PDF Version

State of Working Massachusetts - Labor Day 2012

August 31, 2012
  • Table of Contents
    • Introduction
    • Jobs
    • Unemployment
    • Wages
    • The Wage-Education Link
    • The Minimum Wage
    • Income
    • Poverty
    • Appendix A: Poverty and Social Programs
    • Appendix B: Methodological Notes


Workers across Massachusetts are still hampered by the lingering effects of the recent recession. Even though the economy has stopped shrinking and begun to grow, that growth is too slow to fully restore the losses experienced by workers and families in recent years. There are still fewer jobs, higher unemployment, and higher levels of poverty than before the recession began.

For low- and middle-income families, the burden of this long recession has been especially heavy. Partly, this is because lower-income families always have fewer resources to fall back on in difficult economic times. But, in this case, that is exacerbated by the growth of inequality.

Thirty years ago, wages for low-income workers tended to grow at roughly the same rate as wages for high-income workers. By the 2000s, that had changed. People at the top of the Massachusetts wage-ladder saw gains while workers at the bottom actually saw their wages decline. This left low- and middle-income people more vulnerable to the recession.

There is some good news for workers in Massachusetts: they are faring better than workers elsewhere in the nation. At 6.1 percent, our unemployment rate is lower than the national rate of 8.2%, and even when the recession was at its most severe, our economy outperformed most other states. There are many reasons for Massachusetts's relative strength, but one of them has to do with the policy choices we made to develop a highly-skilled and highly-educated workforce. Today, Massachusetts has the best-educated workers in the country, which has helped to make ours a relatively strong, high wage economy.

Ultimately, though, the fact that Massachusetts is doing less badly than other states is cold comfort. The state of working Massachusetts, today, is still far removed from the kind of economy where workers can find good jobs and families can afford decent lives.

This Labor Day update to the State of Working Massachusetts includes recent wage, job, and employment data available from Census Bureau's Current Population Survey and the Current Employment Statistics survey conducted by the Bureau of Labor Statistics. A fuller look at national trends will be included in the Economic Policy Institute's forthcoming 12th Edition of The State of Working America, scheduled for release on Tuesday September 11, 2012.


The most recent recession is the worst on record for the U.S. since the Great Depression in terms of the peak share of jobs lost. By this measure, the depth of the recession in terms of share of jobs lost was reached in February of 2010, with the US having shed 6.1 percent of the jobs it had in December 2007, the official start of the recession.

The economic recovery following the Great Recession has been too weak to restore all the jobs that were lost during the downturn. As of June 2012, fifty-four months from the official start of the recession, the U.S. still has 3.6 percent fewer jobs than were available at the recession's start. In each recession since the Great Depression this many months after the onset of the downturn, the US has regained all the jobs lost during the recession and in most cases added substantially to the number of jobs available (see chart, below). The slow progress in closing the jobs gap and returning to more robust economic growth is most likely the result of a failure to adopt economic policies that adequately respond to the depth and severity of the recession and provide the sort of stimulus needed to increase demand.1

In Massachusetts, however, a far smaller share of the total number of jobs that were available just prior to the onset of the Great Recession remain lost (1.5 percent) than is true for most other states (US average = 3.6 percent). Note that July jobs data released as this report was being finalized show a slight uptick in the number of jobs in the state, but also show a slight uptick in the unemployment rate, as more workers enter the labor market.

The jobs that have been lost in Massachusetts to date—a net loss of more than 48,000 jobs since the start of the recession in December 2007—have not been spread evenly across all employment sectors. Both Construction and Manufacturing jobs have been hard hit, while the Education and Health sector and the Leisure and Hospitality sector both have seen modest growth in the number of jobs available in Massachusetts. With few exceptions, a similar pattern of sectoral job losses/gains has been true for the US as a whole as well.

At the same time that the nation lost a record share of its jobs during the Great Recession, it also took far longer than usual for the economy to begin growing again and thus to stop shedding and instead begin adding jobs. Simultaneously, the population of working age adults has grown throughout this period. The result is that 10.8 million jobs are now needed in order to restore the US economy to rates of employment in effect prior to the recession's onset.

1 While there is good evidence that the federal stimulus prevented an even longer and deeper recession (see, for example, Alan Blinder and Mark Zandi, How the Great Recession Was Brought to an End, July 27, 2010), the weakness of the recovery suggests a need for additional measures. See Generate Jobs Now, a recent commentary on this topic from the Economic Policy Institute.


Fifty-four months after the start of the Great Recession Massachusetts has lower levels of unemployment than most other states. With the Massachusetts unemployment rate at 6.0 percent, however, this still means that about 208,000 workers are actively seeking work but remain without jobs, while thousands of others have given up the search entirely. For those Bay Staters experiencing the income loss and uncertainty that accompanies short or long-term unemployment, the impact of the current downturn is both very real and deeply problematic. (Note that July employment data released as this report was being finalized showed a slight uptick in the unemployment rate, to 6.1 percent. At the same time, however, the number of employed people also increased, and so the uptick in the rate is likely due to an increase in workers entering, or re-entering, the labor market.)

Massachusetts has maintained lower unemployment rates than the US throughout the duration of the Great Recession and its aftermath.

The overall lower rate of general unemployment that Massachusetts has maintained relative to other states, however, conceals the large disparities in unemployment rates experienced by different groups of workers within the Massachusetts economy. Massachusetts workers with lower levels of education have experienced much higher levels of both regular unemployment (unemployed for 26 weeks or less) and long-term unemployment (unemployed for more than 26 weeks) than have Massachusetts workers with more education.

Because educational attainment is closely correlated with annual earnings,2 these less educated workers are very likely to be among the portion of the Massachusetts workforce with low incomes. They are also likely to have experienced little or no growth in earning power in recent decades (see discussion below). As a result, the burdens of unemployment experienced by workers in the Bay State during this historic downturn have been borne disproportionately by those residents with the fewest personal financial resources upon which to draw. In 2010 (the most recent year for which full data are available for both measures) the long-term unemployment rate for workers with only a high school degree was nearly three times as high as the rate for workers with a B.A. degree or higher.

This pattern also holds true for workers experiencing a period of unemployment lasting for more than one year. In the aftermath of the recession, Massachusetts workers with lower levels of education experienced much higher rates of unemployment that lasted over a year than did workers with college degrees or higher.

Looking at a broader measure of workers whose employment status has been negatively affected during the current downturn—a measure termed "underemployment"—it is similarly clear that rates among Massachusetts workers vary greatly by educational attainment. Once again, workers with low levels of education have been affected disproportionately, with underemployment rates almost four times higher than those for workers with college degrees or higher. Underemployed workers are defined as: 1) unemployed workers, 2) workers who are working part time because full time opportunities are not available, and 3) workers defined as only "marginally attached" to the labor force.3

2 Bauer, Paul et al. "State Growth Empirics: The Long-Run Determinants of State Income Growth" (Working Paper 06-06, May, 2006, Federal Reserve Bank of Cleveland).

3 The definition for "underemployed" used by the U.S. Census bureau includes all three of these groups of workers. The figures presented in this chart for Massachusetts, however, do NOT include those workers defined as "marginally attached". The sample size in Massachusetts was too small for the Census to provide accurate estimates of the number of workers in the marginally attached group and hence this group is omitted from the totals.


Wages account for the majority of income received by most workers—on average, about two-thirds of income is derived from wages, but this proportion is generally much higher for middle- and low-wage earners who do not receive income from capital gains or other investments.4 Because of the importance of wages to working families, looking at wage levels and growth in real wages (i.e., wages that have been adjusted for inflation) over time is one way to measure the well-being of workers in Massachusetts. When we do so we see a picture similar to that of other aspects of the state's economy—that is, while Massachusetts is generally doing better than the rest of the country, not all residents of the state have shared equally in the gains of recent decades. Lack of substantial real growth among lower-wage workers, along with more rapid growth at the high end, has widened the wage gap in Massachusetts.

In 2011, the average median hourly wage in Massachusetts was $19.81 per hour, among the highest in the country, and 23 percent higher than the U.S. average median wage of $16.06 per hour.

The relatively high median wage level in Massachusetts follows three decades during which the median wage grew faster in this state than in the country as a whole. In 1979 the median hourly wage in the state was $15.99 per hour in inflation-adjusted (2011) dollars ($5.16 in nominal dollars), slightly lower than the inflation-adjusted U.S. average of $16.33 per hour ($5.27 nominal). By 2009 the Massachusetts median hourly wage had risen to nearly $21.00 hour and, despite a decline over the past two years, it remains almost a quarter higher than it was in 1979 and 23 percent higher than the U.S. median wage, which has seen a slight decline in value.

But while median wage earners may be better off in Massachusetts compared to other parts of the country, the experience among workers within Massachusetts has varied in recent years. During the 1980s, real hourly wages grew at roughly the same rate (about 1.8 percent a year) for workers at the 20th, 50th, and 80th percentiles. In the 1990s, however, real wages began to grow at very different rates for these groups. Growth in real wages for the top 20 percent of workers was robust, especially in the 2000s. Meanwhile, wages for the middle group grew at less than half the rate the higher earners experienced and wage earners at the 20th percentile saw a decline in real wages during most of the last two decades.

The recent recession led to a drop in wages for all earning groups in Massachusetts, but the gap between hourly wages for high and low-income workers remains wide. Between 1984 and 2011 (in each case two years after the official end of a recession), this gap grew by nearly $9.00, adjusted for inflation. While all wage earning groups saw declines in hourly wages over the last few years, the drop for lower-wage earners has been steeper, and thus the gap has continued to grow. If the trend of unequal wage growth resumes as the economy and wages begin to grow again, this gap will widen further.

Of course, inequality in wages and income has increased across the U.S., but the gap between high and low earners has widened more in Massachusetts. The figure below reflects the growth in hourly wages for the 20th and 80th percentiles of earners in both Massachusetts and across the U.S., and shows that the two groups have pulled further apart in Massachusetts than in the country as a whole. The Massachusetts gap narrowed somewhat during the recession, but remains wider than the U.S. average. The increase in wage inequality is largely the result of a lack of growth in wages at the bottom and suggests that many Massachusetts workers have not shared in the benefits of the state's economic growth in recent decades.

4 For a breakdown of sources of income for different income groups, see Mishel, Lawrence et al, The State of Working America 2008-2009 (Cornell University, 2009), p. 82.

The Wage-Education Link

While a variety of factors affect wage levels, one likely explanation for Massachusetts' relatively higher median wage and stronger growth in recent decades is the higher level of education among the state's workforce. Recent academic work has pointed to a connection between education and income and suggested that a state's high school and college attainment rates is an important factor in explaining its per capita income growth relative to other states over time.5

This work is consistent with a change in the correlation between state median wage and educational attainment levels that can be seen between the years 1979 and 2011 As the next two figures show, there was a much weaker correlation between these two variables in 1979, while in 2011 higher educational attainment is more strongly linked to higher wages on the state level. The change is likely part of larger structural changes in the labor market—thirty years ago it was easier to find manufacturing jobs that paid relatively well even for people with lower levels of education; today many of these jobs have disappeared and higher-paying jobs tend to require a college degree.

The charts below illustrate the changing relation between education and wages. Each diamond refers to a single state, and the slope of the trend line shows how strongly workforce education is correlated with median wages across states.

In Massachusetts, the share of workers with a Bachelor's degree or higher has more than doubled over the past thirty years (see first figure, below). In fact, as the second figure below shows, Massachusetts now has the highest proportion of workers with a Bachelor's degree or higher of all U.S. states. Given the correlation between a state's educational attainment level and its median wage, it is not surprising that Massachusetts has high wage and income levels compared to many other states (see separate discussion of Massachusetts income in the following section).

In 2011, the median hourly wage for a Massachusetts worker with a Bachelor's degree or higher was more than twice as much as that for a worker with a high school degree (see first figure, below). This higher level follows three decades during which real hourly wages for college-educated workers grew at an annual rate nearly one percentage point higher than inflation, while wages for those with some college education have remained flat and workers with high school degrees saw a slight drop in real median wages (see second figure below).

5 Bauer, Paul et al. "State Growth Empirics: The Long-Run Determinants of State Income Growth" (Working Paper 06-06, May, 2006, Federal Reserve Bank of Cleveland).

The Minimum Wage

The preceding discussion suggests that a high-quality educational system that ensures broad access to college and university training is an important tool for helping workers move to higher-paying jobs. Indeed, it is likely that our highly educated workforce helps attract high-wage employers to Massachusetts. It is also the case, however, that some jobs do not require a higher degree. In the past, federal and state requirements for a minimum wage have helped ensure adequate salaries for such jobs. In Massachusetts close to 130,000 workers earn hourly wages that are at or near the minimum wage of $8.00 (this total includes workers earning up to $8.25 per hour).

A majority of workers earning at or near the minimum wage—63 percent—are 20 years or older and already have a high school degree. Two-thirds of minimum wage earners work at least 20 hours per week, and one-third work full time. Minimum wage workers are concentrated in a few industries—in fact 60 percent work in either the Retail or Leisure and Hospitality industries.

Industry sectors with high concentrations of minimum and low-wage workers have experienced more robust job growth in recent years than have sectors with low concentrations of such workers, and have recovered more quickly from recessions. The growth has occurred despite three sets of increases to the minimum wage that were implemented during this period (in 1996-1997, 2000-01, and 2007-08). A recent report that looks at national data on low-wage workers shows that the majority of minimum and low-wage workers are employed by large businesses, most of which have recovered from the financial crisis and have higher revenues than before the recession.6

Minimum wage standards tend to benefit other low-income workers who earn wages above the minimum wage. This is so because labor contracts may tie wages to the value of the minimum wage or the overall wage structure of an employer may shift upward when the minimum wage increases. This phenomenon can be seen in the last decade in Massachusetts. Increases in the required state minimum wage from $5.25 per hour to $6.00 per hour in 2000 and $6.75 in 2001 ($7.84 and $8.57 in real 2011 dollars) were followed by increases in the average real hourly wage for workers at the 20th percentile in subsequent years (from $11.33 to $11.56 in 2001 and to $12.07 in 2002); a similar pattern followed statutory increases in 2007 and 2008.

Over the past forty years the value of the minimum wage has been eroded by inflation. A minimum wage earner working full time in Massachusetts will earn about $16,000 this year, higher than the poverty threshold for a an individual (a little over $11,700 in 2011), but lower than the threshold for a family of three (around $18,000). That same worker would have earned about $21,040 back in 1968 (measured in real, inflation-adjusted dollars).

6 See "Big Business, Corporate Profits, and the Minimum Wage", National Employment Law Project Data Brief, July, 2012.


As noted in the preceding discussion of wages, for the majority of lower and middle income workers, hourly wages are the principal source of these workers' annual household income. While wage data allows us to gauge how well workers are being compensated on an hourly basis, we must turn to data on income in order to understand how workers are faring in terms of changes in their total annual purchasing power. At both the national and state levels, the data show that incomes for average working households have not fared well during the current economic downturn and weak recovery.

Data from the U.S. Census Bureau's annual American Community Survey (ACS) show that median household income in Massachusetts fell to $62,072 in 2010. This is a statistically significant decline of $3,067 or 4.7 percent from the 2009 level of $65,139 (adjusted to 2010 dollars). For the U.S. as a whole, the ACS data show median household income stood at $50,046 in 2010, a statistically significant decline of $1,004 or 2.0 percent from the 2009 level. Since 2007 (in the final month of which, the nation officially fell into recession), median household income in Massachusetts dropped an inflation-adjusted $3,509 or 5.4 percent. During the same period, U.S. median household income fell $3,310 or 6.2 percent. [For a discussion of the differences between the American Community Survey and the Current Population Survey, please see the "Methodological Notes" section at the end of this report.]

Despite the drop, Massachusetts's 2010 median household income continued to compare favorably to those of other states, remaining among the highest in the nation. As discussed elsewhere in this report, hourly wages (and therefore the resulting annual incomes which we are discussing here) are directly linked to educational attainment, an area in which the Massachusetts labor force has a substantial advantage over other states. The comparative strength of Massachusetts' median incomes (relative to median incomes in other states) is closely related to the high levels of education seen in Massachusetts' workforce.

Not only is median household income higher in Massachusetts than in most other states, over the past three decades Massachusetts has seen among the highest percent increases in median household income in the nation. Between 1980 and 2010, median household income in Massachusetts increased by more than 26 percent (adjusted for inflation). By comparison, the national median household income increased 11 percent during this period.

As is true of hourly wages (discussed earlier in this report), median household annual income for Massachusetts began to diverge markedly from the U.S. average back in the late 1970s and early 1980s. In 1980, the Massachusetts median household income was $48,800, which was modestly higher than the national median household income of $44,500 (1980 figures are adjusted for inflation to 2010 dollars). Since then, median income in Massachusetts has grown at an average annual rate of close to 0.8 percent, while for the U.S. median income has grown less than half as quickly, at a rate slightly below 0.4 percent. As a result, by 2010 the Massachusetts median household income of $61,300 was $11,800 (or nearly 25 percent) higher than the national median household income of $49,500.7

The data showing an increase in median household income in Massachusetts over the last three decades, however, do not reveal important differences among households at different income levels. In 1980, the difference in income between Massachusetts households with high incomes (families in the 90th percentile of all incomes) and those with lower incomes was much smaller than it is today. Thirty years ago, households in the 90th percentile (i.e., those families whose incomes were higher than 89 percent of other Massachusetts families) had median incomes about five times larger than the median incomes of households in the 20th percentile (i.e., those families whose incomes were lower than 80 percent of other Massachusetts families). By 2010, these higher income households had median incomes that were eight times higher than the median incomes of households in the 20th percentile .8

Looking at the differences in the rate of growth of median household incomes tells the same story in clearer terms. From 1980-2010, low-income Massachusetts households (those in the 20th percentile) saw an average income gain of just 0.3 percent each year over that full thirty-year span, adjusted for inflation.9 Households in the 50th percentile (i.e., those households in the very middle of the income distribution, with half of all households making more and half making less than they do) saw their incomes rise by 0.8 percent each year over the same period.10 By contrast, households in the 90th percentile saw their inflation-adjusted incomes grow by 1.7 percent each year, or nearly six times faster than the rate of income growth experienced by low-income Massachusetts households.

7 The median household income figures presented here are derived from U.S. Census Current Population Survey data and therefore differ somewhat from the median household income figures provided in the preceding discussion, which drew upon American Community Survey data. For a more detailed discussion of the differences between these two data sets, please see the "Methodological Notes" section at the end of this report.

8 Ibid

9 Specifically, this statistic compares the incomes of those households that were in the 20th percentile in 1980 with the incomes of those households that were in the 20th percentile in 2010 (adjusted for inflation). It does not track the same set of households over that 30-year period. It therefore is not necessarily the case that the same households were included in that 20th percentile in both years.

10 The median household income figures under analysis here are derived from U.S. Census Current Population Survey data and therefore differ somewhat from the median household income figures provided in the earlier income discussion, which drew upon American Community Survey data. For a more detailed discussion of the differences between these two data sets, please see the "Methodological Notes" section at the end of this report.


The poverty rate measures the percentage of people who live in households with cash income below a certain level, known as the poverty threshold (cash income includes wages, social security, and unemployment benefits, but excludes non-cash benefits, such as food stamps, and the effect of tax payments and credits received.) For 2010, the poverty threshold was about $11,300 for a single person under the age of 65 and a little over $22,000 for a family of four (thresholds for individuals and couples over the age of 65 are somewhat lower).

In 2010, Massachusetts was one of ten states with the lowest poverty rates in the country. The rate was 11.4 percent, almost four percentage points lower than the U.S. average of 15.3 percent (for a discussion of how this would change under the census bureau's new supplemental poverty measure, see Appendix A).

While Massachusetts compares favorably with much of the rest of the U.S., we saw in preceding sections that wages and income have either stagnated or dropped at the lower end of the economic spectrum and thus it is not surprising that poverty has increased here. After remaining relatively stable for four years, the poverty rate in Massachusetts showed a statistically significant increase last year, growing from 10.3 percent in 2009 to 11.4 percent in 2010. This means that between 700,000 and 750,000 people had incomes below the official poverty threshold in Massachusetts in 2010. Around 38 percent of people in poverty who are 16 years or older worked either full or part time in 2010.

The poverty rate among children also showed a statistically significant increase in 2010, rising to 14.3 percent from 13.1 percent in 2009. This means that in 2010 a little more than 200,000 Massachusetts children lived in families with incomes below the poverty line. The childhood poverty rate in Massachusetts has risen by nearly two percentage points from its pre-recession level of 12.4 percent in 2006.

Nationally, the rate of deep poverty (defined as people with incomes at or below 50 percent of the poverty line) is at its highest point since the Census Bureau began to measure it. In Massachusetts the rate of deep poverty has also risen, from 4.0 percent in 2000 to 5.3 percent in 2010. This translates into some 325,000 to 350,000 Massachusetts residents living on incomes of less than about $5,600 for a single non-elderly person or about $11,000 for a family of four.

Appendix A: Poverty and the Role of Social Programs

The Supplemental Poverty Measure

The analysis of poverty above is based on the official poverty measure that is used by the U.S. Census Bureau. Because it counts only cash income and does not take into account the effect of income transfer programs, such as the Supplemental Nutrition Assistance Program (SNAP, formerly known as Food Stamps), the official poverty measure may overstate the level of poverty by counting people who would fall into poverty without these programs. At the same time, the official poverty measure may also understate poverty levels because it uses a threshold that is based on the cost of food in 1963 (multiplied by three) and does not account for housing or other costs. It also does not take into account geographical variation in expenses, and does reflect costs related to employment, such as child care or transportation costs, or out-of-pocket expenses for medical care. Likewise, the official poverty measure does not include the effect of income, payroll, and other tax payments or the effect of tax credits, such as the Earned Income Tax Credit (EITC) received by families.11

Recently the Census Bureau has begun to publish a Supplemental Poverty Measure (SPM) based on a set of calculations that incorporates the elements discussed above. The SPM calculates income using a wider definition of resources that includes the EITC and SNAP, as well as other resources such as housing subsidies, reflects tax payments and credits, and subtracts out-of-pocket medical and other costs. It measures income against a threshold that uses current data on consumer expenditures on food, clothing, shelter and utilities, and varies depending on a household's housing status. Under this measure, the U.S. poverty rate is somewhat higher than under the official measure, 16.0 percent compared to 15.3 percent. The demographic composition of those in poverty also shifts—the number of children in poverty drops from 21.6 percent to 18.2 percent, while poverty rates among adults and the elderly rise.

The Role of Social Programs

The Census bureau has provided some data showing how the poverty rate changes when certain components of the SPM calculation are excluded. This information allows us to see more clearly the role of social programs in reducing poverty. The table below shows the number of people in poverty throughout the U.S., as calculated using the SPM, and then shows how these numbers would change if EITC and SNAP benefits were not included in the calculation (as is the case with the official poverty measure). As the table shows, when the EITC is excluded around 6 million additional people, just over half of them children, are in poverty; similarly about 5 million people, just under half of them children, are in poverty if SNAP benefits are excluded. To put it another way, the EITC lifts 6 million people out of poverty, and SNAP raises 5 million people above the poverty threshold. As the table also shows, counting payroll taxes and medical expenses adds to the number of people counted as being in poverty.

The Census Bureau has not calculated the SPM on a state-by-state level, so we do not know exactly how using this measure would change the official estimate of the Massachusetts poverty rate described above. To the extent that Massachusetts has higher costs, especially high housing costs, the SPM calculation would likely yield a higher poverty rate, while the existence of relatively strong safety net programs in Massachusetts would tend to lower it. We do know that nearly 378,000 taxpayers claimed the federal EITC credit last year. If the ratio of people lifted out of poverty by the EITC to the number of EITC filers on the national level (about 23 percent) holds in Massachusetts, close to 90,000 people in Massachusetts would have avoided poverty thanks to the EITC. Accounting for additional benefits from the Massachusetts State EITC would likely raise this number further. Likewise, Census data show that between 281,000 and 296,000 households received SNAP benefits in 2010; again if the federal ratio between households receiving SNAP benefits and people raised above poverty level held, then SNAP benefits help move 107,000 people out of poverty.

11 For a description of the official poverty measure and a description of differences between it and the new Supplemental Poverty Measure, see "The Research Supplemental Poverty Measure: 2010 (U.S. Census Bureau, November 2011).

Appendix B: Methodological Notes

CPS vs. ACS Data

Data on income, wages and poverty (among other information) are collected by the U.S. Census Bureau using two different annual surveys, the Current Population Survey (CPS) and the American Community Survey. Due to the large number of households that are surveyed annually by the U.S. Census Bureau for the ACS, the ACS is understood to provide the most accurate data on state-level measures such as median household income. The Census Bureau, however, first began collecting data using the ACS in 2000, meaning that ACS data exist only for the last ten years. To examine changes over longer periods of time, the best data available come from the Census Bureau's Current Population Survey (CPS) which provides data running back to the 1940s.

Because the data from the ACS and the CPS are collected using different methodologies, there are differences in the results from the two surveys; estimates of various measures (such as median household income or poverty rate, for example) may not agree perfectly between the two surveys. These discrepancies can be observed in some of the charts and discussion presented in this report.

Like most researchers, for recent state-level data we rely on the ACS, while for analyses of historical trends, we rely on data from the CPS.

Statistical Significance

Because the CPS and ACS data are based on surveys of a random sample of individuals in the population being measured, one cannot be certain that the estimate produced by the sample reflects the actual rate for the entire population. For each number it publishes, the Census Bureau therefore provides a margin of error using a 90 percent confidence level that defines the range within which its estimate falls. A 90 percent confidence level means that there is a 90 percent likelihood that the true population measure falls within the margin of error or range established around the estimate.

When two estimates (for example, a single measure taken in two different years) are compared, the margin of error can be used to calculate whether the two measurements reflect a real change in the population rather than a difference occurring by chance because of variation in the samples. When we say that two estimates are not statistically significant we mean that we cannot be certain that the estimates—even though numerically different—provide an accurate measure of a real difference in the true population. In most cases, larger sample sizes allow for greater certainty concerning estimates even for relatively small year-to-year changes.