Opinion pieces and speeches by EPI staff and associates.

[ THIS TESTIMONY WAS PRESENTED BEFORE THE COMMITTEE ON FINANCIAL SERVICES OF THE U.S. HOUSE OF REPRESENTATIVES ON FEBURUARY 16, 2007. ]

Monetary Policy and the State of the Economy

By Jared Bernstein

Introduction
I thank Chairman Frank and members of the committee for the opportunity to testify on the critical issue of the importance of full employment in today’s economy.

In recent months, top policy officials and economic commentators have wondered why there seems to be more economic anxiety among working Americans than might be expected given the low unemployment rate and solid macroeconomy.  Though the Iraq war was obviously prominent in voters’ minds during the midterm elections, polls revealed that significant majorities were concerned about policy makers’ handling of economic issues as well as their own economic circumstances.¹

While some officials remain puzzled by this apparent disconnect between macroeconomic performance and perceptions of economic well-being, Figure 1 suggests it should not be such a “head-scratcher.”  After rising at rates close to that of productivity over the latter 1990s, the real median wage flattened in real terms, as did the average real wage of high school and even college-educated workers.  Granted, the figure compares the second half of a recovery (1995-2000) with a period that includes a recession, but unemployment was below five percent by late 2005, and productivity growth has been even stronger in the 2000s than in the latter 1990s. 

In other words, the productivity/wage gap towards the end of the figure is not wholly a cyclical phenomenon.  Nor can it be explained away, as some have tried to do by citing the growth of employer-provided fringes.  Over the five years of economic recovery—2001q4-2006q4, real compensation grew at an annual rate of 0.8% while productivity grew 2.8% per year.² In fact, as is extremely clear in Figure 2, which plots the long-term relationship between productivity and the compensation of non-managerial workers, this gap between growth and living standards of working families has been a prominent and unsettling feature of the economy over recent decades.³

There are many reasons why this gap has evolved but in this testimony I focus on two related determinants: the absence of full employment and the loss of worker bargaining power.  The key points are as follows:

• The combination of slack labor markets and diminished distribution mechanisms such as unions, minimum wages, and more balanced trade, has led to a persistent gap between growth and living standards for many working families.

• One reason for this imbalance has been allegiance to a “natural rate” theory of unemployment.  Despite little evidence to support its contemporary use, this theory has led policy makers to give greater weight to inflation relative to unemployment concerns, and has thus been partly responsible for years of unnecessary slack in the job market.

• The post-1973 growth slowdown does not explain the extent to which middle incomes have lagged growth.  Other factors are in play, including slack job markets and diminished bargaining power.

• During periods of full employment, middle-family wage and income growth has been much stronger, and inflation has grown no faster than periods of slack job markets.

• Economic elites have been operating from a playbook with an inherent bias against broadly shared prosperity.  There is, at this point in time, no legitimate macroeconomic rationale for this bias, and I urge policy makers to chart the course back to a policy set that restores full employment, worker bargaining power, and much more equitable outcomes.

The Evolution of Natural Rate Theory and Its Impact on Working Bargaining Clout
Any discussion of full employment must deal with the concept of the NAIRU, the non-accelerating inflationary rate of unemployment.  For the purposes of this panel, the important point is that the NAIRU model generates the “natural rate of unemployment,” the rate below which inflation would not merely rise, it would continuously accelerate until unemployment went back up to the so-called natural rate.  The Congressional Budget Office currently estimates this rate to be five percent.

Its relevance to our discussion today is clear.  If this concept is wrong in that it fails to accurately describe the relationships between these key variables, using the NAIRU to set monetary policy could lead to unnecessary slack in the job market, which is particularly devastating to the least fortunate and demonstrably harmful, as I show below, to the income growth of middle-income households as well.  Similarly, if the concept is right but we cannot accurately estimate the correct value—the unemployment ‘floor,’ as it were—we also risk the danger of erring on one side or the other.

Like many economists today, I consider the NAIRU to be a poor guidepost for policy makers.⁴ Events have overtaken the original model, and while some useful advances have been made (Ball and Moffitt, 2001), the evidence I present below suggests that subscribing to the model risks persistent and unnecessary slack in the economy, wasting billions of dollars and consigning millions of potential workers to fewer job opportunities and lower wages than should be the case.

In brief, and simplifying considerably, the original NAIRU story was based on the idea that the unemployment rate is usually at equilibrium, i.e., just where it should be in terms of supply, demand, and stable prices.  Those unemployed at this “natural rate” can’t find work at the wage they think they deserve, but that’s because they have an upwardly skewed view of their worth (or marginal product).  However, the monetary or fiscal policy authorities want to lower the unemployment rate, and get these folks a job, so they undertake stimulative actions (lower interest rates, tax cuts, etc.).

Wage offers do rise, and these formerly jobless workers leave the sidelines and join the job market.  However, firms offset their now higher labor costs by raising prices.  Soon, the new workers recognize that they’ve been tricked: their reservation wage, though finally met, is not going as far as they think it should.  They then demand yet higher wages—and this is a key piece of the theory that I critique in a moment—and the wage/price spiral is underway.  The only way to stop it is for unemployment to return to the natural rate, but by then, the inflationary damage is done.

When this model was developed in the 1960s, it had strong predictive power and seemed to accurately describe events, particularly in the latter 1970s.  But evidence over the past decade, stemming both from research and real events, suggests that the model no longer holds and that policy maker’s allegiance to it has done more harm than good—in my view, a lot more.

Most recently, the NAIRU took a beating in the latter 1990s, a period of great relevance to my testimony regarding the benefits of full employment.  At that time, economists believed that full employment was consistent with an unemployment rate of six percent, so when the jobless rate began to fall well below this alleged natural rate, the NAIRU-ites expected inflation to accelerate, especially given the increase in real wages.  To the contrary, inflation decelerated.

Undaunted, NAIRU devotees made two adjustments.  First, they allowed that because of changes in the economy and workforce, the value of the natural rate varied over time.  This led to estimates like those in Figure 3, created by top macro-economists in a 2001 publication.⁵ The figure shows that the TV (time-varying) NAIRU is just about the same as (is statistically indistinguishable from) the unemployment rate.  If this is true, then we can’t know the NAIRU in real time (since the trend will be changing in unknowable directions in real time) and the concept becomes a very weak guidepost.  As Leone noted, “If the ‘time-variant NAIRU’ looks to all the world like a smoothed version of the actual unemployment rate…then policymaking based on the NAIRU comes perilously close to a formalized version of the simple rule of thumb: ‘keep the unemployment rate near its average of recent quarters.’” ⁶

Second, economists began to add productivity growth to the model, a variable that was not in the original formulation.  This has turned out to be an important development indeed.  Alan Greenspan, for example, appears to have recognized that with productivity accelerating in the latter 1990s, higher labor costs could be non-inflationary, as they were offset by steady unit labor costs.  This makes sense: firms in the latter 1990s generally kept their profit margins intact, and paid higher wages out of more efficient production, obviating any pass-through to prices.

It makes sense, but it’s not a NAIRU model, and this is part of my message.  It is not at all clear that the Federal Reserve is operating from this model.  The unemployment rate has been well below the CBO NAIRU, yet the Fed is into a sustained pause regarding interest rates and inflation appears to be decelerating.  While concerns about tight job markets generating wage-push inflation are of course evident in their statements and speeches, the actions of neither the Greenspan (in the second half of the 1990s) nor the Bernanke Fed appear to be dominated by concerns about the natural rate.

I believe one reason for this is the diminished bargaining power of the American worker.  Over the post-1979 period when inequality and wage stagnation were most pronounced, unionization rates fell by half, from about 25 to 13 percent.  The real value of the minimum wage fell to its lowest level in 50 years.  Trade penetration has climbed steadily (exports plus imports are up from about 19% to 28% of GDP since 1979) and our international indebtedness has grown to historically unprecedented levels, as globalization and offshoring increasingly put U.S. workers in competition with workers around the world.

With significantly diminished bargaining power, the NAIRU story falls apart.  Unemployment is not voluntary, and workers can’t push for ever higher wages.  Econometricians can make these models fit by allowing them to move with trend unemployment or productivity growth, but these changes violate the basic premise of the model, which is based on workers’ behavior and expectations, and these behaviors have changed in ways that render the model ineffective.

The structure of the economy has changed significantly over the past few decades, and these changes have had a major impact on the distribution of power, and thus the distribution of economic rewards.  Simply put, in a labor market that lacks the institutions and norms that provide workers with some bargaining power, in an economy where those whose access to power and assets gives them a huge upper hand in the distribution of wealth, the predictable outcome is precisely the surge in inequality we have seen over the past few decades.

In this context, full employment partially takes the place of those institutions and norms that previously ensured a more equitable distribution of growth.  It is one of the few reliable sources of bargaining power available to the American worker today.

Full Employment, the Natural Rate, Inflation, and Median Family Income Growth:
A Long-Term Perspective
With the caveat that I’m reducing many big and complicated changes over many years into a few key variables, the following two tables make an important and compelling case for a new economic policy that gives great prominence to concerns regarding bargaining power, full employment, and inequality.  

The first table shows the annual growth rates in the real income of the median family, productivity growth (the source of improved living standards), and average unemployment, a rough proxy for full employment/bargaining power.  Obviously, big changes occurred in our economy over these long periods.  Changes in family and labor force composition, family labor supply, immigration, technology, trading regimes, and much else are embedded in these trends. 

In the first period, real median income and productivity grew at about the same rate, while average unemployment was 4.8%.  In the second period, median income grew much more slowly, and the average unemployment rate was considerably higher: 6.3%.  The lack of income growth over this second period is typically ascribed to the slowdown in overall growth—shown in the table as a point less per year of productivity growth (1.8% vs. 2.8%).

But these data show that it is not correct to stop with the insight that both the economy and income grew more slowly, post 1973.  The deceleration in middle-class income growth, shown in the bottom line of the table, was much greater than the slowdown in productivity growth.  Note also that this income slowdown occurred despite the large increase in women’s labor supply; Mishel et al (2006, Table 1.24) show that among middle-income families with children, 1979-2000, working wives added over 500 hours, the equivalent of over three months of full-time work in the paid labor market. 

In other words, along with the productivity slowdown, the list of bargaining-power-reduction factors noted above came into play, and what growth there was flowed less to the broad majority of working families.  At the same time, unemployment was 1.5 points higher on average.  The fact that there was less growth to go around is only part of the story. Because of shifting bargaining power and the loss of full employment conditions, middle-income families worked harder but gained less. 

Again, these outcomes have many determinants, but one is the shift away from a macroeconomics based on full employment to a microeconomics based on managing individual incentives and expectations.  Moreover, despite statistics like those above, the consensus view among economists and policy elites is that the policy shift has been a smashing success.  Paul Krugman, in a recent essay on Milton Friedman, a founder of the natural rate theory, made this same point: “…given the common assumption that the turn toward free-market policies did great things for the US economy and the living standards of ordinary Americans, it's striking how little support one can find for that proposition in the data.”⁷

A final table, similar to the one above, adds a NAIRU analysis to the picture.  Using CBO historical estimates of the NAIRU, we can determine when the actual unemployment rate was above or below the “natural rate.”  The first column of the table accumulates the annual percentage-point differences over the two time periods.  Thus, if CBO’s NAIRU was 5% and the actual jobless rate was 4.5%, this would show up as a -0.5 percentage point in our analysis.

Over the period when middle-incomes tracked productivity, the unemployment rate was often below the NAIRU, cumulatively 19 percentage points below over the years 1949-73.  This happens to be about the same number of points that unemployment was above the NAIRU in the latter period.  Not only was middle-income growth much higher in the period when we were often below the NAIRU, but inflation was lower as well (and note that the latter-year inflation comparison leaves out the years of very high inflation in the early 1980s to avoid an upward bias; nor do I adjust latter period income growth for families’ additional hours of work).⁸ 

Clearly, from the perspective of middle-class incomes, tight labor markets, even below the supposed natural rate, were associated with much better income growth.  As these numbers suggest, the concept of the natural rate has not been helpful to the median family.

 

Sources: CBO NAIRU estimates; Census Bureau, median family income (RS deflator);BLS, unemployment; BLS, CPI-RS deflator.

Conclusion

The data in the table above cover many years and many regimes, leading some, no doubt, to discount their relevance. But consider that over the current business cycle, productivity growth has been stellar, growing once again at an annual rate of 3% (2000-06), while real median family income is down 0.5% per year (2000-05, most recent data). And look back at Figure 1 for a graphic depiction of a recovery that, at least from the perspective of middle-income workers, has been productivity rich but income poor.

I believe a reason for this productivity/income gap is that policy makers, led by neoclassical economics, have abandoned the two main goals of the economics of an earlier era: (1) ensuring that we as a society tap our collective potential and fully employ our economic resources, especially people, and (2) providing individuals with ample protections and publicly provided insurance against undesirable market outcomes—weak job creation, high unemployment, rising poverty rates, falling real incomes—and other challenges like aging out of the workforce or becoming disabled.

This policy regime has been replaced by one devoted to: (1) getting rid of the policy set associated with the old economics and (2) making sure that individuals are offered the optimal incentives, the ones that should lead them to behave in ways that, according to the models, bring about the most efficient results.

In other words, the target of economic policy has shifted away from maximizing society’s potential through promoting full employment and insurance against market failures, and towards incentivizing the individual’s interactions with the market.

Today, we’re seeing the results of this shift: greater inequality, a fiscally strapped federal government, the shifting of risk from the government and the firm to the individual, the drying up of large risk pools, and the loss of the systems and institutions—like pension coverage, minimum wages, overtime rules, and a durable safety net—that smoothed some of the rough edges of our market economy, without diminishing its growth potential.

Can we, however, realistically take lessons from the past given the myriad changes that have occurred over the last few decades?  As Andy Stern recently pointed out, “We’re as far from FDR as he was from Lincoln, and I doubt he was looking to Honest Abe for policy guidance.”

My colleagues at the Economic Policy Institute and I believe we can construct a policy architecture that both respects the opportunities and constraints of the modern, global economy, while lastingly reconnecting the living standards of the median family to the growth of productivity.  Our project, the Agenda for Shared Prosperity , is well underway, and we encourage committee members to learn more about the ideas that we, along with some of the most progressive and innovative thinkers in economics today, are developing.

Along with health and retirement security, fair trade, education and fiscal policy ideas, we are developing an agenda for rewarding work, and maintaining full employment is at its heart.  This, unsurprisingly, includes a role for the Federal Reserve.  This role may entail rebalancing the inflation/unemployment trade-off with less emphasis on fighting the phantom menace of wage-push inflation.  This committee may choose to take another look at the Humphey-Hawkins Act with an eye towards elevating the goal of full employment.  The Fed chair, in his or her HH testimony to this body, might be required to explain whether the labor market is at full employment, and if not, what steps would be taken to get there.

For groups that are chronically underemployed—those out of the reach of even tight job markets—we might want to consider direct job creation.  In Savner and Bernstein (2004), we argued that “…when labor markets are slack, we create public-service jobs to keep people gainfully employed, drawing on the successful experience of transitional jobs programs that have sprung up around the country using public funds to create work for people struggling to get a foothold in the labor market. Such jobs could meet important community needs and let people use their newly minted skills. What’s more, the message is clear and consistent with values we all agree with: Everyone who wants to work should have the chance to do so.”¹⁰    

Of course, given that diminished worker bargaining power is at the heart of productivity/wage gap, measures that rebalance these dynamics will also facilitate a more equitable distribution of growth.  Measures to level the playing field for union organizing, like the Employee Free Choice Act, make sense, as does the recent legislation to raise the Federal minimum wage.  EPI’s policy project suggests a framework for pushing back against the downsides of globalization while preserving the gains, including more aggressive tactics to reverse growing international indebtedness, and getting Congress back in the business of writing more balanced trade agreements.¹¹
 
Opponents and defenders of the status quo may well argue that such proactive efforts will distort the invisible hand, lead to spiraling inflation, embolden workers to push for wages beyond their “marginal product,” spook financial markets, and so on.  But the historical evidence undercuts these arguments, showing no discernable utility of the “natural rate” model in the current context.   Moreover, the magnitude of the productivity/wage gap, the extent of poverty amid plenty, the staggering levels of wealth concentration, the abuse of executive pay-setting practices, and the depth of economic insecurity among the American middle class suggest that the status quo is unacceptable.  I urge the committee to consider these and other ideas as you undertake to chart a path back to broadly shared prosperity.

The author thanks Ross Eisenbrey and James Lin for helpful comments and edits.

Bibliography

Bernstein, Jared. All Together Now: Common Sense for a Fair Economy. San Francisco: Berrett-Koehler Publishers, 2006.

Bernstein, Jared and Dean Baker. The Benefits of Full Employment: When Markets Work for People. Washington D.C.: The Economic Policy Institute, 2003.

Faux, Jeff. “Globalization that works for working Americans.” Economic Policy Institute Agenda for Shared Prosperity. 11 Jan. 2007. < http://www.sharedprosperity.org/ bp179.html>.

Krugman, Paul. “Who was Milton Friedman?” The New York Review of Books. 15 Feb 2007. < http://www.nybooks.com/articles/19857>.

Leone, Richard. Foreword. The Roaring Nineties: Can Full Employment be Sustained? Alan B. Krueger and Robert M. Solow, ed. New York: The Century Foundation Press, 2002. xi-xiii.

Mishel, Lawrence, Jared Bernstein, and Sylvia Allegretto. The State of Working America. Ithaca: Cornell University Press, 2007.

Savner, Steve and Jared Bernstein. “Can Better Skills Meet Better Jobs?” The American Prospect. 1 Sept 2004. < http://www.prospect.org/web/printfriendly-view.ww?id=8357>.

Endnotes
[1] For example, according to the New York Times exit poll, two-thirds of voters reported that they are either just maintaining their living standards (51 percent) or falling behind (17 percent).
[2] Real compensation is from the Employment Cost Index, and productivity is nonfarm business.
[3] Wage data in Figure 2 are from the BLS series on average hourly wages of production, non-managerial workers (about 80% of the workforce).  These values are scaled up to include fringes by multiplying them by the ratio of compensation to wages from the NIPA accounts.
[4] International Economist, Benefits of Full Employment, Chapter 2.
[5] Staiger, Stock, and Watson (2001).
[6] Richard Leone, in foreword to Krueger and Solow, 2001.
[7] See Krugman, http://www.nybooks.com/articles/19857. 
[8] Core inflation follows a similar pattern, 2.9% in period one vs. 4% in period two, also omitting 79-82 from period two.  BLS core inflation data begin in 1958, so period one goes from 1958-73 is this comparison.
[9] Some of the material in this section is from Bernstein, 2006, Chapter 2.
[10] http://www.prospect.org/web/printfriendly-view.ww?id=8357. 
[11] See Faux 2007, http://www.sharedprosperity.org/bp179.html.

---

Jared Bernstein is an economist at the Economic Policy Institute in Washington, D.C.

[POSTED TO VIEWPOINTS ON FEBRUARY 16, 2007. ]