The State of Working Connecticut 2011: Wages, Job Sector Changes, and the Great Recession

Transcription

1 The State of Working Connecticut 2011: Wages, Job Sector Changes, and the Great Recession Sarah Esty Orlando Rodriguez, M.A. December 2011 Produced with the generous support of the Melville Charitable Trust, the Stoneman Family Foundation and the staff of the Economic Policy Institute.

2 The State of Working Connecticut 2011 Wages, Job Sector Changes, and the Great Recession I. Introduction...2 II. Statewide Wage Trends A. Overview of Wages and Earnings...3 B. Trends for High and Low Wage Workers...6 C. Wage Inequality...8 D. Wages by Sector E. Hours Worked by Sector III. Wages and Demographics A. Gender Wage Gap B. Minority Wage Gap C. Value of Education IV. Labor Market Area Breakdown A. Cross-LMA Comparison B. Bridgeport-Stamford C. Danbury D. Enfield E. Hartford F. New Haven G. Norwich-New London H. Torrington I. Waterbury J. Willimantic-Danielson V. Conclusion and Policy Recommendations VI. Appendix A Note on Terminology Connecticut Voices for Children 1

3 I. Introduction The Great Recession, which extended from March of 2008 through January 2010 in Connecticut, reshaped our economic landscape. 1 The effects of the recession and extent of post-recession recovery vary greatly among individuals, employment sectors, and regions of the state. Our previous report, The State of Working Connecticut, 2011 Jobs, Unemployment, and the Great Recession, addressed the way in which the Recession impacted employment levels across various regional, industrial, and demographic groups. This report focuses on wage trends, considering how the recession and recovery reshaped compensation for different groups of workers in Connecticut. Broadly, the last five years (2006 to 2010) including the Great Recession and its aftermath have resulted in gains for those who were doing better than most, but stagnation or decline for many of the rest. Wages rose most for the top percentiles of wage earners, those with college degrees, men, whites, and those in high-wage manufacturing, education, and healthcare jobs. These improvements were not enjoyed by the many workers without a college degree, those in low-wage jobs in the retail and food services sectors, blacks, and those in the poorest regions of the state, for whom wages increased little or even fell from 2006 to Disproportionate wage gains for already advantaged workers and a lack of similar gains for the disadvantaged have compounded disparities in un- and under-employment along racial, gender, regional, and educational lines. Inequality increased in nearly all categories from 2006 to 2010; the gaps widened between wages for men and women, whites and blacks, and those with more and less education. The closing Hispanic wage gap bucked this trend, but this change was in part due to growing Hispanic unemployment and job losses for Hispanics in lower-wage jobs that artificially increased median wages. Finance and insurance sector wages declined most during the recession, causing the two major economic regions of Fairfield County (the Bridgeport-Stamford and Danbury Labor Market Areas) to face similarly large reductions in average wages. However, the finance sector, and Bridgeport-Stamford and Danbury along with it, has rebounded strongly and seems poised to make a full recovery. Wages in other sectors and regions that declined from 2006 to 2009 continue to fall or have regained only a fraction of their losses, suggesting that the post-recession wage recovery has been and may well remain very uneven. This report contains four major sections. First, we discuss statewide trends in wages and earnings, considering in particular those changes occurring with different employment sectors, geographies, and wage levels. Next, we look in-depth at changes within and between subgroups of the population, with a focus on disparities by gender, race and ethnicity, and level of educational attainment. Recognizing the significant differences between regions of the state in numerous areas, including size of certain industrial sectors and cost of living, the third section of the report considers wage changes within the six largest employment sectors in each of Connecticut s nine Labor Market Areas (LMAs). 2 The final section of the report offers broad public policy solutions. It recommends policy makers focus on education and careful economic planning if they hope to reduce inequalities and promote wage growth for all workers. 1 The Connecticut Department of Labor reports the official period of recession in CT ran from March 2008 to January 2010 (Jungmin Charles Joo, Connecticut Recovery Began in January 2010, The Connecticut Economic Digest, Vol. 16 No. 3: March 2011, available at For the purposes of this report, recession will be taken to mean this period unless otherwise noted. 2 The United States and Connecticut Departments of Labor establish the borders of the LMAs. An LMA is defined as an economically integrated geographic area within which individuals can reside and find employment within a reasonable distance or can readily change employment without changing their place of residence. ( Labor Market Areas, 2011, U.S. Department of Labor Bureau of Labor Statistics, March 2011, pg iii, available at: Connecticut Voices for Children 2

4 II. Statewide Wage Trends From 2006 through 2010, median weekly and median hourly wages increased statewide. This, however, has not necessarily translated into increased income for most Connecticut residents, as the rise in wages ran parallel to rising unemployment among lower-wage workers, which likely had the effect of increasing median wages estimates. Average annual earnings peaked before the recession in 2007 then plummeted in They have slowly rebounded in the last two years, but still have not returned to 2007 levels. Furthermore, recent changes in wages have been uneven, with drastically different magnitudes and directions based on wage level, industrial sector, type of compensation, and geographic area. For years, Connecticut s strong middle class helped keep down wage inequality levels. However, rapidly rising wages for the highest paid workers particularly with the post-recession recovery of the financial sector produced a sharp rise in inequality in A. Overview of Wages and Earnings Both median hourly and weekly wages in Connecticut rose through the Great Recession, gaining consistently from 2006 through 2009, but falling slightly from 2009 to 2010 (Figure II-1) 3. Connecticut s median weekly and hourly wages remained significantly above both our peer state average (MA, NY, NJ, RI) 4 and the national average throughout the last five years. Connecticut increased this advantage over the course of the recession; other states experienced declining median wages from 2006 to 2008 while median wages continued to grow in Connecticut over the same period. Median hourly and weekly wages followed parallel trends over the last five years in Connecticut, our peer states, and the country as a whole. This suggests that changes in the hourly wage have been the primary drivers of changes in weekly earnings, as opposed to changes in hours worked or overtime payment. 3 This report uses several sources of wage and employment data which are roughly comparable, but are collected from different populations, using different methods, and with slightly different definitions of wages and earnings. Throughout this report, CPS is used to refer to the U.S. Census Bureau Current Population Survey yearly data, available through the American FactFinder 2.0 tool ( as well as the U.S. Department of Labor Bureau of Labor Statistics website for employment and wage data ( The QCEW is the Quarterly Census of Employment and Wages. Reports with quarterly and annual QCEW data for Connecticut and individual Labor Market Areas are available from the Connecticut Department of Labor ( See the Appendix A Note on Terminology for a further discussion of sources and definitions. 4 To help benchmark changes in our state s economy, this report draws comparisons on a number of metrics between Connecticut and the average value for our peer states. We selected Massachusetts, New York, New Jersey, and Rhode Island for this peer state group because of their geographic proximity, similarities in socioeconomics, and similarities to Connecticut s economic structure. Connecticut Voices for Children 3

5 Hourly Wage Weekly Wage $22 Median Hourly and Weekly Wages $981 $976 $1,000 $21 $20.70 $950 $20 $889 $19.97 $900 $19 $18 $19.20 $821 $864 $861 $18.56 $18.38 $850 $17 $17.93 $800 $16 $16.02 $16.21 $16.00 $751 $747 $750 $15 $ CT - Hourly Wage Peer States - Hourly Wage US - Hourly Wage CT - Weekly Wage Peer States - Weekly Wage US - Weekly Wage Source: CT Voices and Economic Policy Institute analysis of CPS data. Dollar values are inflation adjusted to 2010 dollars using the CPI-U-RS/BLS CPI Calculator. Figure II-1 While hourly and weekly median wages have risen, this has not translated into an increase in income for all residents. In fact, median weekly wages rose in parallel with unemployment, suggesting that the increase in median wages may be the result of the disappearance of lower income jobs rather than a rise in wages for all workers or growth in middle- or high-wage sectors (Figure II-2). $700 Connecticut Voices for Children 4

6 $1,000 $980 $960 $940 $920 Connecticut Unemployment and Wages % 10% 8% 6% $900 $880 $860 Median weekly wages Unemployment $840 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Source: CT Voices analysis of CT Department of Labor Local Area Unemployment Statistics and CPS data. Dollar values are inflation adjusted to 2010 dollars using the BLS CPI Calculator. Figure II-2 Trends in average annual earnings for people in Connecticut support the conclusion that rising median wages do not reflect increases in income for most people. Average annual earnings peaked in 2007, before the onset of the recession and the start of the rise in unemployment in 2008, and declined significantly from 2007 to 2009 as unemployment rose (Figure II-3). Annual earnings began to rebound in 2010, returning to just above 2006 levels, slightly preceding the declines in unemployment over the course of The movement in average annual earnings better captures the true declines in income for most workers, declines which are not captured by apparently increases in median weekly wages that are driven by shrinking employment in low-wage jobs. Falling average annual earnings over a period of rising median hourly and weekly wages also illustrates the impact of the financial sector s decline during the Great Recession; financial sector compensation comes largely in the form of bonuses and stock options, which are not captured in the weekly wages but do appear in the annual earnings calculations. Furthermore, compensation in the finance sector is significantly higher than other sectors, causing it to disproportionately impact state average earnings without similarly driving changes to the median. 4% 2% 0% Connecticut Voices for Children 5

7 $70,000 $68,000 $66,000 $64,000 $62,000 $60,000 Connecticut Average Annual Earnings and Unemployment $59,288 $61,027 10% 9% 8% 7% 6% $58,000 $56,000 $54,000 $59,142 $58,718 $52,000 Average Annual Earnings Unemployment $50,000 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 $59,463 Source: CT Voices analysis of CT Department of Labor Local Area Unemployment Statistics and QCEW data. Dollar values are inflation adjusted to 2010 dollars using the BLS CPI Calculator. Figure II-3 B. Trends for High and Low Wage Workers Through the recession, wages increased most for the highest-paid workers and least for the lowest paid workers (Figure II-4), widening already significant gaps between the richest and poorest Connecticut residents. Median wages (50 th percentile 5 ) rose 4 percent (adjusted for inflation) between 2006 and 2010, less than one-third the rate of growth of very high wages. Real wages increased by 14.4 percent for very high wage workers, but only 1.3 percent for very low wage workers, who saw the slowest wage growth of all workers. 5% 4% 3% 2% 5 Percentiles are the values below which a certain percentage of observations fall. For example, the 10 th percentile wage is the wage at which 10% of workers earn less and 90% earn more (not the average wage for the bottom 10%). In this report, we define very low wage workers as those at the 10th percentile, low wage workers as those at the 20th percentile, median wage workers as those at the 50th percentile, high wage workers as those at the 80th percentile, and very high wage workers as those at the 90th percentile. Connecticut Voices for Children 6

8 16% 14% Percent Change in Wages from 2006 to % 12% 10% 9.2% 9.7% 8% 7.6% 6% 4% 4.4% 4.6% 4.0% 2% 1.3% 1.6% 0% 10th Percentile (Very Low) 20th (Low) 30th 40th 50th 60th 70th 80th (High) 90th Percentile (Very High) Source: CT Voices and Economic Policy Institute analysis of CPS data. Dollar values are inflation adjusted to 2010 dollars using the CPI-U-RS.. Figure II-4 Hourly wages have grown consistently and significantly over the last five years for the highest paid workers. Wages stagnated for middle- and low-wage workers from 2006 to 2009 and fell from 2009 to 2010 (Figure II-5). Wages have increased most for very high (90 th percentile) wage earners, who made approximately six dollars per hour more in 2010 than they did in High wage earners (80 th percentile) wages grew by about three dollars per hour over the same time period. In contrast, very low wage workers (10 th percentile) wages increased by only about eleven cents per hour, and experienced a twenty cents per hour decline from 2009 to Median wages grew by about eighty cents per hour from 2006 to These small real increases 6 in wages for middle and low income workers, while keeping up with inflation, are minor compared to wage increases for the highest paid workers. Furthermore, most workers saw a decline in real wages from 2009 to 2010 except for the very highest paid, who saw continued growth in their already high wages. 6 Real wages are inflation adjusted, as opposed to nominal wages, which are in dollars for the year in which they were earned. Connecticut Voices for Children 7

9 $50 $45 $40 $35 $30 $25 $20 $15 $10 $5 $41.79 $32.34 $19.20 $47.81 $35.66 $35.47 $20.70 $19.97 $10.96 $11.28 $11.13 $8.79 $9.10 $8.90 $ Figure II-5 C. Wage Inequality Hourly Wages by Percentile Source: CT Voices and Economic Policy Institute analysis of CPS data. Dollar values are inflation adjusted to 2010 dollars using the CPI-U-RS. 90th percentile (Very High) 80th percentile (High) 50th percentile (Median) 20th percentile (Low) 10th percentile (Very Low) Highest wage year ( ) Wage inequality between very high and very low wage earners has risen significantly since 2006 in Connecticut, surpassing the national and our peer state average (Figure II-6). In 2006, very high wage earners made approximately 4.75 times the wages of very low wage earners, reflecting greater inequality than the national average but slightly less inequality than our peer states. However, by 2010, Connecticut s ratio of very high to very low wages had risen to 5.37, higher than both the national average of 4.71 and our peer states average of Connecticut s rate of increase in wage inequality was greater from 2009 to 2010 than during the prior five years and was likely driven by the significant increase in wages in the highly paid finance sector. Connecticut Voices for Children 8

10 Ratio of 90th Percentile to 10th Percentile Weekly Wages 5.6 Inequality Between Very High and Very Low Wage Earners Connecticut US Peer State Average Source: CT Voices and Economic Policy Institute analysis of CPS data. Dollar values are adjusted to 2010 dollars using the CPI-U-RS. Figure II-6 From 2009 to 2010, inequality also increased rapidly between very high and median wage earners in Connecticut, vaulting us above our peer states and matching the national average (Figure II-7). For years, Connecticut s strong middle class helped keep wage inequality levels down. However, rapidly rising wages for the highest paid workers particularly with the post-recession recovery of the financial sector produced a sharp rise in inequality in Connecticut Voices for Children 9

11 Ratio of 90th Percentile to 50th Percentile Weekly Wages 2.50 Inequality Between Very High and Median Wage Earners Peer State Average Source: CT Voices and Economic Policy Institute analysis of CPS data. Dollar values are adjusted to 2010 dollars using the CPI-U-RS. Figure II-7 Connecticut US D. Wages by Sector Wages vary tremendously by sector (Figure II-8). Average weekly wages in the financial and insurance sector were nearly double those in manufacturing, the next highest-paid major sector, and eight times greater than those in the lowest-paid accommodation and food services sector. Real wages in most of Connecticut s largest employment sectors have stagnated over the last five years. Weekly wages in the low-paying sectors like retail trade and accommodation and food services have declined slightly from 2006 to 2010 (3 percent and 1 percent respectively). Workers in better-paying areas, such as health care, public K-12 education, and manufacturing, have seen minor increases in wages since 2006 (between 2 percent and 4 percent growth from 2006 to 2010). The only significant fluctuations occurred in the finance and insurance sector, where wages declined 14 percent from 2007 to 2009, then rebounded from 2009 to 2010, growing 9 percent in one year. Connecticut Voices for Children 10

12 Average Weekly Wage for Largest Employers $3,000 $2,500 $2,816 $2,959 $2,782 Finance and insurance (7% of jobs in CT) Manufacturing (10%) $2,000 $1,500 $1,000 $500 $0 $1,372 $1,411 $1,174 $1,141 $1,144 $963 $998 $894 $913 $909 $613 $359 $ $593 $356 Source: CT Voices analysis of QCEW data. Dollar values are adjusted to 2010 dollars using the BLS CPI Calculator. Figure II-8 Local Gov't Educational services (6%) Health care and social assistance (15%) Retail trade (11%) Accommodation and food services (7%) Statewide Average Represents highest wage year ( ) Workers in higher-paid sectors experience large seasonal variation in earnings, reflecting the significant share of compensation coming from bonuses and other non-wage income (Figure II-9). Quarterly earnings data include bonuses and stock options that are not counted in hourly or weekly wages but are often significant portions of the yearly earnings of many workers in Connecticut. Non-wage income can contribute to much greater income inequality than is captured by gaps in hourly or weekly wages because non-wage income goes primarily to the highest paid workers. 7 For example, workers in the financial sector can earn yearly bonuses as high as three months wages, while those in lower-wage jobs usually do not receive any bonuses. Thus, the already significant gaps between weekly wages in various sectors actually under-represents the true inequality in yearly earnings. 7 See, for example, research by the Independent Budget Office of the City of New York, which found that, Though all types of income are concentrated at the top of the distribution, the concentration of non-wage income is especially large Almost the entire non-wage portion of AGI consists of dividends and interest, net business income (gains minus losses), or net realized capital gains (gains minus losses). Most filers have none of these non-wage types of income, and the top one percent of all filers received more than half of each: 58.3 percent of dividends and interest, 84.6 percent of business income, and 84.7 percent of capital gains. (Letter to Councilmember James Oddo, City of New York Independent Budget Office, December 6, 2011, available at: Connecticut Voices for Children 11

13 $60,000 Statewide Average Quarterly Earnings for Largest Employers $50,000 $40,000 $30,000 $20,000 $10,000 $0 All Industry Average Retail trade Health care and social assistance Manufacturing Finance and insurance Accommodation and food services Source: CT Voices analysis of QCEW data. Dollar values are inflation adjusted to 2nd Quarter 2011 dollars using the CPI-U-RS monthly data. Figure II-9 E. Hours worked by Sector In most sectors, weekly hours worked declined over the course of the recession but began to rebound in 2010, signaling the official end of the recession in Connecticut in 2010 (Figure II-10). Hours worked varied widely across employment sectors, with those in manufacturing working over 50 percent more hours than those in leisure and hospitality. While hourly wages vary significantly between sectors, differences in hours worked also contribute to differences in weekly and quarterly wages. For example, workers in the leisure and hospitality field work an average of 26.2 hours per week, 34 percent less than those in the manufacturing field, and also earn average weekly wages that are 75 percent lower. 8 Differences in hours worked have held steady throughout the recession, except in the financial sector, where hours worked steadily increased from 2007 to Overall, from 2007 to 2010, hours worked declined by 2 percent to 5 percent in all sectors except for finance. Coupled with declines, or stagnation, in hourly wages, the combination contributes to declines in earnings for most Connecticut workers. Change in hours worked parallels changes in yearly income, illustrating the impact of hours on total pay (Figure II-11). 8 Wage data come from the QCEW, while hours worked are from the CES, so the industrial categories are not identically defined (QCEW uses the category Accommodation and Food Services, while CES uses the category Leisure and Hospitality, but the workers contained within each are similar enough for a rough comparison). Connecticut Voices for Children 12

14 Average Weekly Hours Worked Average Annual Earnings Average Weekly Hours Worked (Private Sector) Statewide Average Manufacturing Construction Financial Activities Prof. and Bus. Services Trade, Trans., and Utilities Ed. and Health Services Leisure and Hospitality Source: CT Voices analysis of Current Employment Statistics (CES) data. Figure II Average Weekly Hours Worked and Annual Earnings in Connecticut, (Private Sector) $61,500 $61,000 $60, $60,000 $59,500 $59,000 $58,500 $58,000 $57, $57, Average Weekly Hours Worked Average Annual Earnings Source: CT Voices analysis of Current Employment Statistics (CES) and QCEW data. Dollar values are inflation adjusted to 2010 dollars using the BLS CPI Calculator. Figure II-11 Connecticut Voices for Children 13

15 What Others Report Economic Policy Institute (Ten Facts About the Recovery, July 2011) Persistent high unemployment is contributing to low wage growth for workers. Nominal hourly wages have grown 1.8% over the last year, below the growth rate at the end of the recession (2.6%) and half the rate from before the recession. Brookings (What is Happening to America s Children? A Look at the Widening Opportunity Gap for Today s Youth, Oct. 2011) Annual family earnings are an important measure of resources available to children, with opportunity implications that extend into adulthood. In terms of family earnings, the median child is no better off today than she was 35 years ago. Gains in family earnings have gone only to half the children disproportionately wealthiest within that group and half of America s children are worse off than their counterparts in Connecticut Voices for Children 14

16 III. Wages and Demographics The large wage disparities faced by women and blacks were narrowing prior to the recession, but increased sharply in 2008 and were still larger in 2010 than they were in Hispanics, though also experiencing an increase in wage disparity in 2008, saw gains in wage equality in 2010 and appear to have made progress since However, the narrowing of the wage gap by Hispanics may be a statistical artifact of large unemployment among the lowest paid Hispanic workers. Significant disparities still remain for women, blacks, and Hispanics, who earn, respectively, 76 percent, 67 percent, and 69 percent of the hourly median wage of men and whites. Connecticut s disparities are larger than the national average for blacks and women. Median wages vary dramatically by level of educational attainment, with those holding a bachelor s degree earning wages approximately twice as large as those with only a high school degree. Wages have increased most since 2006 for those with more education and those earning the highest wages within each level of education. Because of the strong impact of education on wages, racial and ethnic disparities in levels of educational attainment are important drivers of the racial wage gaps. A. Gender Wage Gap In 2010, the median weekly wage for women in Connecticut was 76 percent of that paid to men a percentage which is virtually unchanged from 2006, though it has fluctuated during that period (Figure III- 1). The wage gap had been narrowing prior to the onset of the recession, but increased dramatically from 2007 to From 2008 to 2010, the gap has begun to narrow again, but remains larger than before the recession. Movement towards gender wage parity can come from positive sources (rising wages and new high-paying jobs for women) and negative sources (falling wages for men, job losses for low-paid women), so analyzing the gap itself is insufficient to determine whether women are better or worse off than they were five years ago. To that end, we find that women s median weekly wages increased 9.4 percent from 2006 to 2010, while men s wages grew 11.3 percent. The positive direction of this trend in women s weekly wages is promising. However, the fact that women s wages grew less quickly than men s suggests that women are moving backwards in terms of obtaining wage parity with men. Connecticut Voices for Children 15

17 $763 $798 $782 $838 $835 $989 $1,004 $1,071 $1,117 $1,101 $1,300 $1,200 $1,100 $1,000 Median Weekly Wages by Gender 79% 77% 76% 75% 73% 80% 78% 76% 74% 72% 70% $900 68% 66% $800 64% 62% $700 60% Women's Wage Men's Wage Women's Wages as a % of Men's Source: CT Voices analysis of CPS data. Dollar values are inflation adjusted to 2010 dollars using the BLS CPI calculator. Figure III-1 Since 2006, changes in women s median weekly wages do not capture the true changes to their income because of the impact of shifts in the workforce. Progress towards wage equality appears inversely correlated with increases women s employment levels as the number of women in the workforce climbed, women s median wages fell and the wage gap grew. Conversely, as the number of women in the workforce declined, median wages rose and the gap narrowed (Figure III-2). These trends suggest that jobs women gained as they re-entered the workforce were lower-paying. Supporting this conclusion, national data indicate that coming out of the recession, women had lost jobs in higher-paying construction, utilities, information, and manufacturing sectors (where 2010 average annual earnings in Connecticut ranged from $58,600 to $103,500), to be replaced with jobs in lower-paying mining and logging and education and health services (where the average annual salaries were only $47,300 to $68,400). 9 9 Changes in employment by sector come from Rakesh Kochhar, In Two Years of Economic Recovery, Women Lost Jobs, Men Found Them Pew Research Center (July 6, 2011), (available at: Gender_FINAL_ pdf), 10. The Pew data shows percentage change in employment in that sector, which inflates the importance of changes in sectors which employ relatively few women (such as mining and logging or construction), but the trends in which sectors have gained and lost employment and general magnitude of those changes are still useful to consider. Data on average annual earnings in Connecticut in 2010 come from the 2010 QCEW. Connecticut Voices for Children 16

18 Percentage of Men's Wages Employment Median Weekly Wages and Employment by Gender 79% 80% 850,000 78% 76% 77% 75% 76% 800,000 74% 73% 750,000 72% 70% 68% 702, , , , , , ,000 66% 64% 62% 550, , , , , , ,000 60% ,000 Women's Employment Men's Employment Women's Wages as a % of Men's Source: CT Voices analysis of CPS data.. Figure III-2 Compared to the national average and our peer states, Connecticut s women make a considerably smaller percentage of men s wages and have fallen behind in narrowing the wage gap with men (Figure III-3). In 2010, Connecticut lagged behind the nation and our peer states, with women making only $0.76 on the dollar compared to men, versus $0.81 nationwide and $0.82 in our peer states. Differences in unemployment rates further complicate the understanding of the sources of Connecticut s high gender wage gap. While unemployment is higher among men than women, both nationwide and in Connecticut, the gender difference in unemployment is smaller in Connecticut than nationally, and fewer Connecticut workers are unemployed overall. In 2010, while 18 percent more Connecticut men were unemployed than women (unemployment rates of 9.9 percent and 8.4 percent respectively), nationally, 22 percent more men were unemployed than women (10.5 percent unemployment for men and 8.6 percent for women). 10 Combining the information in Figures III-1 and III-2, we see that the 2007 to 2009 decrease in employment for men in Connecticut was correlated with a rise in median wages for those who still had jobs, and a rise in employment for women from 2007 to 2008 led to a decline in median wages over that period. This suggests that higher unemployment for men and lower unemployment for women (a larger ratio of unemployment men to women) would result in a larger gap in median wages. Yet the larger percentage of unemployed men versus women nationally than in Connecticut did not result in larger a national wage gap than that of Connecticut quite the opposite. Connecticut s wage gap has been consistently larger than the national wage gap throughout the last five years. This suggests that other structural factors, such as the prevalence of high-wage male-dominated sectors in the state, including manufacturing and finance, may be more 10 CT Voices analysis of data from the 2010 Current Population Survey, see Matt Santacroce and Orlando Rodriguez, The State of Working Connecticut 2011: Jobs, Unemployment, and the Great Recession, Connecticut Voices for Children (Sept. 2011) (available at Connecticut Voices for Children 17

19 73% 75% 76% 77% 79% 79% 79% 80% 80% 80% 81% 81% 80% 81% 82% significant drivers of the long-term gap, while overall employment contributes more to short term fluctuations. 82% Women's Median Weekly Wages as a Percentage of Men's 80% 78% 76% 74% 72% 70% US CT Peer State Average Source: CT Voices analysis of CPS 2010 data. Figure III-3 What Others Report US Census Bureau (Education and Synthetic Work-Life Earnings Estimates, Sept. 2011) Men s median annual earnings are on average $12,741 greater than women s after accounting for differences in education level, race/ethnicity, age, citizenship, English proficiency, and geography. Bureau of Labor Statistics (Women's earnings and employment by industry, 2009, Feb. 2011) The gender wage gap varies significantly across sectors, with particularly large gaps in finance and manufacturing and small gaps in construction and agriculture. In sectors with the largest numbers of women, such as education and health services, the gap tends to be larger. US Government Accountability Office (Gender Pay Differences: Progress Made, but Women Remain Overrepresented among Low-Wage Workers, Oct. 2011) Young women are now receiving college degrees at higher rates than men, but women are still overrepresented in low-wage industries. Institute for Women s Policy Research (Separate and Not Equal? Gender Segregation in the Labor Market and the Gender Wage Gap, Sept. 2010) Occupational segregation by gender declined in the 1970s and 1980s, but progress has stalled since the 1990s and many industries remain dominated by a single gender. This is of concern because it is economically inefficient and is a major cause of the persistent gender wage gap. B. Minority Wage Gap Black, Hispanic, and white workers all earn higher median hourly wages in Connecticut than the United States as a whole (Figure III-4). White workers have higher median hourly wages than blacks and Hispanics both in Connecticut and nationwide. However, the gap between white and minority wages is particularly Connecticut Voices for Children 18

20 large in Connecticut because whites in Connecticut earn much more than whites nationwide, whereas blacks and to some extent Hispanics in Connecticut earn more, but not much more, than the national average. It is positive that median wages for blacks and Hispanics are higher in Connecticut than the national average. However, the cost of living is also significantly higher in Connecticut than nationally, such that these slightly higher wages do not necessarily lead to a better standard of living for minority workers in the state. 11 $25 $ Median Hourly Wages by Race and Ethnicity $21.55 $17.64 Connecticut United States $15 $14.39 $13.68 $14.96 $12.11 $10 $5 $0 White Non-Hispanic Black Hispanic Source: CT Voices and Economic Policy Institute analysis of CPS data. Dollar values are inflation adjusted to 2010 dollars using the CPI-U-RS. Figure III-4 Connecticut blacks earned only 67 percent the median hourly wages of whites a wage gap that is significantly larger the US average of 78 percent. In Connecticut (unlike the US as a whole) Hispanics earn median wages that are slightly higher than those earned by blacks. This leads Connecticut s Hispanic-white wage gap to be slightly smaller than the national average (see Figure III-5). Differences in educational attainment have been shown to be a major contributor to wage gaps for both Hispanics and blacks, 12 and Connecticut s Hispanic population is better educated than the national Hispanic population, while the state s black population is less well educated than the US average. 13 Section III-C delves further into this intersection of race, education, and wages. 11 Prices in Connecticut are approximately 22% greater than those nationally, and even higher in certain regions like the Bridgeport-Stamford-Norwalk metro area, where they are over 40% higher than the country overall. See Bettina H. Aten and Roger J. D Souza, Regional Price Parities: Comparing Price Level Differences Across Geographic Areas, US Dept. of Commerce Bureau of Economic Analysis (Nov. 2008) (available at: 12 Among men in Northeastern states, controlling for educational attainment, age, and English proficiency explains the nearly the entire wage gap between whites and blacks, Hispanics, and Asians. See, Dan Black, Amelia Haviland, Seth Sanders, and Lowell Taylor, Why Do Minority Men Earn Less? A Study of Wage Differentials Among the Highly Educated The Review of Economics and Statistics, February 2006, 88(1): (available at 13 See Figures III-16 and III-17 Connecticut Voices for Children 19

21 Ratio of Minorities' Median Hourly Wages to Whites' 77.6% Connecticut United States 66.8% 69.4% 68.7% Black Hispanic Source: CT Voices and Economic Policy Institute analysis of CPS 2010 data. Figure III-5 Nationally, the black and the Hispanic wage gaps have remained essentially unchanged over the last five years. In Connecticut, there has been more volatility, but with a clear positive trend for Hispanics and stagnation for blacks. Since 2006, the Hispanic wage gap has narrowed by 7.6 percentage points (Figure III- 6), while the black wage gap has widened by 0.6 percentage point (Figure III-7). Both groups saw increasing wage disparity as a result of the recession, and improving wage equality as the recession ended. 85% 80% 75% 70% 65% Hispanics' Median Hourly Wages as a Percentage of Whites' 69.8% 70.7% 70.8% 69.8% 68.7% 69.4% CT US 60% 55% 61.1% 61.1% 58.3% 60.4% 50% Source: CT Voices and Economic Policy Institute analysis of CPS data. Figure III-6 Connecticut Voices for Children 20

22 100% 95% 90% 85% 80% 75% 70% 65% 60% 55% 50% Blacks' Median Hourly Wages as a Percentage of Whites' 78.1% 67.4% 76.2% 76.1% 72.4% 62.3% 62.4% 77.6% 77.6% 66.8% Source: CT Voices and Economic Policy Institute analysis of CPS data. Figure III-7 An analysis of wage trends over the last five years reveals that Connecticut Hispanics have seen relatively consistent increases in median hourly wages, which have helped narrow the wage gap even as wages grew for whites. Blacks experienced increasing wages before the recession, but saw a significant decline from 2007 to 2008, and have not seen wages return to 2007 levels (Figure III-8). This relative stagnation in blacks wages, when combined with increasing wages for whites, has led to slight widening in the size of the blackwhite wage gap. By 2010, the wage gap for Hispanics was less than that of blacks. Rising median wages must also be considered in the context of changes in the labor force. These changes may reflect rising unemployment, rather than rising wages for most workers. Unemployed workers and the long-term unemployed who have become discouraged and left the labor force are not factored into median wage calculations. Therefore, if rising unemployment, or departure from the labor force, has disproportionately affected racial and ethnic minorities, changes to the median wage may reflect changes in the total size of the workforce (denominator) or the exodus of particular types of workers (if those losing jobs were concentrated in high- or low-wage sectors) rather than movement in wages for most workers. There is in fact high minority unemployment (17.7 percent for Hispanics and 15.6 percent for blacks versus 7.5 percent for whites in Connecticut in 2010). 14 Furthermore, unemployment rose disproportionately for minorities over the last five years, increasing by 7.3 percentage points for blacks and 9.5 for Hispanics, but only 4.2 percentage points for whites. Labor force participation rates 15 rose for whites and Hispanics (by 1 percent and 0.8 percent respectively), while it fell for blacks by 5.4 percent. 16 The increase in unemployment and decrease in labor force participation rates for blacks suggests that large numbers of black workers are out of work and many have stopped looking entirely. This means that changes to median wage and the black-white wage gap may have been driven significantly by changes in the composition of the labor force in addition to changes in wage level for workers who remained employed throughout this period. Supporting this conclusion, national research suggests that blacks have struggled in CT US 14 Orlando Rodriguez and Matt Santacroce, State of Working Connecticut 2011: Jobs, Unemployment, and the Great Recession, Connecticut Voices for Children (Sept. 2011) (available at: Labor force participation rate is the labor force (number of employed and unemployed persons looking for work) as a percentage of the civilian non-institutional population. See, Labor Force and Labor Force Participation Rate in Glossary, Bureau of Labor Statistics, (Feb. 2008) (available at; 16 State of Working Connecticut 2011: Jobs, Unemployment and the Great Recession Connecticut Voices for Children 21

23 $14.04 $12.73 $15.19 $12.82 $13.41 $12.56 $14.20 $13.73 $14.39 $14.96 $20.83 $20.98 $21.54 $22.74 $21.55 the recession due to high un- and under-employment as well as falling wages for those who are employed. 17 If mostly those employed in low wage occupations had lost jobs, median wages for blacks would have risen. Instead, the stagnation of median wages suggests that the jobs lost were more evenly spread across income levels and included many middle class jobs. 18 The significant rise in unemployment among Hispanics suggests that changes to the labor force also were an important cause of changes in median wages, and that the consistently rising median wage over the last five years may represent job losses amongst low wage workers rather than increases in wages for most workers. For example, the recession-driven contraction of the construction sector nationwide resulted in job losses for many Hispanics earning low wages, which likely contributed to the appearance of rising median wages. 19 $28 $26 $24 $22 $20 $18 Median Hourly Wages by Race 72.4% 69.4% 67.4% 62.3% 62.4% 66.8% 61.1% 61.1% 60.4% 58.3% 80% 70% 60% 50% 40% $16 $14 $12 $10 30% 20% 10% $ White Black Hispanic Blacks' Wages as a % of Whites' Hispanics' Wages as a % of Whites' Source: CT Voices and Economic Policy Institute analysis of CPS data. Dollar values are inflation adjusted to 2010 dollars using CPI-U-RS. Figure III-8 0% 17 Alexandra Cawthorne, Weathering the Storm: Black Men in the Recession, Center for American Progress (April 2009) (available at: 18 Recent research has highlighted the impact of public sector job losses and the disproportionate damage to black middle class working families. See, for example, Steven Pitts, Black Workers and the Public Sector, UC Berkeley Center for Labor Research and Education (April 2011), (available at 19 The Hispanic Labor Force in the Recovery, US Department of Labor (March 2011) (available at Connecticut Voices for Children 22

24 What Others Report US Department of Labor (The Hispanic Labor Force in the Recovery, March 2011) Nationally, labor participation rates are falling amongst all but the oldest Hispanics, and falling most for the young. Unemployment rates are highest for Hispanics with the least education and those years old. US Bureau of Labor Statistics (Labor Force Characteristics by Race and Ethnicity, 2010, August 2011) Black and Hispanic workers are disproportionately underrepresented in management, professional, and related occupations - the highest paying major job category. Earnings disparities exist across all major occupations for men, and are wider for women in some industries (management and professional, for example) but disappear for women in others (such as production, transportation, and material moving). C. Value of Education In 2010, those with a four-year college degree earned approximately twice as much per hour as those with less education. Those with a four-year college degree have also seen their median wages rise consistently since 2006 for a total increase of 8 percent, while wages fell by a net of 5 percent for those with only a high school education (Figure III-9). Median wages for those who attended some college, or obtained a two-year degree, are less than 10 percent higher than wages for those with only a high school diploma in $35 $30 Median Hourly Wages by Educational Attainment $29.29 $28.54 $29.94 $30.64 $30.76 $25 $20 $16.81 $16.96 $16.92 $15.96 $16.78 $15 $16.07 $15.89 $15.38 $15.98 $15.26 $10 High school $5 Some college Bachelor's or higher $ Source: CT Voices and Economic Policy Institute analysis of CPS data. Dollar values are inflation adjusted to 2010 dollars using CPI-U-RS. Figure III-9 Connecticut Voices for Children 23

25 $8.18 $13.13 $9.60 $10.07 $18.15 $11.15 $12.45 $13.05 $14.78 $15.26 $16.78 $22.18 $17.77 $19.77 $26.39 $20.03 $23.22 $30.76 $23.94 $28.07 $29.03 $35.55 $40.78 $48.36 $74.33 In 2010, those with a bachelor s degree or higher earned significantly higher hourly wages than those with only a high school diploma or some college, and this gap widened for the higher wage percentiles (Figure III-10). The lowest-paid group of high school graduates (those at the 10 th percentile - the wage at which 10 percent of workers earned less money and 90 percent earned more) earned $8.18/hour, about 38 percent less than the lowest paid college graduates, who earned $13.13/hour. The highest-paid (90 th percentile) college graduates earned 156 percent more than the highest-paid high school graduates ($74.33/hour compared to $29.03/hour). The returns to a college degree were very significant the median college graduate earned more than the highest paid high school graduates ($30.76 vs. $29.03). However, attending college without earning a bachelor s degree produced only minor wage benefits. Across all percentiles, those with some college fared only slightly better than those with only a high school diploma. $70 $ Hourly Wages By Education Level $50 $40 High School Some college Bachelor's or higher $30 $20 $10 $0 10th percentile 20th 30th 40th 50th 60th 70th 80th 90th percentile Source: CT Voices and Economic Policy Institute analysis of CPS data. Data for the 10th and 90th percentiles of workers with some college were not available due to small sample size. Dollar values are inflation adjusted to 2010 dollars using CPI-U-RS. Figure III-10 Across all levels of educational attainment, the top percentiles of Connecticut workers saw increases in their hourly wages from 2006 to For those with only a high school diploma, the bottom 70 percent of workers saw decreases in their hourly wages during that period (Figure III-11). In contrast, only the bottom 10 percent of workers with a four-year college degree saw a decline (Figure III-13). The top half of those with some college saw wage increases, while the bottom half experienced falling wages (Figure III-12). By far the largest wage increases occurred for the highest-paid workers with a college degree. The 90 th percentile wage for college degree-holders increased by 35 percent compared to 3 percent growth for the 90 th percentile high school graduates wages. Connecticut Voices for Children 24

26 $8.53 $10.68 $10.07 $12.65 $12.45 $14.91 $14.78 $16.81 $16.78 $19.33 $19.77 $22.09 $23.22 $26.49 $28.07 $33.77 $8.59 $8.18 $10.03 $9.60 $11.98 $11.15 $14.05 $13.05 $16.07 $15.26 $17.93 $17.77 $20.63 $20.03 $23.35 $23.94 $28.10 $29.03 $ and 2010 Hourly Wages for High School Graduates $ $20 $15 $10 $5 $0 Source: CT Voices and Economic Policy Institute analysis of CPS data. Dollar values are inflation adjusted to 2010 dollars using CPI-U-RS. Figure III-11 $ and 2010 Hourly Wages for People with Some College $30 $ $20 $15 $10 $5 $0 Source: CT Voices and Economic Policy Institute analysis of CPS data. Data for the 10th and 20th percentiles were not available due to small sample size. Dollar values are inflation adjusted to 2010 dollars using CPI-U-RS. Due to insufficient sample size, data were not available for 10th and 90th percentile earners in Figure III-12 Connecticut Voices for Children 25

27 $13.15 $13.13 $17.21 $18.15 $21.66 $22.18 $25.64 $26.39 $28.54 $30.76 $32.44 $35.55 $38.02 $40.78 $43.82 $48.36 $55.19 $74.33 $70 $ and 2010 Hourly Wages for People with College Degrees $50 $ $30 $20 $10 $0 Source: CT Voices and Economic Policy Institute analysis of CPS data. Dollar values are inflation adjusted to 2010 dollars using CPI-U-RS. Figure III-13 The significant education-related wage differentials have important racial implications because of large disparities in educational attainment between whites, blacks, and Hispanics. Forty percent of Connecticut whites over age 25 have a four-year college degree, compared to 14 percent of Hispanics and 18 percent of blacks. Thirty-three percent of Hispanics have not completed high school or earned a GED, which is true of 18 percent of blacks and 7 percent of whites (Figure III-14). Educational Attainment by Race in Connecticut, Ages 25 and Older 14% 18% 21% 32% 29% 40% 25% Bachelor's or Higher Some College or Assoc. Degree HS Diploma or GED 35% 33% 18% 28% 7% Less than HS Diploma Hispanic Black White, Non-Hispanic Source: CT Voices analysis of ACS 2010 data. Figure III-14 Connecticut Voices for Children 26

28 Differences between the educational attainment of Connecticut s white, black, and Hispanic populations and those groups nationally help account for the differences between Connecticut and national minority wage gaps. Connecticut s white population is markedly better educated than the national average (40 percent have bachelor s degrees in Connecticut compared to 31 percent nationally), which likely contributes to higher wages earned by white Connecticut workers (Figure III-15). Educational Attainment for Non- Hispanic Whites Ages 25 and Older 40% 39% 25% 25% 31% 30% Bachelor's or Higher Some College or Assoc. Degree HS Diploma or GED 28% 28% 29% Less than HS Diploma 7% 8% 9% Connecticut Peer State Average United States Source: CT Voices analysis of ACS 2010 data. Figure III-15 Hispanics in Connecticut are also better educated than the national average, with fewer having less than a high school degree (33 percent in Connecticut versus 38 percent nationally) and more with a college degree (14 percent versus 13 percent) (Figure III-16). The higher level of education of Connecticut s Hispanics helps explain why the white-hispanic wage gap is narrower in Connecticut compared to the national wage gap. While Connecticut s Hispanics are markedly more educated than Hispanics nationally, the wage gap is only marginally narrower in Connecticut (Hispanics in Connecticut earn median wages 69.4 percent as high as those earned by whites, versus 68.7 percent nationwide). 20 This is due in large part to the significantly higher educational attainment of Connecticut whites and their subsequently much higher wages compared to the national average. Connecticut has a slightly smaller percentage of Hispanics without a high school diploma than our peer states (33 percent versus 34 percent). However, a slightly smaller percentage of Connecticut Hispanics have a college degree (14 percent versus 16 percent) and have attended some college (21 percent versus 22 percent) than in our peer states. 20 See Figure III-5 Connecticut Voices for Children 27