people are oversaving, this may show that they recognize the political risk involved in the Social Security program, and are compensating for it. In a recent essay, "Social Security - Political Risk = Private Accounts," I argued that the political risk in the current program is worse than the market risk that would obtain under private accounts.

Comments for this entry have been closed
Return to top