PEOPLE: Demographics + Households

 Findings on Low-Wage Status by GenderLow-Wage Status by Race & Ethnicity, Low-Wage Status by Age Group, Low-Wage status and Educational Attainment, Low-Wage Status and Citizenship, and Low-Wage Status and Nativity were derived through author tabulations of 2012 American Community Survey (5% IPUMS-USA). The same data were used in our examination of the makeup of Low-Wage Households and of low-wage income as it contributes to overall household income. Low-wage in these instances is defined as <150% of poverty threshold for two-person household.  

PEOPLE: Underemployment

Occupational Employment Statistics (OES) data from the US Bureau of Labor Statistics was used to examine workers in low-wage occupations who are overqualified based on their level of education.

INDUSTRIES: Low-Wage Industries

To determine which industries were low-wage industries in North Carolina, we used 2012 data (the most recent full year available) from the Bureau of Labor Statistics’ Quarterly Census of Employment and Wages (QCEW). For each industry NAICS code (from the high-level 2-digit level to the granular 6-digit level), the QCEW reports total employment, total number of establishments, aggregate annual wages, and average weekly wages. Using data for the state of North Carolina at the 4-digit NAICS level, we took our low-wage cutoff of $11.34/hour, multiplied it by 40 hours per week, and used the result ($453.60) as a cutoff for average weekly pay – any industry below that cutoff is a low-wage industry. Because we wanted to focus on industries that are having a significant impact on North Carolina’s workforce, we chose to report industries that are below the low-wage cutoff and that have at least 25,000 employees in North Carolina. We used the same dataset and methodology to identify the highest wage industries in North Carolina.

INDUSTRIES: Wage and Employment Growth

To calculate growth in employment and wages in each industry, we used QCEW and processed it the same way as we did for “Low-Wage Industries.” In order to compare wages on a “purchasing power” basis, we converted 1992 dollars to 2012 dollars using an inflation factor based the Bureau of Labor Statistics’ CPI Inflation Calculator. All percent changes shown are absolute percent changes from the base year.

OCCUPATIONS: Low-Wage Occupations

To determine which occupations were low-wage occupations in North Carolina, we used the Bureau of Labor Statistics’ state-level Occupational Employment Statistics (OES) for 2013. For each SOC occupation code, the OES reports total annual employment, as well as the 10th, 25th, 50th (median), 75th, and 90th percentile annual and hourly wage. We identified low-wage occupations as those whose median hourly wages fell below $11.34/hour.

OCCUPATIONS: Wage and Employment Growth

To calculate growth in employment and wages for each occupation, we found the absolute percentage change between 2003 data and 2013 data for each occupation. To calculate wage growth, we used median hourly wage. So that we could compare wages on a “purchasing power” basis, we converted 2003 dollars to 2013 dollars using an inflation factor based the Bureau of Labor Statistics’ CPI Inflation Calculator.

OCCUPATIONS: Low-Wage Occupations in Industries

The Bureau of Labor Statistics’ Industry-Occupation Matrix (IOM) shows what percentage of workers in each NAICS industry belong to each SOC occupation, as well as what percentage of workers in a given SOC occupation work in each NAICS industry. IOM data is reported at the national level only. At the 2-digit NAICS level, for each industry, we added up the percentage of workers working in any occupation we identified as low-wage. We then applied those percentages to the number of workers in each industry in North Carolina in 2012.

INDUSTRY SNAPSHOTS

We compiled information from industry reports, including both public sources and paid research databases (accessed through the University of North Carolina at Chapel Hill Library) like IBISWorld. Each snapshot includes specific citations and links.

GEOGRAPHY: Low-Wage Workers

To show the percentage of total jobs that are low-wage and the magnitude of low-wage jobs, we used data provided by the US Census Bureau through the Longitudinal Employer-Household Dynamics (LEHD) Origin-Destination Employment Statistics (LODES) datasets for the year 2011 (the most current year of data available). We collected and organized the data at the County FIPS code level using STATA Statistical Software v.13. The NC Department of Commerce Tier designations were collected from the NC DOC’s website. We used ArcGIS Online as our visualization platform because it offers interactive capabilities and multi-layer analysis.

GEOGRAPHY: Urban Centers

To generate the dot density visualization for low-wage jobs, we used data provided by the US Census Bureau, made available through the Longitudinal Employer-Household Dynamics (LEHD) Origin-Destination Employment Statistics (LODES) datasets for the year 2011 (the most current year of data available). The Workplace Area Characteristics file, which shows where jobs are physically located, was used for this analysis because it provides the best spatial orientation. Using STATA Statistical Software v.13 we were able to organize the LODES data to show employment information down to the Census Tract FIPS code level. Using dot density techniques, we visualized low-wage jobs at the census tract level by joining the LODES data to current census tract maps using ArcGIS 10.2.

To show median household income, we used US Census Bureau data from the 5-year American Community Survey (2008-2012) because it most closely matched the timeframe for the LEHD-LODES dataset. It was gathered using the American FactFinder website, which provides the data at the census tract level. ArcGIS was then used to transpose the data onto maps.

GEOGRAPHY: Low-Wage Job Trends

To show the percentage change in low-wage jobs over time (2002-2011), we used data provided by the US Census Bureau through the Longitudinal Employer-Household Dynamics (LEHD) Origin-Destination Employment Statistics (LODES) datasets for the years 2002 and 2011. Please note that 2011 represents the most current statistics available from the Bureau of Labor Statistics. We collected and organized the data at the County FIPS code level using STATA Statistical Software v.13 and Microsoft Excel. We used ArcGIS Online as our visualization platform to host our findings.

GEOGRAPHY: Low-Wage Job Projections

Using the selected low-wage industries [link to page with Industry group’s definition], we assessed the number of workers in each of the 11 industries for each county. Data for the number of workers in each industry was derived from the National Establishment Time Series (NETS) Database for 2012. From North Carolina’s Department of Commerce regional employment projections, we were able to project the number of workers in low-wage industries based on the annual growth rate in each region for the select industries. Counties within the nine regions were assumed to have the same growth rate as the region as a whole.

GEOGRAPHY: Housing Affordability

Housing affordability was determined using the US Department of Housing and Urban Development's HUD 2015 fair market rents data. Housing affordability was defined as an individual's ability to rent a two-bedroom home using no more than 30% of his/her income. Why a two-bedroom home for one person? We focused on an individual supporting a two-person household, as highlighted in our definition of low-wage.

POLICY: Cost to Government

Using Integrated Public Use Microdata Series (IPUMS) census microdata, low-wage workers were derived by selecting people in North Carolina who reported doing any work for pay in the previous week or worked at least 15 hours in the previous week without pay. To determine whether a worker was a low-wage worker we reduced their earnings to a calculation of hourly pay, even if salaried. If the hourly pay fell at or below $11.34 per hour then the worker was designated low-wage.

For salaried workers, hourly wage was calculated by using the variable for annual income, dividing that by the variable for hours usually worked in a week and then dividing that figure by the variable that asks how many weeks an individual worked in a year. For individuals paid hourly, we used the variables asking if an individual was paid hourly (selecting those that answered yes) and asking an individual’s hourly wage. Summing these two categories of workers yielded total low-wage workers.

Distribution of WIC Spending and Amount Spent on Low-Wage Workers

We used WIC data provided by the Federal Department of Agriculture’s Food and Nutrition Service. The Current Population Survey (CPS) conducted by the Bureau of Census for the Bureau of Labor Statistics contains information on individuals receiving WIC; thus, individuals determined to be low-wage workers and also to receive WIC assistance were included. The USDA data show the number of individuals in North Carolina receiving WIC assistance. Using our constructed number of North Carolina low-wage workers and the federal data we determined the percentage of WIC recipients that were low-wage workers. Since our data included all workers at work, we could also find the number and percentage of WIC recipients that were not low-wage workers by the same process. The USDA data also contain the total dollar amount spent on WIC in North Carolina. We assumed this figure was distributed evenly among all recipients, and applied the percentage of WIC recipients who were low-wage workers to the total dollar amount spent on WIC in North Carolina to find the percentage and corresponding amount spent on low-wage workers.

Distribution of SNAP Spending and Amount Spent on Low-Wage Workers

SNAP data provided by the Federal Department of Agriculture’s Food and Nutrition Service. The CPS contains information on individuals receiving SNAP assistance; thus, individuals determined to be low-wage workers and also to receive SNAP assistance were included. The USDA data show how many individuals in North Carolina received SNAP assistance. We divided our constructed number of North Carolina low-wage workers into the federal data of total SNAP assistance recipients to find the percentage of SNAP assistance recipients that were low-wage workers. Since our micro data included all workers at work, we could also find the number and percentage of SNAP assistance recipients that were not low wage workers by the same process.

The USDA data have the total dollar amount spent on WIC in North Carolina. We assumed this figure was distributed evenly among all recipients, and applied the percentage of SNAP assistance recipients who were low-wage workers to the total dollar amount spent on WIC in North Carolina to find the percentage and corresponding amount spent on low-wage workers. The same process was repeated for non-low-wage workers at work that received SNAP assistance. The rest of SNAP assistance recipients were assumed to be not at work.

Medicaid Expenditures on Low-Wage Workers

The CPS contains information on individuals receiving Medicaid; thus, individuals determined to be low-wage workers and also to receive Medicaid assistance were included. We combined our constructed low-wage worker variable with individuals that receive Medicaid to find the number of low-wage workers that received Medicaid. The CPS also has a variable labeled “person value of Medicaid,” which tracks how much Medicaid money was spent per individual. We summed the person values of Medicaid of all low-wage workers who received Medicaid to find the total value of Medicaid expenditures on low-wage workers.

Cash Public Assistance and SNAP in Distressed Urban Census Tracts

We used census tracts in Durham, Mecklenburg and Wake Counties identified in a 2014 report from the University of North Carolina’s Center for Urban and Regional Studies (CURS). The US Census Bureau's American Community Survey (ACS) contains data at the census tract level on whether an individual received cash public assistance or SNAP benefits. We combined the census tracts by their counties and found the percentage who had received cash public assistance or SNAP benefits by dividing the number who responded yes to the question by the total residents of the census tracts.

Public Health Insurance in Distressed Urban Census Tracts

We used census tracts in Durham, Mecklenburg and Wake Counties identified in a 2014 report from the University of North Carolina’s Center for Urban and Regional Studies (CURS). The US Census Bureau's American Community Survey (ACS) contains data at the census tract level on whether an individual reported being covered by public health insurance. We combined the census tracts by their counties and found the percentage that were covered by public health insurance by dividing the number who responded yes to the question by the total residents of the census tracts.

POLICY: Minimum Wage

Using Integrated Public Use Microdata Series (IPUMS) census microdata, minimum wage workers were found by isolating the number of people in North Carolina who reported doing any work for pay in the previous work or worked at least 15 hours in the previous week without pay. To determine whether a worker was a minimum wage worker we reduced earnings to a calculation of hourly pay, even if salaried. If the hourly pay fell at or below $7.25 per hour (North Carolina’s minimum wage) then the worker was identified as a minimum-wage worker. The same process was performed to identify workers who earned more than minimum wage but who would still be impacted by a minimum wage increase to $10.10 per hour—in other words, workers earning between $7.25 and $10.10 per hour. The process was utilized again to find workers earning less than the minimum wage.

For salaried workers, hourly wage was calculated by using the variable for annual income, dividing that by the variable for hours worked in a typical work week and then dividing that figure by the variable that asks how many weeks an individual worked in a year. Hourly wages returned were used to sort individuals into the following categories: those earning below minimum wage (<$7.25/hour), those earning minimum wage ($7.25/hour), and those earning more than minimum wage but who would still be impacted by the proposed minimum wage increase ($7.25-$10.10/hour). We selected individuals who reported being paid hourly and those who reported or who were determined to have an hourly wage at or below $10.10 per hour, the combination of which produced the total number of minimum wage workers.

Wages Added by Minimum Wage Increase and Ripple Effect

To determine the amount of new wages that would be created by a minimum wage increase to $10.10 per hour, we used data from the Current Population Survey (CPS) conducted by the Bureau of Census for the Bureau of Labor Statistics to target individuals earning less than $10.10 per hour and to pinpoint their earnings. We then summed the difference between each of these individuals’ respective hourly wages and $10.10, which yielded the total amount of an increase to $10.10 per hour for one hour of work. That amount was then multiplied by 2,080, the widely-used baseline for annual full-time employee hours. This produced the yearly amount of new wages created by a minimum wage increase to $10.10 per hour.

A similar process was used to analyze the ripple effect of a minimum wage increase. We supposed individuals earning up to $12 per hour might also get a nominal raise as a result of an increased minimum wage. We first assumed those individuals might receive a 10% raise. We multiplied the amount of people earning a given amount above $10.10 per hour by their hourly wage times 0.1. To model a 5% raise we did the same process but multiplied the amount of people earning a given amount above $10.10 per hour by their hourly wage times 0.05. For the both the 10% and 5% raise, we summed the amount found by the previous process and multiplied that amount by 2,080, resulting in the yearly amount of new wages created.

Minimum Wage Households

To find the households in North Carolina that earned minimum wage we used the Current Population Survey (CPS) variable on household income. Multiplying the North Carolina minimum wage ($7.25/hour) by 2,080 annual full-time employee hours yielded an annual minimum wage income of $15,080 for a one-person household. To construct annual minimum wage incomes for households greater than one-person, we first used the CPS variable on number of individuals in a given household to group households by size.

We then used the federal poverty line as a ratio measure of household income to household size to determine how an increase in household size would affect income. As 150% of the federal poverty line is our threshold for low-wage work, we used the income thresholds at 150% of the federal poverty line as our guide. From this we derived the difference in incomes for a one-person household and for a two-person household at 150% of the federal poverty line, and applied that difference to minimum wage income. For example, the income at 150% of the federal poverty line for a two-person household is 34.7% greater than that of a one-person household. We found the differences for households up to six people, then used the same differences on households of different size earning minimum wage. For example, for a two-person household we set the minimum wage income at 34.7% greater than $15,080 (the minimum wage amount for a one-person household, based on one person earning minimum wage over the course of a year.)

Having found the number of households earning minimum wage for households up to six people, that figure was divided by all households by corresponding household size. This gave us the percentage of households by size earning minimum wage.

POLICY: Pay Gap Between Executives and Average Employee

The AFL-CIO’s Executive Paywatch uses public data from the Russell 3000 and the S&P 500 to produce a list of the sixty highest paid executives of North Carolina-based companies. The list includes executive names, companies, and salaries (2013). Using the ReferenceUSA tool, we were able to link each company to its primary (six-digit) NAICS code. Primary NAICS codes correspond to specific industry classifications such as home centers or women’s clothing stores. The Bureau of Labor Statistics provides average worker income data (2013) for industries, based on primary NAICS codes. We used this average annual employee income data for the companies' respective industries and compared these with their executives’ income in 2013 to find the executive-worker pay ratio.

We also examined the companies with North Carolina’s sixty highest paid executives at the two-digit NAICS code level. Two-digit NAICS codes provide broad industry classifications such as manufacturing and retail trade. Using the executive-worker pay ratio calculated earlier, we grouped the companies of North Carolina’s sixty highest paid executives by two-digit NAICS codes to provide a broad illustration of those industries with the largest gaps between executive pay and average worker pay.