Wages and Salaries as a Share of Net Domestic Product

Since 1970:

These two figures tell different stories. The difference between them is that non-wage non-salary benefits--overwhelmingly health and pension--are included in the second and not in the first.

From the perspective of workers who are not part of the salaried part of the upper middle class--for whom benefits are, we think, a large part of compensation--the first diagram is more relevant. It shows (a) the explosion of the wage-and-salary share during the late 1960s, (b) the rollback during the 1970s and the further downward push of the wage share in the early 1980s, (c) constancy up until the late 1990s, (d) a wage-and-salary share boom during the "new economy" boom, and (e) the collapse of the wage-and-salary share since 2000.

From the perspective of the salaried part of the upper middle class (or of the declining fraction of other workers with ample benefits) the second diagram is more relevant. It shows (a) the explosion in the labor share of income in the late 1960s, (b) continued rise through the 1970s, (c) a rollback during the early 1980s, (d) a decline in the labor share and then a recovery in the Clinton years, and (e) nothing terribly unusual for the state of the business cycle going on since 2000.

I find this frustrating. The national income accounting tells us that what is going on with benefits and their distribution has a profound effect on how we understand the recent economic history of the distribution of income between labor and capital, and I don't think I have a good grasp of what is going on with benefits and their distribution.

Comments

[Barry]

Brad, remember that when you look at these and extrapolate to people, you're taking an average. If an oeverall % remains constant, but you know that the upper decile of the population has been experiencing an increase, then you should realize that that implies that it's getting worse for some of the other deciles.

[masaccio]

It looks like those upward slopes come on Carter's and Clinton's watches, and the downslopes come on the watches of Reagan, and Bush and Bush. Hmmmm.

[me]

masaccio - I noticed that right off, good read.

I am also wondering about those "benefit" numbers. Who is getting them. Salaried workers have lost their pensions in favor of cash balance plans or 401Ks that greatly reduce a company's cost.

On the healthcare side, every year employees pay more money for less coverage. If you read the financial statements of big companies like IBM, they are paying less in employee healthcare today than they were 10 years ago.

[CapitalistImperialistPig]

Is this data based on an exhasutive survey, or is it just another one of those phony statistical arguments? ;)

[Movie Guy]

"I don't think I have a good grasp of what is going on with benefits and their distribution."

I believe that general trend in benefit packages is becoming reasonably clear. They are being tightened by many companies and corporations.

Apparently, the push is on to shift the increasing share of healthcare coverage costs to employees. Sure, the companies and corporations provide the vehicle (the plan) for coverage, but the employees are having to assume a larger share of the plan costs as corporations bargain down the employee negotiation positions if such negotiations occur at all. I see no reversal of this effort. If anything, the intent to float all healthcare costs to the employees is the latest tactic, though it may not be widespread across all industries. Yet.

There are benefit cuts being experienced by some individuals in the salary/bonus range of $150,00 to $350,000 as well. Primarily healthcare cost transfers to employees from what I have heard or read.

Lower salaried workers are certainly taking it on the chin. In some manufacturing operations, a common practice is to manage employee shift hours so well that many are ineligible for healthcare benefit coverage. It's a weekly drill. Elsewhere, we're seeing the tradeoff between wage increases and continuation of healthcare plans and maintenance of pension contribution percentages by employers.

I expect that a deeper examination will also reveal that companies and corporations are making progress in stepping out of large positions in all pension plans, whether DB or contributory.

I have no doubt that within ten to fifteen years, it will be common practice that the majority of companies and corporations will provide no significant contributions toward sponsored healthcare plans or pension funds. The Ownership Society...

Sounds like a good time to conduct a 1,000 sample poll.

[anne]

ME

'If you read the financial statements of big companies like IBM, they are paying less in employee healthcare today than they were 10 years ago.'

This comment surprises me quite a bit, and I will look to the data to understand. Thank you.

[anne]

Brad DeLong

'The national income accounting tells us that what is going on with benefits and their distribution has a profound effect on how we understand the recent economic history of the distribution of income between labor and capital, and I don't think I have a good grasp of what is going on with benefits and their distribution.'

Agreed.

[andres]

Brad, it is only a small step from this analysis to argue that the continuing concentration of income, the decline in the wage and salary share in particular, jeopardizes the macroeconomic well-being of the country and makes it harder for the Fed to maintain high employment rates. If you're not careful, post-Keynesians will start citing these posts, and your departmental colleagues will start giving you strange looks.

[piotr]

Employee compensation - wages increased in 4 years from 11% to 13% of NDI.

Costs of medical care increased by ca. 40%.

I surmise that (a) almost half of the increase of the medical cost was shifted from employer costs to employee costs, and about half increased the employer costs. If so, than employee compensation - employee medical costs share of NDI decreased quite precipitously.

What is very curious is what parts of NDI increased so dramatically -- statistic discrepancy? By the way, are those pre-tax of after-tax figures?

[pi]

I think the standard reference on the distribution of benefits receipts and the impact on trends in inequlity is Brooks Pierce's paper in the the Quarterly Journal of Economics (Vol. 116, No. 4, November 2001, pp. 1493-1525). As I recall, the conclusion is that the increase in wage rate inequality looks larger when one includes the value of benefits.

[ken melvin]

Thanks Prof. DeLong. I've long wanted to see the these ratios. I think they tell much a story and I really appreciate you doing this.

[Michael Cain]

"Apparently, the push is on to shift the increasing share of healthcare coverage costs to employees."

TTBOMK, in every large developed economy except the US, health insurance costs are not born by private companies. In an increasingly globalized world economy, it seems unsurprising then that US companies would seek to shed employee/retiree health insurance obligations as quickly as possible simply from a competitive perspective.

Again TTBOMK, in every large developed economy except the US, health insurance for common sorts of care (eg, you break a leg or develop cancer) is provided in a collective fashion. The details vary a lot, from Britain's NHS to Japan's system of private but very heavily regulated insurers and care providers. The funding details also vary widely, but in general the amount paid by an individual looks a lot like a percentage of income. In such a system, the poor are not paying their full bill, and the rich are paying more than theirs -- if your income is $10M, and the percentage is 4%, you probably didn't get $400K worth of medical care.

The US, or at least the present administration, seems to be determined that rather than "we" being responsible for health insurance collectively, "I" will be responsible for my own health insurance individually. I suspect that in the medium-term (say, 10-15 years), this will turn out badly. But it will take a near-catastrophic failure of the system in order to change the outlook.

[simon]

Brad two qusetions -

1)How does one make inferences about income deciles causally? Can the difference between the two sets of data reflect that companies have poured more money into their pension plans? That is are corporations paying for pension obligations for workers that they incurred in prior pay periods?

2) Why are we speaking about the difference between upper and income households here? Where is the data to support this inference? What exactly are we trying to say with the inference?

[Maynard Handley]

Brad, what you and and every poster have not commented on is that things are not what they seem with the second graph either. The "extra compensation" of that graph is not real money, it is debased money. Health dollars, in the sense that they can be spent only on health care, and only in a ridiculously inefficient fashion, are not as good as real dollars. Likewise pension dollars, given what we all know of PBGC finances and of corporate welshing in the last few years, are something of a crapshoot, good for maybe 70c on the dollar.

Measured in terms of real value, I believe the second curve has also gone down quite a bit.

[Bernard Yomtov]

I take it these are pre-tax numbers. If so, the gap is bigger than it looks, since the benefits are tax-free or tax-deferred.

[Alex Tabarrok]

Isn't the first graph the same puzzle as the declining labor share of the population? In other words the decline is in the number of workers not the wage per worker. The second graph is less affected because the decline in the number of workers has been at the low end of the market.

Best

Alex

[dryfly]

Soon we'll see folks with signs reading...

"Will Work For Insurance"

[ken melvin]

"Will Work For Insurance" Many already do work for the insurance. Banks, Starbucks, . . . in many cases, the insurance is worth more than the wages. Some can't get health insurance at any price, many are quoted $18k per year or more, and lots pay $4.5k/yr per person with a $2k deductible.

[Jorma]

Do stock options fit into this? I'm thinking they might be in the compensation numbers, not the wages.

I have no gross stats on option grants to employees but the numbers must be huge. Those charts have essentially a 5 percentage point range. I would not be surpised if option grants account for more than 1% of compensation. Having a strong mitigating effect on the post 2000 plunge in wages vs comp.

I'm sure most can get where I am going with this.

[me]

anne, I think you will find 10 years ago IBM's cost was close to $500 miillion and now it it under $350 million.

You need to remember that IBM (and most other large companies) offered company paid healthcare for life. Recently, IBM retirees owe the company more for healthcare than their pension is. Benefits are now a profit center, not the devastating expense their propaganda would have you believe.

[triticale]

There are, as a matter of fact, other employee benefits in some lines of work. I do contract work in cellular infrastructure, and even as a contractor I generally get free cellular, sometimes even with unmetered long distance. On my current contract, where an out-of-town contractor was expected, I have the use of a rental car much better than my own vehicle, with subsidized gasoline. I would presume that the cost of these would show up in that second graph. By the way, I'm now earning twice what I did when I entered this field six years ago.

[bakho]

Don't the peaks in the wages and salaries graph correspond to the hikes in the Federal Minimum Wage? In the mid 80s, salaries fall (Reagan didn't believe in minimum wage) but benefits don't fall as much. It seems that Reagan increased the employer contribution to SS by a couple of percentage points.

All the top guns take their salaries as stock options or other benefits that are taxed at a lower rate.

[RP]

Hmmmm....so I get the same health care as last year or the year before, but I continually pay more contribution per paycheck, and more per co-pay, and there is a bigger number on my paycheck for the "health care" account, which causes some bean counter to conclude that my compensation has risen!

Meanwhile, it merely points out that my real wage is going down faster than anyone is willing to admit.

[cm]

triticale: Why don't you tally up the combined value you are getting out of your fringe benefits (and don't forget to consider the imputed income effect), and compare it to your hospital copay or out-of-pocket even for non-complicated procedures. And also consider how much of these "benefits" you use on behalf of your job, and deduct accordingly. BTW does the insurance that comes with the rental car, or the corporate travel policy, cover accidents on non-job related travels? You may be surprised how much of a "benefit" you are actually getting!

[cm]

Michael Cain: Whichever cost the employer "bears", unless they are churning through their capital or the investors' money, it is generally safe to assume it comes from the revenue, hence arguably money "made" for the company by its employees (including not just the line workers, but also e.g. salespeople, support staff, and executives).

So in the aggregate, every dollar my employer nominally spends on my healthcare is a dollar they don't pay me nominally in wages for my efforts, provided they run a positive cashflow and don't pile up debt.

[derek]

Brad, can we see the signal obtained by subtracting those two graphs?

[anne]

Me

"I think you will find 10 years ago IBM's cost was close to $500 million and now it is under $350 million."

Even accounting for reduction in number of employees, this appears striking, and suggests looking to closely. Thank you.

[SamChevre]

From the perspective of workers who are not part of the salaried part of the upper middle class--for whom benefits are, we think, a large part of compensation--the first diagram is more relevant.

I doubt this--I don't have a good source, but I am inclined to think that benefits are a larger portion of compensation in the $25k-$50k compensation range than above it (except for very very high earners who run into the $1 mm cap). Health benefits are expensive, and they are basically a per-employee cost--they become relatively cheaper as compensation goes up. And most large employers do supply health benefits to full-time employees. I know a lot of people earning $10-$15 an hour, or teaching, whose benefits are probably worth at least 50% as much as their wages.

[ken melvin]

Within the wages and salaries group composite, would it be safe to assume that executive salaries have gone up a lot while workers wages may have gone down quite a bit and that a plot of workers' wages only might show their share even more drastically reduced?

[David Yaseen]

The total comp - wages/salaries spread increased from about 5% in 1960 to 7% in 1970 to about 11% in 1980 to 12% in 1990 and then dropped to 11% in 2000 (stock options?). Since then, it has gone up again to around 13%.

I had been under the impression that, since more employees were unionized on the left side of the period graphed, more of them would therefore have received company pensions and health care than they did later on. Yet nonwage compensation increased even as these benefits were withdrawn. Increasing inefficiency in health care spending and increasing income inequality via bonuses go some way toward explaining this shift, but it seems that something else is clearly missing.

Have there been shifts in what businesses/corporations are able to write off as employee compensation? Have unemployment/disability insurance increased during the period?

[piotr]

We need to know the definitions, but I am aware of these benefits:

1. FICA

2. Pension contribution

3. Medical insurance

4. Life insurance

5. Workers' compensation

Managers can get large pension benefits, but they do not get FICA (not much), and the medical costs decrease as the proportion of the total compensation, for this reason I think that income inequality is not a major source of the growth of benefits. OTOH we know that medical costs increase - CPI is about 30% over the last 4 years.

[Old Ari]

Extrapolating by eye, it seems that compensation will go down to zero about 2050.

[Douglass Carmichael]

Did i miss it? Who gets the non wage non salary income? Is that interest and dividends?

I've wondered what percentage of total income in the US is from interest?

[Ian D-B]

First, non-wage, non-salary income goes to rent and interest, corporate profits, and surplus of government revenues.

Second, yes, options are included, but not when granted. They are counted as salary income as the gap between the strike price and the exercise price. Performance based options are untaxed and not counted in wage and salary numbers, but they don't account for much.

Third, medical costs going up may eat into wage and salary numbers, but that was inevitable. People demand a certain level of medical care from employers, and if they want it it's got to come from somewhere. This is the same as your real income falling because gas prices are higher. The options are to either get crappier insurance or earn less money.

Fourth, the difference between this chart when it's a percent of net dom. inc. as Brad posted, and when it's gross income, is interesting. The only measurement difference between the two is capital consumption, but this causes labor's share in 2004 to match 1997 as the lowest year since 1966.

Finally, Piketty and Saez wrote the best paper on top percentiles income distribution, using tax data going back to 1917. Bascially it says that even while labor's share is constant, much of it is going to the top 1%.

[piotr]

The difference between medical costs and gasoline is that in the latter case we deal with an imported commodity, so the price is in the first approximation an external variable.

Medical costs are pretty much determined by bussiness practices and regulations within USA. The increase in costs cannot be attributed to employees suddenly demanding more care than before nor to some objective scarcity.

[nate]

http://www.nytimes.com/2005/07/30/technology/30nocera.html

any thoughts on this? how does this figure into compensation, payrolls and benefits?

[Joerg Wenck]

First you show us in a very graphical way "what is going on with benefits" - and then you complain about not having a good grasp of it? Does that mean you don´t believe the data to be correct?

[Ian D-B]

I should have done this a long time ago, but I finally got around to plotting benefits as a percentage of total compensation. It's amazing how much the've rise, essentially uninterrupted, since the 1950's. The BEA has reasonably good data on the type of benefits that money goes to, and it looks like the biggest contributions are from contributions to social insurance (50%) and group health insurance (33%). The rest is in pensions and profit sharing (13%), worker's comp (5%), and group life insurance (1%). So much of this could be due to our absurd spending in the medical sector.

[nate]

it would be interesting to know the profile of people who get benefit compensation (eg, is it government employees or private sector employees or both? is it low, middle, upper or all incomes?)

Ian D-B: does the "group health insurance" (33%) or other benefits include retirement medical benefits?

[Ian D-B]

I'd imagine group health insurance doesn;'t include retirement insurance -- that should be with retirement benefits. thgough I'm not entirely sure. The BEA isn't very good about defining their data precisely or giving methods.

Your question about who gets benefits is the key, and I don't think there's any good data on it. Benefits aren't reported to the IRS, and companies don't necessarily provide good records of what they give employees. This is just as much of a problem, if not more so, than figuring out who gets stock options and what they're worth.

[Ian D-B]

I'd imagine group health insurance doesn;'t include retirement insurance -- that should be with retirement benefits. thgough I'm not entirely sure. The BEA isn't very good about defining their data precisely or giving methods.

Your question about who gets benefits is the key, and I don't think there's any good data on it. Benefits aren't reported to the IRS, and companies don't necessarily provide good records of what they give employees. This is just as much of a problem, if not more so, than figuring out who gets stock options and what they're worth.

[nate]

it would be very nice and interesting to see how retiree medical expenses figure into this. also, does anyone know how well-funded retiree medical is?