Comments on: WHy does the government measure savings as it does?

Question: WHy does the government measure savings as it does?

IndigoJones — Tue, 10 Apr 2007 17:34:23 -0800

The US government does not consider IRA money or 401K money (or stock market investment of any kind) to count as "savings" when they calculate US savings rates. Thus our perennially lousy savings rates vs. other countries. Or so I read. If true, what is our government's reasoning?

One argument I've heard re 401s at least is that as they are not really dynamic or choice ridden (automatic withholding doing the job of will power on the biweekly pay packet), they therefore do not show as a clear a picture of civilian economic sentiment on a day by day basis. Not sure I believe that, or that I think it's useful. What do you think?

Bonus question- do other high savers (Germany, Japan, Scotland) have this kind of automatic savings and if so, how is it counted by their (or our) economists?

By: mhum

mhum — Tue, 10 Apr 2007 18:21:25 -0800

The US government does not consider IRA money or 401K money (or stock market investment of any kind) to count as "savings" when they calculate US savings rates

What is your basis for this belief?

According to the BEA (the relevant gov't agency), personal saving is defined as "personal income less the sum of personal outlays and personal current taxes". No mention of dynamism or choice. In fact, on page 9 of this paper outlining their definitions and methodology, 401k contributions are explicitly counted in wage and salary accruals (i.e.: a part of personal income).

By: b1tr0t

b1tr0t — Tue, 10 Apr 2007 18:27:28 -0800

I don't know how the numbers are computed, but many 401(k) accounts are invested in stocks or mutual funds. That is, they are investment accounts, not savings accounts.

By: smackfu

smackfu — Tue, 10 Apr 2007 18:33:41 -0800

A good reason not to include it is because these retirement plans replace pension plans, which were never historically included. So savings rates would go up but it would be meaningless.

mhum, I think the issue is that personal income in the definition starts with "compensation, received", and 401k is pre-tax so is paid to employee but not received and so not counted.

By: mhum

mhum — Tue, 10 Apr 2007 19:21:47 -0800

smackfu: I think the issue is that personal income in the definition starts with "compensation, received"

Except, if you follow the link to "Compensation of employees (received)" it leads you to "Wage and salary accruals and disbursements" which includes "voluntary employee contributions to certain deferred compensation plans, such as 401(k) plans".

I remain skeptical of the proposition that 401(k) plans are not included in personal income.

By: mhum

mhum — Tue, 10 Apr 2007 19:38:15 -0800

And, in case anyone wants to argue that 401(k)s are counted as a personal consumption expenditure because they're an investment (and hence not part of personal savings), please take a look at this document (particularly tables 2.4 and 2.5 in the Appendix) or at this table detailing the breakdown of the products included in PCE. It boils down to durable goods (e.g.: cars, furniture, etc...), nondurable goods (e.g.: food, clothing, fuel, etc...), and services (e.g.: housing, medical care, etc...). Note that they explicitly include brokerage charges (under services), but nowhere do they include the, say, stocks or mutual funds you bought which incurred the brokerage charges.

By: sfenders

sfenders — Tue, 10 Apr 2007 19:53:14 -0800

The US government does not consider IRA money or 401K money (or stock market investment of any kind) to count as "savings"

Well, money put in any kind of brokerage account (or whatever an IRA is) probably should be counted as savings. But stock market investment is not saving money. You didn't save it, you spent it on buying stocks. Owning some investment instrument is really not the same thing at all as having money in the bank. It will be counted as income/savings when you sell it for a profit, so it gets counted then.

I'm pretty sure that counting stock market investment as "saving" is not standard anywhere in the rest of the world either. You might as well count everything of value and just measure "net worth" if that's what you want to do.

By: russilwvong

russilwvong — Tue, 10 Apr 2007 20:03:31 -0800

IndigoJones: The US government does not consider IRA money or 401K money (or stock market investment of any kind) to count as "savings" when they calculate US savings rates. Thus our perennially lousy savings rates vs. other countries. Or so I read.

This is false. You might want to track back where you read it.

sfenders: But stock market investment is not saving money. You didn't save it, you spent it on buying stocks.

Flagged. This is also false. mhum is correct: household savings = household income minus consumption. Buying an investment is like moving money from one pocket to another pocket; it's not the same as buying a TV.

Regarding comparison of household savings rates: this OECD report discusses the difference between household savings rates in the US (2.4% in 2002), the EU (9.6% in 2002), and Japan. After accounting for three possible factors (different levels of public expenditure, reliance on different taxes, differences in pensions), the difference is magnified, not reduced.

By: smackfu

smackfu — Tue, 10 Apr 2007 20:27:55 -0800

Do mortgage payments count as savings then? It's an investment for most people.

By: sfenders

sfenders — Tue, 10 Apr 2007 20:35:22 -0800

Yes, it seems I am more insane than usual tonight, sorry. Here is a nice document that explains it all. It seems capital gains are entirely excluded. I cannot guess why.

By: mhum

mhum — Tue, 10 Apr 2007 21:52:57 -0800

smackfu: Do mortgage payments count as savings then? It's an investment for most people.

If I'm interpreting the section on Owner-occupied Housing on page 8 of this document correctly, the interest and property tax parts of mortgage payments are counted as expenditures. The rest, presumably, count towards savings.

By: Gerard Sorme

Gerard Sorme — Tue, 10 Apr 2007 22:02:21 -0800

The OP (indigojones) is absolutely right. Here's some recent discussion on the topic and why our rates are skewed:
http://blogs.dailymail.com/donsurber/2007/02/02/the-myth-of-a-low-savings-rate/

By: mhum

mhum — Tue, 10 Apr 2007 22:04:29 -0800

sfenders: I'm pretty sure that counting stock market investment as "saving" is not standard anywhere in the rest of the world either.

For what it's worth, it seems like the Canadian definition of savings is in line with the American definition, at least with respect to stock market investment. According to this document, "capital transfers have no effect on the saving of the donor or recipient". I'm fairly certain that purchasing stock in a company counts as a capital transfer.

By: mhum

mhum — Tue, 10 Apr 2007 22:16:28 -0800

Gerard Sorme: Here's some recent discussion on the topic and why our rates are skewed

The discussion linked to makes a fairly serious (but apparently quite common) error in interpreting the BEA's defintion of savings. The definition as presented in the Crutsinger article is:

The savings rate is computed by taking the amount of personal income left after taxes are paid, an amount known as disposable income, and subtracting the amount of spending.

This is interpreted by some to mean that only "taxable" income is included as income and since 401(k) contributions aren't counted as "taxable" they are excluded. This is false. The BEA's methodology documents and glossary (both of which I've linked to above) make this patently clear. Both 401(k) contributions and pensions are counted as income, the former under "Wage and salary accruals and disbursements" and the latter under "Supplements to wages and salaries".

By: commander_cool

commander_cool — Wed, 11 Apr 2007 05:14:24 -0800

Do the government economic statistics include equity investments in housing? I'm spending 15% of my pre-tax income building equity in my condo and don't personally consider that an outlay because it's reducing my debt. Americans are more likely to be homeowners than other citizens of the world, so that may explain some of the difference in savings rates.

By: IndigoJones

IndigoJones — Wed, 11 Apr 2007 05:45:57 -0800

Do the government economic statistics include equity investments in housing?

It is my understanding that they do not.

The source for my original proposition is most recently the current issue of Forbes, but I have read it elsewhere many times. See also Gerard Sorme above.

Many thanks to all who have answered, and to whose who may yet still.

By: IndigoJones

IndigoJones — Wed, 11 Apr 2007 05:54:49 -0800

Owning some investment instrument is really not the same thing at all as having money in the bank.

But if the bank lends it to business, and they do, then the money is tied to business performance much as if you have bought stock, no? Would the feds consider my buying corporate bonds as saving?

By: mhum

mhum — Wed, 11 Apr 2007 08:28:33 -0800

commander_cool: Do the government economic statistics include equity investments in housing?

As I linked to above, the owner-occupied housing section of personal consumption expenditure appears to include only interest and taxes. So, if my interpretation is correct, the equity you put into your house would be counted as saving.

By: mhum

mhum — Wed, 11 Apr 2007 08:51:39 -0800

IndigoJones: The source for my original proposition is most recently the current issue of Forbes, but I have read it elsewhere many times. See also Gerard Sorme above.

Unless the Forbes article is referring to an alternative measure of personal saving (which would be strange, I think like talking about alternative measures of GDP without explicitly mentioning the fact), there really shouldn't be any dispute. It's all spelled out by the BEA -- the part of the Dept. of Commerce which tracks the National Income and Product Accounts and hence computes the personal saving rate -- in its own glossary:

Personal saving = personal income less the sum of personal outlays and personal current taxes

Personal income = the sum of compensation of employees (received), supplements to wages and salaries, [and other stuff...]

Compensation of employee (received) = wage and salary disbursements and supplements to wages and salaries received by U.S. residents, including wages and salaries received from the rest of the world. (Note: this is last bit is what distinguishes Compensation (received) from Compensation (paid) -- the latter excludes non-US payments)

Wage and salary accruals and disbursements = The monetary remuneration of employees, including [... among other stuff ...] voluntary employee contributions to certain deferred compensation plans, such as 401(k) plans

Supplements to wages and salaries = employer contributions for employee pension and insurance funds and employer contributions for government social insurance.

I'll see if I can find a copy of Forbes today, but this seems pretty clear-cut to me. This notion that "401(k)s aren't counted as savings by the government" seems to be some kind of urban legend about gov't statistics. Kind of like how people persist in believing that they measure the unemployment rate by counting the people receiving unemployment benefits. They don't, they use a phone survey.

By: mhum

mhum — Wed, 11 Apr 2007 09:00:58 -0800

IndigoJones: Would the feds consider my buying corporate bonds as saving?

I'm fairly certain (from reading this) that the purchase of the bond would be counted as saving (or, more precisely, not counted as an expenditure). The interest you receive on the bond would be income for you and an expenditure for the corporation.

By: mhum

mhum — Wed, 11 Apr 2007 09:29:32 -0800

smackfu: A good reason not to include it is because these retirement plans replace pension plans, which were never historically included.

According to this page of this senate report, pensions were included in national income from the very first analyses in 1934. So, this isn't even a case of something which used to be excluded but only recently included.

By: smackfu

smackfu — Wed, 11 Apr 2007 10:04:40 -0800

Although it's not specific, I assume they are referring to pension disbursements there. Those should count as income by any standard.

By "pension plans", I was referring to the pension fund contributions by the company, which have been replaced by 401k matching. Those are difficult to count as income.

By: russilwvong

russilwvong — Wed, 11 Apr 2007 10:31:02 -0800

Just wanted to thank mhum for trying to straighten out people's misconceptions on this thread, and especially for backing up his or her answers with links to relevant definitions from the Bureau of Economic Analysis and elsewhere.

By: mhum

mhum — Wed, 11 Apr 2007 10:43:50 -0800

smackfu: Although it's not specific, I assume they are referring to pension disbursements there

Indeed, I believe that you are correct. Pensions (as labor income) appear to be lumped in with disability compensation which would suggest that it's the disbursments which are counted as personal income.

However, this raises a further point. Since pension fund contributions by the company were not counted as income to the employee, they would not (as far as I can tell) have been counted as an expense to the company and hence would have been counted as a saving for the company. It appears that, in 1934, they were concerned more about the total national aggregates rather than the current personal/business/gov't breakdown (understably so, since this was their first crack at the problem). Total national savings (computed in the usual way as income minus expenses) would be the same whether the pension fund contributions were attributed to the company or the employee.

By: mhum

mhum — Wed, 11 Apr 2007 11:24:47 -0800

IndigoJones: The source for my original proposition is most recently the current issue of Forbes

I found the article: "Don't Worry" by Ken Fisher on page 180 of the Apr. 16, 2007 issue of Forbes. In that article he asserts that

[a]ny American with net worth tied up in a home's value, a 401(k) or stocks. [...] In [my book] that's a form of savings. The government, though, counts only how fast you pay down your home's mortgage.

This is wrong on 2 out of 3 counts. It is true that the official savings figures don't count capital gains (you can see the explanation for their reasoning in this FAQ). However, it is not true that retirement plan contributions aren't counted. I hope my previous posts are sufficiently persuasive because I'm not sure what else would be more convincing.

As for housing value, he is correct that the gov't counts how fast you pay down your mortgage. However, they also explicitly count housing appreciation as a form of income, specifically as Rental income of persons with capital consumption adjustment. Don't let the term "rental" throw you. For owner-occupied housing, they impute a rental value. See also this FAQ.