The economic recovery from the recession of 2001, both nationally and in Massachusetts, has been not only mixed, but also puzzling in a number of key respects. Nationally, growth has been fairly robust in Gross Domestic Product and corporate profits, and housing prices have risen rapidly in both the state and the nation through the end of 2005. The September 11 terrorist attacks, the war in Iraq, surging oil prices, and weather-related disasters have done little to diminish economic growth nationally since the recession ended in the fourth quarter of 2001. Yet labor markets, both in the nation and especially in our state, continue to lag behind other key economic variables, at least by some key employment measures.

Payroll job growth has been weak, but there are some indications that employment has changed in character more than it has stagnated, with more people working for themselves, as independent contractors, or in the underground economy. And there are troubling signs that immigration, both nationally and in Massachusetts, has gone from a boon to an economic challenge, especially for unskilled and younger workers, including minorities, who are being crowded out of vital first steps in the job market.

The mixed picture of national employment developments comes from the two monthly surveys that the US Bureau of Labor Statistics conducts to provide data on job growth. One of these surveys is the Current Employment Statistics program (CES), which measures the number of wage and salary jobs on the formal payrolls of private and public employers in the nation. Similar surveys take place at the state level. Known informally as the “payroll survey,” the CES measures employment in traditional jobs in the nonfarm sector of the nation’s and the state’s economy.

The rate of new job creation as measured by the payroll survey over the past four years has been slow by historical standards. Table 1 shows that the nation averaged a 10.4 percent rate of new job creation during the first 15 quarters of recovery from the five recessions that occurred from the 1960s through the 1990s. During the current recovery, however, payroll employment levels have increased by just a little over 3 million, or 2.3 percent, since the trough of the recession in the fourth quarter of 2001. This represents a rate of new job creation less than one-fourth that of the historical average for a recovery of this length.

Some labor market analysts have speculated that offshoring has been a chief reason for this slow rate of growth in payroll employment levels over the past few years (see “Offshore Leave,” CW, Summer ’04). Offshoring refers to US firms—especially in the manufacturing and information technology sectors—that purchase goods and, increasingly, services from abroad, in place of inputs formerly provided by in-house staff or by employees of other firms located in the US. Offshoring results in a shift of employment from US business establishments to firms overseas. Low-cost communications, an emerging high-quality labor supply abroad, and low foreign labor costs all combine to shift even highly sophisticated production and business processes overseas.

While there have been substantial reductions in information technology employment levels in the US and Massachusetts over the past few years, the limited research available provides little evidence of offshoring as the major cause of employment stagnation. During 2003, the US Department of Labor’s Mass Layoff Statistics survey found 1.5 million reported layoffs, but only 13,000 of these were attributed to offshoring. In the Monthly Labor Review, Robert Bednarzik reviewed a number of private studies of outsourcing in the IT sector and found an estimated impact on the nation’s information technology sector ranging from 34,000 to 72,000 jobs a year.

Other analysts say that the fundamental reason why GDP has grown far more rapidly than new jobs is accelerated labor productivity growth. Researchers at the Federal Reserve Bank of Boston have asserted that “productivity essentially has been the dominant engine of growth in nonfarm business output during the current recovery.” Still, the authors of this Boston Fed study were themselves somewhat puzzled by the surge in productivity growth. They hypothesized that employers were relying on rising worker productivity to expand output because they were unsure about the long-term strength of the economic recovery and, thus, were reluctant to expand payroll employment levels.

OFF-PAYROLL EMPLOYMENT

But there is another possible explanation for the gap be-tween economic expansion and payroll job growth: that aggregate employment levels have in fact grown faster than the payroll survey shows. This employment growth has occurred primarily outside of the nation’s traditional employer-employee relationships, including self-employment, independent contractor employment, and off-the-books employment, and is not captured by the monthly payroll survey. Indeed, we find that much of the new employment generated in the US and in Massachusetts since the end of the recession has been in off-payroll and under-the-table jobs.

Clues to this phenomenon can be found in a second source of national monthly employment data, the Current Population Survey (CPS), a survey of approximately 60,000 households conducted each month by the Census Bureau for the Bureau of Labor Statistics. The “household survey,” as it is commonly called, counts the number of persons ages 16 and over who are working. Included in this employment count are not only workers in traditional employer-employee jobs, but also self-employed, independent contractors, independent consultants, unpaid family workers, family household workers, and workers, both native-born and immigrant, who hold under-the-table jobs.

Historically, the payroll and household employment surveys have tracked fairly closely over the course of the business cycle. During the recession of 2001, for instance, both surveys generated very similar findings about the magnitude of job losses. The payroll survey found that the number of nonfarm payroll jobs in the US had declined by slightly over 1.7 million between the first and fourth quarters of 2001. The CPS household survey found an employment drop of almost the same magnitude, with the estimated number of employed persons declining by 1.58 million over the same time period.

Since the end of 2001, however, the two surveys’ estimates of employment change have diverged considerably. The payroll survey estimated that the nation added only 3.05 million wage and salary jobs from the fourth quarter of 2001 to the third quarter of 2005. But the household survey reported double the employment gain, nearly 6.3 million more people working, over the same time period (Chart 1). With strong growth in employment outstripping growth of the nation’s civilian labor force, the national unemployment rate declined from a high of 6.3 percent in the summer of 2003 to 5.0 percent in recent months, which some economic analysts consider to be the equivalent of full employment.

A number of national financial and economic analysts have been quick to dismiss the household survey’s findings on employment growth, claiming that only the payroll job numbers are to be given credence. Ironically, these same analysts frequently cite the unemployment rate estimates from the CPS survey as a measure of underlying labor market strength. But one cannot reject the employment estimates from the CPS survey while at the same time accepting the unemployment estimates from the same survey. Indeed, the underlying arithmetic of the unemployment rate says that, if only 3.05 million new jobs had been created over the past four years instead of the 6.3 million revealed in the CPS survey, the nation’s unemployment rate would today be about 7.0 percent, two points above its current 5.0 percent level.

Whether measured by the household survey or the payroll survey, the job market recovery has been much weaker in Massachusetts than in the nation as a whole, but there is still a substantial gap between the two indicators of employment change. Since the national economic recovery began in late 2001, state payroll employment has fallen by 70,600 jobs; the household survey estimates that state employment levels declined by only 40,200 over that time period. Between the peak of employment in the first quarter of 2001 and the first quarter of 2005, payroll employment in Massachusetts declined by 177,000 while household survey employment fell by only 79,000, a gap of nearly 100,000 (Chart 2).

CONTRACTORS, CONSULTANTS, AND IMMIGRANTS

The best evidence suggests that both the household and payroll surveys provide valid measures of labor market activity. Differences between the findings of the two surveys with respect to employment growth have arisen in part because they measure different kinds of employment activities in the job market. The simple fact is that much of the recent employment growth in American labor markets has occurred off the nation’s traditional employer payrolls in labor market segments not counted in the payroll survey but covered by the household survey.

In our view, a major source of the gap between the two surveys’ employment estimates, in both the nation and the state, is attributable to a rise in independent contractor employment and off-the-books workers, along with a sharp increase in the number of foreign immigrant workers. We found in Massachusetts that over the 2000-04 period the number of persons reporting themselves as self-employed increased by 43,000 even as payroll employment levels plunged. Some of these self-employed individuals represent the classic entrepreneur who sees a business opportunity and strikes out on his or her own, engaging in all of the risk-taking, selling, and decision-making that creating a new business requires. However, others in this category are second-option self-employed. These individuals became self-employed after a permanent layoff from their former jobs. Examples abound, including tech workers who offer programming services from home after finding themselves unable to become re-employed in the state’s ailing information technology industry; a physicist who, after losing his job in a research lab, begins a handyman service on the Cape; and high-tech workers starting catering businesses or landscaping firms. The hallmark of these second-option self-employed is their desire to find a permanent wage-and-salary job that utilizes their skills and abilities, rather than stay in business for themselves.

In addition to the growth in the number of self-employed in the state, we find an increase of more than 60,000 workers who are employed either as contract workers or who work in under-the-table, cash-payment jobs. Contract workers, including construction workers, engineers, journalists, MIS analysts, and some health professionals, are employed by firms in jobs that are characterized by an employer-employee relationship, but in which firms treat the workers as self-employed for tax and benefit-eligibility purposes. Research in this area has found that payroll and independent contractor workers are virtually indistinguishable from one another with respect to job duties, supervision, and other work governance issues. Under-the-table workers are simply those who work for cash payments—frequently those in daily labor pools, but also landscapers, construction workers, restaurant workers, and security guards.

More is known about recent immigration inflows and the changing role of new immigrants in the country, especially in Massachusetts. Despite a national economic recession, slow payroll employment growth, and half-hearted attempts to curtail the flow of illegal immigrants in the nation, the number of foreign-born persons who have settled in the US since the beginning of the decade has reached nearly 8 million, making this the highest five-year period of immigration in the nation’s history. About 3.7 million of these new immigrants are estimated to reside in the country illegally.

Our analysis of the most recent CPS survey data for 2005 reveals that more than 4 million of these recent immigrants were employed during the first 10 months of 2005. Over the entire period 2000 to 2005, civilian employment as measured by the household survey increased by slightly less than 4.6 million. Thus, these new immigrants have accounted for nearly 90 percent of new employment growth and well over 100 percent of employment growth among males. At the same time, employment rates of most groups of native-born workers, especially males and those under 30 years old, have declined. During 2000, the share of the native-born civilian, non-institutional, working-age population who were classified as employed was 64.5 percent. This proportion declined by about 2 percentage points to 62.6 percent, by 2005.

It is important to note that the decline in the native-born employment rate was modest among the population ages 30 and older, and the employment rate actually increased among persons 55 and older. Most of the decline in job finding for native-born persons over the last five years, both nationally and in Massachusetts, was concentrated among teens and young adults under the age of 30, especially males without college degrees (Table 2).

Among the native born, teen employment levels fell sharply, with their employment-to-population ratio dropping from 46.3 percent to 38.2 percent nationally over this five-year period, a decline of 8.1 percentage points—the lowest employment rate for teens post–World War II. In Massachusetts, employment rates for native-born teens fell by nearly 6 percentage points over the same time period, from 48 percent to 42 percent. Nationally, employment rates for native-born persons ages 20 to 24 declined from 73.3 percent to 68.7 percent during this time period, while in Massachusetts the decline for this age group was even larger, falling from 71 percent to 64 percent (Chart 3). Even among native-born 25-to-29-year-olds, employment rates have fallen considerably over the last five years. Only persons 55 and older have seen their employment rates rise in both the nation and the state since the end of the labor market boom in 2000. Young black males have experienced disproportionate losses in their employment rates, but declines in employment-to-population ratios were also large among young white male high school graduates and dropouts.

These findings suggest that immigrants are playing a very different role in the job market, both nationally and locally, since 2001. During the 1990s, immigration helped to fuel the strong growth in employment that occurred in the nation and our state. Indeed, during the 1990s economic expansion, payroll employment outpaced employment growth as measured through the household survey. For both new immigrants and the native born, employment was largely “on the books,” as workers of both types primarily held regular wage-and-salary jobs. Importantly, there was little evidence of foreign-born workers displacing native-born workers, even among the young. Teen and young-adult employment rates increased over the decade as the expanding job market absorbed entrants from both groups.

But now there is reason to believe that the work obtained by new immigrants is coming in part at the expense of native-born workers, especially young adults with low education and skill levels. Multivariate analyses by the Center for Labor Market Studies reveal an adverse effect of rising immigrant employment on teen and young-adults employment-to-population rates since 2000. These results hold true across states, including Massachusetts. Even during the 1990s boom, there was evidence that immigrants contributed to real wage declines among high school dropouts and young workers with limited schooling and experience. Now, it appears that these marginally employable young people are losing whatever job prospects they might have to foreign immigrants.

The loss of early work experience among teens and young adults isn’t just about spending money for kids. Early work experience contributes to long-term employment and earnings gains for young people over time. Moreover, some researchers have found that early work experience among minority males, especially blacks, helps bolster the post-secondary enrollment rates of recent high school graduates.

At the same time that the number of employed new immigrants has increased sharply both in the state and nation, new forms of work have emerged, including independent contractors, temporary consultants, and off-the-books workers. Groups of immigrant workers stand in front of convenience stores and gas stations every morning, waiting to be picked up by would-be employers in landscaping, hotels, restaurants, and construction industries, among others.

Public policy-makers have failed to address the changing structure of these employment relationships and their adverse economic and fiscal consequences. Lax enforcement of basic employer-employee relationship requirements under the Social Security Act and the IRS codes, failure to enforce wage-and-hour laws, and inability to halt the flow of illegal workers into the nation’s labor markets have all contributed to this deterioration of labor standards in the state and the nation.

Paul E. Harrington and Andrew M. Sum are associate director and director, respectively, of the Center for Labor Market Studies at Northeastern University.