(Promoted by jasonrosenbaum - An important argument.)

Financing health care reform will be a prime subject of discussion among the Senate and House conferees, specifically how much the financing relies on a tax on high-cost health plans. Supporters of this tax label these “Cadillac” health plans and make the assumption that they provide comprehensive (even lavish) coverage that requires very low out-of-pocket costs from beneficiaries. However, in the dysfunctional health insurance market, high-cost does not equal high-value; and it is not comprehensiveness of coverage that is the primary predictor of who will be affected by this tax, rather it is the size of the firm they work for or the age of their co-workers. The fact that Chevy plans are about as likely to be taxed as Cadillac plans is one reason to be cautious about relying on such a tax. Bivens and Gould (2009) document this as well as other reasons to prefer the more straightforward, progressive financing in the House bill.

One claim for the Senate excise tax has recently surfaced: that health care cost increases have been a major driving force in constraining wage growth and that wages will grow more strongly by curtailing employer health costs via the excise tax. This claim boldly asserts that health care costs are large enough (and the tradeoff with wages is large enough) to drive major changes in overall wages. This is a much stronger claim than saying that there is some tradeoff between higher health costs and wages in the total compensation package.(1)

Jonathan Gruber, an economics professor at M.I.T., argued in an op-ed in the Washington Post on December 28, 2009:

And when firms reduce their insurance generosity, they make it up in higher pay for their workers. We saw this in the late 1990s, when the rise of managed care temporarily lowered insurance costs, and wages rose in real terms for the first time in many years. But as soon as managed care was weakened and health costs rose again, we once again saw flat or declining real wages in the United States. (Gruber 2009)

Others, including prominent and well-respected journalists, have also made the argument recently for a “health care theory of wage determination.”(2) Proponents of this theory point for evidence to the latter half of the 1990s, a five-year period when wages were growing rapidly while growth in employer health care spending was relatively constrained. They contrast the period from 1995 to 2000 with the periods from 1989 to 1995 and 2000 to 2006, when wages stagnated while health care costs grew much more rapidly.

There is logic to their argument, but it is only skin-deep and deeper examination will show it to be simply not true. The logic can be seen looking at trends in health care premiums and wages—wage growth fared better in the late 1990s when health care premiums grew more slowly than in the early 1990s and wages performed poorly in the 2000s, a period when health premiums grew strongly again.(3)

However, digging just a bit beneath the surface reveals the following:

1. Health care costs are not large enough to substantially move wages as these proponents claim;

2. Examination of actual wage and benefit trends confirms that changes in the trajectory of health care costs did not materially affect wage trends over the last 20 years; and

3. The wage behavior described—accelerating in the late 1990s and more slowly thereafter—actually best characterizes wage growth for low-wage workers who have minimal access to employer-based health care. Conversely, this pattern of wage-growth over time is least pronounced for higher paid workers with the most health coverage.

Clearly, this “health care theory of wage determination” is wrong, and other factors explain these overall wage trends. The simple explanation is that productivity accelerated in the mid-1990s, and the low unemployment (and hikes in the minimum wage) facilitated faster wage growth. That this wage growth disappeared entirely in the 2002-07 recovery is not due to faster health care cost increases but to weak employment growth and employers’ ability to achieve increased profitability rather than pass on productivity gains to workers. This reveals a fundamental flaw in our economy: productivity gains are not passed on to higher living standards for workers.


It is easy to understand that health care cost trends have not been a significant driver of wage trends when one examines the scale of employer expenditures on health care. Health care costs were just 7.6% of total compensation and 9.4% of total wages (all wages paid, including premium pay, paid leave, and so on) in 2007.(4) The share of health care in total wages (in nominal, non-inflation adjusted terms) grew from 7.2% in 1989 to the 9.4% in 2007, suggesting that the expanded role of health costs could have reduced wage growth by 2.2% over this entire 18-year period, or 0.12% each year. This assumes a complete tradeoff between health costs and wages (if every dollar of higher health costs reduced wages correspondingly). Consequently, employer health costs can hardly be considered a major determinant of wage growth.

Further, overall benefits’ (health care plus all other fringe benefits) share of total compensation has actually been stable for the last 20 years or so—as health costs expanded, pension and payroll tax shares diminished. Hence, the story of stagnant wages in the U.S. economy is not one of growing non-wage compensation.(5)


Digging deeper, consider the changes in health care costs and wages per hour worked in the early 1990s, the late 1990s, and 2000 to 2006, the periods cited by proponents of the health care theory of wage determination.

The data in Table 1 show that wages and total compensation definitely accelerated in the late 1990s, with hourly compensation growth tripling from a $0.22 to a $0.68 annual growth. This alone disproves the theory that moderated health care costs were the primary driver of wages: it was not a change in the mix of compensation between wages and benefits that drove growth, rather it was the simple fact that total compensation accelerated rapidly. After 2000 compensation growth subsided to $0.41 per hour but still remained faster than that of the 1989-95 period.

Do trends in health care costs explain this behavior in compensation or wages? Employer health care expenditures grew $0.03 in the late 1990s, pretty much the same as in the early 1990s, so that hardly seems an explanation. Health care expenditures did grow more quickly in the 2000-06 period (up $0.09 each year).

This greater growth could at most explain $0.06 of the $0.45 deceleration of wage growth from 2000 to 2006 versus the late 1990s. Interestingly, the growth of pension costs is more important than that of health costs. More important, the health care story misses and cannot explain the substantial deceleration (one-third slower) of overall compensation growth in the 2000s.


Most economists would point to the faster productivity growth of the late 1990s to explain the faster wage and

compensation growth. In the early 1990s, as in the 1980s, productivity growth was about 1.3% each year. Productivity growth doubled in the late 1990s to 2.5% annually and maintained that pace in the 2000s. It is the combination of this trend—a faster growing pie—and the lower unemployment and higher minimum wages that allowed workers the leverage to make sure their wage growth kept pace with overall productivity.

But the lessons of the 2000s are also instructive: despite the faster productivity growth there has been no real wage growth recently, either for those with a high school or a college degree (See Mishel et al. (2009), Figure 3A).

There is something fundamentally broken about our economy when workers gain nothing from productivity

growth, and this should give pause to those who assume that when employers lower their health care expenses they will automatically pass these savings onto workers in the form of higher wages. This is an especially problematic assumption given the very high unemployment expected to prevail over the next five years, an environment where workers will have little leverage.


The last piece of evidence on this issue is data on the wage trends for workers at differing wage levels. The same trends discussed above—accelerated wage growth in the late 1990s and the subsiding of this growth in the 2000s—are evident across the board for segments of the workforce that have extensive health care coverage and those for whom coverage is sparse.

This is a point made by Jared Bernstein and Sylvia Allegretto in an analysis in 2006:

About half of all workers don’t even receive employer-provided coverage. According to the U.S. Bureau of Labor Statistics (BLS), 47% of workers did not participate in employer-provided health care benefit plans in 2005. Thus, there is no health care squeeze that would explain the wage losses of nearly half the workforce. In addition, the BLS data show that among workers whose average wage was less than $15 per hour last year, only 39% participated in employer-provided health plans….. low-wage workers also lost the most ground in terms of real wages. Thus, those least likely to get health care experienced the greatest loss in real wages, the opposite of what the trade-off explanation would predict. (Bernstein and Allegretto 2006)


Table 2 shows the wage growth for workers at every decile over the 1989 to 2007 period, including the relevant sub-periods. Wage growth was far faster from 1995 to 2000 than in the 1989-95 period at every wage level. However, the acceleration of wage growth was far greater for low- and middle-wage workers, the groups with the least coverage by employer-provided health care plans: only 27% and 64%, respectively, of workers in the bottom and middle fifths of the wage distribution received employer-sponsored health insurance in 2000 (see coverage by wage fifth in Mishel et al. (2009), Table 3.12). This further reinforces how health care cost containment of the late 1990s was not the major, or even an important, determinant of wage trends. Note that the acceleration of wages for the two highest-paid groups—at the 90th and 95th percentiles—was half that of what the lowest-paid workers enjoyed even though 80% of the highest fifth of earners received employer-sponsored health coverage. This runs directly counter to the notion that health care costs are driving wage trends. Note also that wage growth was substantially diminished in the 2000s, even though productivity growth continued at the same fast pace. In the recovery period from 2002 to 2007 there was hardly any wage growth at all (see Mishel et al. (2009), Table 3.1). The worst wage growth in the 2000s was for low- and middle-wage workers, the groups with the least health care coverage. So, it does not seem likely that faster health care premium growth in the 2000s can explain the disappointing wage growth.

The recent claims that trends in employer health care expenditures explain the beneficial wage growth of the late 1990s and the disappointing wage growth since 2000 does not hold up to any careful scrutiny. Health care expenditures are relatively small compared to overall wages, and an examination of the actual trends shows that health care cost increases do not correspond to the major movements in wages or compensation. This is especially the case for the wage trends of low- and middle-wage workers: their wages accelerated the most in the late 1990s and grew the least in the 2000s. The fact that these groups have the least participation in employer-provided health plans confirms that health care is not the major factor that the advocates of this new health care theory of wage determination would have us believe. There undoubtedly is a tradeoff between health care costs and wage growth, but this dynamic does not play a leading role in the drama of the stagnant wages facing workers for several decades and the inability of working families to benefit from rising productivity growth.

Bernstein, Jared, and Sylvia Allegretto. 2006. The Wage Squeeze and Higher Health Costs. EPI Issue Brief #218. Washington D.C.: Economic Policy Institute. http://www.epi.org/publications/entry/ib218/
Bivens, Josh, and Elise Gould. 2009. The House Health Care Bill is Right on the Money: Taxing High Incomes is Better Than Taxing High Premiums. EPI Issue Brief #267. Washington D.C.: Economic Policy Institute. http://epi.3cdn.net/d4461bae3920d3a28a_7jm6b9314.pdf
Gruber, Jonathan. 2009. ‘Cadillac’ tax isn’t a tax—it’s a plan to finance real health reform. Op-ed. Washington Post. December 28.
Mishel, Lawerence, Jared Bernstein, and Heidi Shierholtz. 2009. The State of Working America 2008/2009. Washington, D.C: Economic Policy Institute.

1 I remain a skeptic that the tradeoff is 100% over any short-run period, especially the high unemployment period ahead. That is a different topic for another time.

2 Ezra Klein, in his (appropriately) highly regarded blog for the Washington Post and in an op-ed, was the first to enunciate this “health care theory of wage determination”:

From 1989 to 1995, median wages actually fell a bit. Then, managed care kicked in. Annual growth in health-care costs fell from more than 10 percent in the early 1990s to less than 5 percent in the late ‘90s. Meanwhile, wages shot through the roof, rising more than 11 percent from 1995 to 2000. Then we ended the managed-care experiment, and health-care costs resumed their normal speed of growth. Predictably, wages slumped back down from 2000 to 2006. (http://voices.washingtonpost.com/ezraklein/2009/12/lower_health_costs_higher_payc.html)
David Leonhardt, an influential economics reporter for the New York Times, weighed on this as well:
A dollar that an employer spends on insurance is a dollar that’s unavailable for income. This helps explain why the one period of slow growth in medical costs over the last two decades — the late 1990s — was also the one period of rapid income growth. (http://www.nytimes.com/2009/12/23/business/economy/23leonhardt.html?_r=1)

3 See Exhibit One in this compilation from the Kaiser Foundation: http://www.kff.org/insurance/7672/upload/7693.pdf.

4 See the Bureau of Labor Statistics’ survey of Employer Costs for Employee Compensation, Table 9, ftp://ftp.bls.gov/pub/special.requests/ocwc/ect/ececqrtn.pdf.

5 See the Bureau of Labor Statistics’ survey of Employer Costs for Employee Compensation, Table 9, ftp://ftp.bls.gov/pub/special.requests/ocwc/ect/ececqrtn.pdf.

crossposted from Economic Policy Institute, originally published on January 6, 2010.