Click here for State-by-State Interactive Map

(Only available on desktop)

Executive Summary

While the human toll of the opioid crisis is unquestionable, rising opioid dependency is also likely impacting the nation’s labor market and economy. This study examines the impact of the opioid crisis on labor force participation and economic growth, both nationally and in each state. While a number of previous studies have estimated the economic costs associated with the opioid crisis, none has estimated the impact on workers and economic growth in each state. This study updates a previous American Action Forum report on the nationwide impact of the opioid crisis on the labor market and economy. It then analyzes the decline in labor force participation and economic growth due to opioids in each state. It finds:

• Nationwide, in 2015, over 2 million prime-age individuals were not in the labor force due to opioids;
• Between 1999 and 2015, the decline in labor force participation cumulatively cost the economy 27 billion work hours;
• During that period, the reduction in work hours slowed the real annual economic growth rate by 0.6 percentage points, cumulatively costing nearly $1.6 trillion in real output; and
• The largest negative economic effects in the country occurred in Arkansas and West Virginia, where the prime-age labor force participation rate declined by 3.8 percentage points and the real economic growth rate slowed by 1.7 percentage points.

Introduction

The opioid crisis is among the most significant and immediate challenges facing the United States. In 2017, nearly 48,000 people died from an opioid-involved overdose, a 13 percent increase from 2016. While the health and personal consequences of the opioid crisis are apparent, the increase in opioid dependency also is likely impacting the broader U.S. economy. Countless media reports anecdotally highlight businesses that are unable to find workers because too many potential applicants cannot pass a drug test. But there has been little research that quantifies the loss of workers due to the opioid epidemic.

This study builds on previous American Action Forum (AAF) research to estimate the number of prime-age individuals (ages 25-54) absent from the labor force due to opioids and the resulting loss in economic growth. In particular, it both revises the previous study and estimates the number of individuals not in the labor force in each state. It finds that nationally between 1999 and 2015, opioid dependency reduced the prime-age male and female labor force participation rates by 1.4 percentage points and 1.8 percentage points, respectively. Consequently, by 2015 just over 2 million prime-age individuals were absent from the labor force due to opioids. During this period, the loss of labor slowed the annual real (inflation-adjusted) economic growth rate by 0.6 percentage points, cumulatively costing the U.S. economy roughly $1.6 trillion.

This study also finds that the opioid epidemic’s impact on labor force participation and economic growth varies considerably by state. Opioids had the largest negative labor market effects in West Virginia and Arkansas. In each, opioids reduced the prime-age labor force participation rate by 3.8 percentage points and, as a result, slowed the real economic growth rate by 1.7 percentage points. The opioid epidemic also substantially reduced prime-age labor force participation in Missouri (3.0 percentage points), Georgia (2.9 percentage points), New York (2.9 percentage points), and Kentucky (2.8 percentage points). The decline in labor force participation also caused real economic growth to slow significantly in these states.

Overview of the Opioid Crisis

Fueled by the proliferation of opioid prescriptions and exacerbated by an inundation of illegal and even more dangerous opioids, addiction has grown rapidly in recent decades. The result has been an abrupt escalation in drug overdose deaths: In 2017, 47,872 Americans died (131 per day) from opioid-related drug overdoses, according to the Centers for Disease Control and Prevention’s (CDC) preliminary estimate.[1] Since 1999, annual opioid overdose fatalities have risen by 495 percent, or 10 percent per year. ^[2] In recent years, the growth of annual opioid overdose fatalities had also been accelerating: 8 percent in 2013, 14 percent in 2014, 16 percent in 2015, and 28 percent in 2016.[3] As a result, opioids have contributed to the first decline in U.S. life expectancy since the height of the AIDS epidemic in 1993.[4]

The CDC’s recent preliminary data indicate that opioid-involved overdose fatalities rose by 13 percent in 2017.[5] Although opioid-involved overdose fatalities continued to rise to a new record level in 2017, the fact that the growth rate decelerated to the lowest level since 2013 tentatively suggests the opioid epidemic could be tapering.

Individuals are generally first exposed to opioids when doctors prescribe legal opioid painkillers, and trends in overdose fatalities suggest that growing access to prescription opioids built the foundation of opioid dependency in the United States. Yet, the nation’s opioid dependency became a crisis when policy efforts effectively restricted access to prescription opioids. As access to prescriptions declined, users turned to more potent illegal opioids, which caused overdose fatalities to accelerate.

A previous AAF study examined opioid-involved overdose fatalities between 1999 and 2016.[6] It found that between 1999 and 2010, the per capita quantity of legal prescription opioids grew at a 14.3 percent annual rate. Along with this growth in supply, prescription opioid-involved overdose fatalities grew 13.4 percent annually. After 2010, the supply of prescription opioids declined each year, and growth in overdose fatalities involving those prescriptions slowed to just 4.8 percent annually. This shift coincides with several major efforts to reduce access to and abuse of prescription opioids, including a reformulation of OxyContin that made it more difficult to abuse.

Yet, when the supply of prescription opioids declined, dependency remained, and users began turning to more potent illegal opioids instead. Indeed, transnational criminal organizations took advantage of unmet demand for opioids and flooded black markets with heroin and illegal synthetic opioids (e.g., fentanyl). Consequently, despite the growth in prescription opioid-involved overdose deaths slowing, the total number of opioid-involved overdose deaths accelerated because overdose deaths involving illegal opioids spiked. Growth in overdose deaths involving heroin accelerated from 4.1 percent per year before 2010 to 31.2 percent per year after 2010. Likewise, annual growth in overdose deaths involving illegal synthetic opioids accelerated from 13.7 percent to 36.5 percent. Growth in illegal synthetic opioid-involved overdose fatalities has been particularly alarming over the last few years. Between 2013 and 2016, overdose fatalities involving illegal synthetic opioids rose 84.2 percent annually.[7]

Preliminary data indicate that the number of annual overdose fatalities involving all three types of opioids continued to grow in 2017, but at substantially slower rates. Specifically, the number of overdose fatalities involving legal prescription opioids, heroin, and illegal synthetic opioids grew 0.5 percent, 0.8 percent, and 47.6 percent, respectively.[8]

While the human toll of the opioid crisis is alarming, research also indicates that opioid dependency is impacting the labor force and harming the broader U.S. economy. A report by Alan Krueger, former Chairman of the White House Council of Economic Advisers, found that opioids are likely pulling prime-age workers out of the labor force. In particular, Krueger found that nearly half of the prime-age men who are not in the labor force take pain medicine daily. Of those taking pain medicine, two thirds use prescription painkillers.[9] Additionally, Krueger found statistically significant evidence that the rise in opioid prescriptions is tied to the decline in the prime-age labor force participation rate for both men and women. His results suggest that the increase in opioid prescriptions between 1999 and 2015 could account for 20 percent of the decline in the prime-age male and 25 percent of the decline in the prime-age female labor force participation rate over that period.[10]

A recent AAF report used Kreuger’s results to estimate the number of prime-age workers absent from the labor force due to opioids and the resulting impact on economic growth.[11] It found that between 1999 and 2015, opioids pulled a growing number of workers out of the labor force. By 2015, 919,400 prime-age individuals were absent from the labor force due to opioids. Between 1999 and 2015, the decline in labor force participation resulted in a loss of 12.1 billion work hours. The reduction in work hours in turn slowed the annual real gross domestic product (GDP) growth rate by 0.2 percentage points, cumulatively costing $702.1 billion in real output.[12]

Methodology

This analysis is done in two parts. First it corrects for a math error AAF discovered in Krueger’s report and updates the previous AAF report’s nationwide estimates of the lost workers and economic growth due to opioids. Second, it estimates the loss of workers and economic growth in each state (also based on the correction in Krueger’s results) due to opioids.

Nationwide Analysis

This study, like AAF’s previous study, applies the revised implications of Krueger’s report to a combination of data from the Bureau of Labor Statistics (BLS): labor force and unemployment statistics from the Current Population Survey (CPS)[13] and hours and productivity statistics from the multifactor productivity program.[14] Krueger estimates the way opioids impact labor force participation with regression models that measure the prevalence of prescription opioids against the change in prime-age labor force participation rate in each country. Krueger’s metric for the supply of prescription opioids is morphine milligram equivalent (MME) prescribed per capita, a standardized metric that aggregates all types and strengths of prescription opioids. When interpreting the nationwide implications of the results, Krueger measures the nationwide change in MME prescribed per capita during the period studied (1999 to 2015) and then multiplies that change by the point estimate on the regression’s opioid variable. That multiplication calculates the implied decline in the nationwide prime-age labor force participation rate associated with the rise in the supply of per capita prescription opioids.

The aforementioned error is an inconsistent measure of MME prescribed per capita. When Krueger estimates his regression model, he measures MME prescribed per capita in natural logarithms. But when interpreting the results, he measures MME per capita in log base 10 to gauge the growth in prescription opioids. Since Krueger measures MME prescribed per capita in natural logarithms when estimating the regression, the variable should also be measured in natural logarithms when interpreting the results.

Correcting for the error generates a much larger decline in labor force participation than originally reported. Between 1999 and 2015, MME prescribed per capita rose by 256 percent. When measuring MME prescribed per capita in log base 10, the regression results indicate that prescription opioids led the nationwide prime-age labor force participation rate to decline by 0.6 percentage points for men (20 percent of the total decline) and 0.8 percentage points for women (25 percent of the total decline). When correctly measuring MME prescribed per capita in natural logarithms, however, the regression indicates that growth in opioids led the nationwide prime-age labor force participation rate to decline by 1.4 percentage points for men (40 percent of the total decline) and 1.8 percentage points for women (nearly 60 percent of the total decline).

This AAF study updates the previous AAF study by calculating the nationwide loss in prime-age labor and the resulting decline in economic growth each year between 1999 and 2015 using the adjusted interpretation of Krueger’s results. Specifically, this study uses Krueger’s results to calculate alternative prime-age male and female labor force participation rates in each year from 1999 to 2015 had the participation rate not declined due to opioids. The entire effect for each gender implied by Krueger’s results is spread evenly across the years during the period so that it accumulates to the full effect in 2015. For instance, the decline in the prime-age male labor force participation rate increases at a constant rate in each year from 1999 to 2015 and reaches the full 1.4 percentage point decline in 2015.[15] The study calculates the workers absent in each year due to opioids by subtracting labor force participation in the non-opioid alternative scenario from the actual labor force participation rate measured by the CPS.

To estimate the resulting loss of real economic growth, this study first estimates for each year the number of these individuals who would have been employed had they been in the labor force. To do so, the study uses each year’s average unemployment rate for prime-age men and women, as measured by the CPS, and assumes that, had the individuals not working due to opioids been in the labor force, they would have had the same unemployment rates. Then using historical annual data from the multifactor productivity program, the study calculates total work hours lost by multiplying average annual hours per each employed person by the loss of employed people due to opioids. Finally, the study multiplies the labor productivity (average real output per worker per hour) by total hours lost to estimate in each year the total real economic production lost due to opioids.

Finally, to estimate the impact on the annual real GDP growth rate, this study calculates an alternative real GDP baseline that adds the foregone real output to the actual real GDP in each year from 1999 to 2015, as reported by the Bureau of Economic Analysis (BEA).[16] The study then calculates the annual growth rate of the alternative real GDP baseline and compares it to the actual growth rate to derive the difference.

State-by-State Analysis

To estimate the change in labor due to opioids and the resulting impact on economic growth in each state, this study performs a similar analysis for each state. It uses a combination of state-level opioid data from the CDC,[17]^{, [18]} labor force data from the BLS’s Local Area Unemployment Statistics (LAUS),[19] and hours and productivity statistics from the multifactor productivity program.

This study uses state-level opioid data to measure the change in the supply of opioid prescriptions between 1999 and 2015 in each state, and then applies Krueger’s estimates to calculate the resulting decline in male and female prime-age labor force participation. This analysis faces the challenge of opioid data being more limited on the state level than the national level. At the state level, the metric MME prescribed per capita is only available for the years 2010 and 2015. Since the national supply of prescription opioids declined after 2010, it is vital to get a sense of the growth in each state before 2010. As previously noted, an earlier AAF report found that between 1999 and 2010, nationwide growth of MME prescribed per capita tracked closely to the growth of overdose fatalities involving prescription opioids (14.3 percent vs 13.4 percent annually). Data on annual prescription opioid-involved overdose fatalities are available for each state. So, as a second-best option, this study measures the growth in the supply of prescription opioids back to 1999 in each state by assuming that from 1999 to 2010, the growth in prescription opioid-involved overdose fatalities matched the growth in MME per capita.[20] It then uses the CDC state-level MME per capita data to calculate the growth in prescription opioids from 2010 to 2015. Finally, it combines the pre- and post-2010 measured growth to calculate the growth in the supply of prescription opioids for the entire period, 1999 to 2015.

As with the updated nationwide analysis, for each state this study applies Krueger’s results, adjusted using natural logarithms, to calculate the percentage point change in the prime-age labor force participation rate associated with the state’s change in opioids from 1999 to 2015. Next, the study applies the percentage point decline in labor force participation in each state to LAUS labor force data to estimate the number of prime-age workers absent from the labor force due to opioids in each year from 1999 to 2015. Specifically, it uses Krueger’s results to calculate alternative prime-age male and female labor force participation rates in each year had they not changed due to opioids. This study also assumes the total implied opioid effect is reached in 2015 after rising at a constant rate since 1999. The study calculates the number of workers absent in each year due to opioids by comparing the alternative labor force participation rate with the actual one.

To estimate the resulting loss of real economic growth in each state, the study first estimates the number of these individuals who would have been employed in each year had they been in the labor force. It does so with the state’s average unemployment rates for prime-age men and women for each year, derived from the LAUS, and assumes that the workers not in the labor force due to opioids would have had the same unemployment rates. The study then calculates total work hours lost by multiplying nationwide average annual hours per employed person, based on the same historical annual data from the multifactor productivity program, by the loss of employed people in the state due to opioids. Finally, the study multiplies the average labor productivity by total work hours lost in the state to estimate the state’s real economic production lost due to opioids.

While applying the nationwide average annual hours and labor productivity data to state-level labor loss figures is not ideal, no similar hours and productivity data are available at the state level. Additionally, to test whether the nationwide productivity data make a meaningful impact on the variation in results between the states, the author of this study performed a separate, unreported, analysis using state-level productivity estimates. In particular, the separate analysis measures worker productivity in each state by dividing state-level real gross domestic product by employment in each year from 1999 to 2015.[21] It then multiplies the worker productivity measure in each state, which effectively is output per employed person, by the loss of employed persons. The state-by-state variation in the change in economic growth derived from this method is not substantively different from the variation in economic growth derived with the nationwide hours and productivity data. This study reports the results based the nationwide hours and productivity data because they are more comparable to the nationwide analysis.

Finally, to estimate the impact on the states’ annual real GDP growth rates, this study calculates alternative real GDP baselines that add the states’ foregone real output to the actual real GDPs, as reported by the BEA, in each year from 1999 to 2015.[22] The study then calculates the annual growth rate in each state under the alternative real GDP baseline and compares it to the actual growth rate to derive the difference.

Nationwide Results

Applying the corrected interpretation of Krueger’s results to labor market data reveals that millions of prime-age individuals are absent from the labor force due to opioids, costing the economy about $1.6 trillion in foregone economic growth between 1999 and 2015.

Labor Force Participation

Between 1999 and 2015, per capita prescription opioids increased 256 percent, or about 8 percent annually. Table 1 contains the associated change in the prime-age labor force participation rate over the same period, and the resulting number of workers absent from the labor force as of 2015.

Table 1: The Impact of Opioids on Prime-Age Labor Force Participation, 1999-2015

Gender	Prime-Age Labor Force Participation Rate, 1999-2015 (in percentage points)	Workers, 2015 (in thousands)*
Total	-1.6	-2,006.1
Men	-1.4	-860.0
Women	-1.8	-1,146.2
*Estimates for each gender may not add to total due to rounding.

The rise in opioid prescriptions from 1999 to 2015 led the labor force participation rate for both prime-age men and women to decline substantially. As previously mentioned, opioids lowered the participation rates of prime-age men and women by 1.4 percentage points and 1.8 percentage points, respectively. The 1.4 percentage point decline in prime-age male labor force participation equates to roughly 40 percent of the overall decline in prime-age male labor force participation between 1999 and 2015. For prime-age women, the 1.8 percentage point decline tied to opioids equates to almost 60 percent of the entire decline in their participation rate.

The decline in the prime-age male labor force participation rate means that in 2015 860,000 men were absent from the labor force due to opioids. The steeper decline in prime-age female labor force participation means that even more women were absent from the labor force due to opioids. In 2015, opioids kept over 1.1 million women out of the labor force. Together, the growth in per capita prescription opioids from 1999 to 2015 caused the total prime-age labor force participation rate to decline by 1.6 percentage points. That decline translates to a loss of 2 million workers as of 2015.

Work Hours

From 1999 to 2015, the rise in opioid dependency and resulting decline in prime-age labor force participation cumulatively cost the economy nearly 30 billion work hours. Table 2 contains the cumulative loss of work hours associated with the decline in labor force participation.

Table 2: Impact of Opioids on Work Hours, 1999-2015

Gender	Work Hours, Cumulative 1999-2015 (in millions)
Total	-27,001
Men	-11,593
Women	-15,408

As the number of individuals absent from the labor force due to opioids grew, the economy lost more and more work hours. Between 1999 and 2015, the loss of workers cost the U.S. economy over 27 billion work hours. Since opioid dependency led more women out of the labor force than men, the majority of the lost work hours was attributed to the decline in female labor force participation. Specifically, the economy lost 15.4 billion work hours due to absent female workers. 11.6 billion lost work hours were tied to absent male workers.

Real Economic Growth

The tens of billions in lost work hours resulted in a major drag on U.S. economic growth. Table 3 contains the cumulative reduction in real economic output due to the opioid crisis and the associated decline in the real GDP annual growth rate.

Table 3: Impact of Opioids on Real Economic Growth, 1999-2015 (in 2009 dollars)

Gender	Real Output, Cumulative 1999-2015 (in billions)	Annual Real GDP Growth Rate, 1999-2015 (in percentage points)
Total	-$1,574.5	-0.6
Men	-$676.0	-0.3
Women	-$898.5	-0.3

From 1999 to 2015, the opioid-induced decline in labor force participation was a major cost to the U.S. economy. During that timeframe, the U.S. economy cumulatively lost nearly $1.6 trillion in real economic output, which translates to the annual real GDP growth rate slowing by 0.6 percentage points. To put this loss in perspective, from 1999 to 2015, real GDP grew 2 percent annually. Had opioids not drawn 2 million prime-age workers out of the labor force, real GDP would have grown 2.6 percent each year, a 30 percent increase.

Since more women left the labor force due to opioids than men, the decline in female labor force participation resulted in a larger portion of the economic cost. The decline in female labor resulted in a cumulative loss of $898.5 billion in real output between 1999 and 2015. The decline in male workers cost the economy $676 billion. The difference, however, is not large enough to translate to a substantially different decline in the economic growth rate, as the lost labor associated with each gender slowed the real GDP growth rate by about 0.3 percentage points.

State Results: Overview

The opioid epidemic’s impact on labor force participation and the economy vary considerably by state. The map below provides an overview of the central findings in each state.

Click on any state for a more detailed examination of the local labor and economic effects of the opioid crisis.

The states hit hardest by the opioid epidemic are Arkansas and West Virginia. Specifically, opioids led the prime-age labor force participation rate in each state to decline by 3.8 percentage points. That loss of labor caused the annual real GDP growth to decline by 1.7 percentage points in each state.

For perspective, the negative labor market effects of opioids in Arkansas and West Virginia were substantially larger than those in Missouri, the state with the next largest decline in labor force participation. Opioids caused Missouri’s prime-age labor force participation rate to decline by 3.0 percentage points. That led annual real GDP growth in Missouri to decline by 1.3 percentage points.

Opioids also drew a substantial portion of workers out of the labor force in Georgia, New York, and Kentucky. In those states, opioids were associated with the prime-age labor force participation rate declining by 2.9, 2.9, and 2.8 percentage points, respectively. As a result, the annual real GDP growth rate in Georgia, New York, and Kentucky declined by 1.2, 0.8, and 1.3 percentage points, respectively.

In absolute terms, opioids tended to have the largest negative impact on the labor force and economy in the more populous states. New York and Texas lost the most workers, work hours, and real economic output due to opioids. No state had more individuals absent from the labor force due to opioids than New York. As of 2015, 225,900 people were absent from New York’s labor force due to opioids. Likewise, in Texas opioids kept 171,100 people from the labor force. The decline in labor in these states also resulted in the largest loss of economic output. In particular, the decline in labor force participation between 1999 and 2015 cumulatively cost New York and Texas 3.1 billion and 2.2 billion work hours, respectively. That, in turn, cost New York $179.4 billion and Texas $128.8 billion in real output.

Somewhat unexpectedly, combining prescription opioid overdose fatality data with MME prescribed per capita resulted in two states, California and New Mexico, recording a slight net decline in the supply of prescription opioids between 1999 and 2015. The supply of opioids in those states started fairly high at the beginning of the period and did not increase substantially through 2010. But, after 2010, when access to prescription opioids were more restricted nationally, the supply of opioids in those states declined just like in nearly every other state. That resulted in a net decline in the supply of opioids by 2015.

The small decline in prescription opioids in California and New Mexico caused the labor forces and economies in those states to grow slightly. In California, the labor force participation rate rose by 0.1 percentage points, adding a total of $12.9 billion in real economic output. The additional growth in California, however, resulted in less than a 0.1 percentage point increase in the state’s annual real GDP growth rate. In New Mexico, the decline in opioids caused the prime-age labor force participation rate to increase by 0.3 percentage points, which led the state’s real GDP growth rate to increase by just 0.1 percentage point.

Of note, adding up the results from each state results in total figures that are similar to those derived from the nationwide analysis. Adding lost workers in 2015 from each state yields a total of 2.1 million workers absent due to opioids. Adding up the cumulative impact on work hours and real output results in a total loss of 28.1 billion work hours and $1.6 trillion in real economic output between 1999 and 2015. For comparison, the nationwide analysis reported in the previous section found a loss of 2 million workers, 27 billion work hours, and $1.6 trillion in real output. The comparability between the two methods suggests that, in the state-level analysis, the assumption that MME per capita grew at the same rate as prescription opioid-involved fatalities resulted in a fairly accurate measure of the growth of MME per capita in each state before 2010.

Conclusion

The U.S. economy depends on prime-age workers because they are among the most productive workers in the labor force. Yet, the growth in opioid dependency over the past two decades has contributed to their falling labor force participation rate.

As of 2015, over 2 million workers were absent from the labor force due to opioids. Between 1999 and 2015, the growing loss of labor cumulatively cost the economy 27 billion work hours and nearly $1.6 billion in lost output, slowing the annual real GDP growth rate by 0.6 percentage points.

Additionally, the opioid crisis’s effect on labor and economic growth varies considerably by state. The states most dramatically impacted by the opioid crisis are Arkansas and West Virginia. Opioids lowered the prime-age labor force participation rate by 3.8 percentage points in both Arkansas and West Virginia, which in turn slowed their annual real GDP growth rates by 1.7 percentage points.

In absolute terms, larger states lost the most workers and real output due to opioids. As of 2015, opioids kept 225,900 people out of New York’s labor force and 171,100 out of Texas’s labor force. That cost these state economies $179.4 billion and $128.8 billion, respectively, in real output.

As federal, state, and local policymakers consider ways to grow the economy and boost the labor supply, addressing the opioid epidemic must be part of the solution.

 

[1] “12 Month-ending Provisional Number of Drug Overdose Deaths by Drug or Drug Class,” Provisional Drug Overdose Death Counts, National Center for Health Statistics, Center for Disease Control and Prevention, accessed August 21, 2018, https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm.

[2] Authors’ analysis of data from “12 Month-ending Provisional Number of Drug Overdose Deaths by Drug or Drug Class,” Provisional Drug Overdose Death Counts, National Center for Health Statistics, Center for Disease Control and Prevention, accessed August 21, 2018, https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm.

[3] Authors’ analysis of data from “Data Brief 294. Drug Overdose Deaths in the United States, 1999-2016,” Figure 4, Centers for Disease Control and Prevention, https://www.cdc.gov/nchs/data/databriefs/db294_table.pdf.

[4] Deborah Dowell et al., “Contribution of Opioid-Involved Poisoning to the Change in Life Expectancy in the United States, 2000-2015,” JAMA, 2017;318(11):1065–1067, doi:10.1001/jama.2017.9308, https://jamanetwork.com/journals/jama/fullarticle/2654372.

[5] Authors’ analysis of data from “12 Month-ending Provisional Number of Drug Overdose Deaths by Drug or Drug Class,” Provisional Drug Overdose Death Counts, National Center for Health Statistics, Center for Disease Control and Prevention, accessed August 21, 2018, https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm.

[6] Ben Gitis & Isabel Soto, “The Types of Opioids Behind Growing Overdose Fatalities,” American Action Forum, April 11, 2018, https://www.americanactionforum.org/research/types-opioids-behind-growing-overdose-fatalities/.

[7] Ibid.

[8] Authors’ analysis of data from “12 Month-ending Provisional Number of Drug Overdose Deaths by Drug or Drug Class,” Provisional Drug Overdose Death Counts, National Center for Health Statistics, Center for Disease Control and Prevention, accessed August 21, 2018, https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm.

[9] Alan Krueger, “Where Have All the Workers Gone? An Inquiry into the Decline of the US Labor Force Participation Rate,” Brookings Papers on Economic Activity, Sept 7, 2017, p. 25, https://www.brookings.edu/bpea-articles/where-have-all-the-workers-gone-an-inquiry-into-the-decline-of-the-u-s-labor-force-participation-rate/.

[10] Ibid., pp. 48-54.

[11] Ben Gitis & Isabel Soto, “The Labor Force and Output Consequences of the Opioid Crisis,” American Action Forum, March 27, 2018, https://www.americanactionforum.org/research/labor-force-output-consequences-opioid-crisis/.

[12] Ibid.

[13] Labor Force Statistics from the Current Population Survey, Bureau of Labor Statistics, https://www.bls.gov/cps/tables.htm.

[14] Multifactor Productivity, Bureau of Labor Statistics, https://www.bls.gov/mfp/.

[15] It is important to note that this method is somewhat different than the one used in the previous AAF study. Krueger’s reported 0.6 percentage point and 0.8 percentage point declines in the prime-age participation rates equated to 20 percent and 25 percent of the total decline in prime-age male and female labor force participation, respectively. The previous study simply calculated the total number of additional prime-age male and female workers who would have been in the labor force in each year from 1999 to 2015, had the participation rates remained at 1999 levels, and assumed that 20 percent and 25 percent of the difference for men and women, respectively, was due to opioids. This study more precisely calculates the impact of opioids on the male and female prime-age labor force participation rates in each year from 1999 to 2015 (assuming the effect implied by Kruger’s results are spread evenly throughout the period) and then calculates the number of additional workers who would have been in the labor force without the impact of opioids.

[16] Interactive Data, Bureau of Economic Analysis, U.S. Department of Commerce, https://www.bea.gov/itable/index.cfm.

[17] “Opioid Overdose Deaths by Type of Opioid,” State Health Facts, Henry J Kaiser Family Foundation, https://www.kff.org/other/state-indicator/opioid-overdose-deaths-by-type-of-opioid/?currentTimeframe=0&sortModel=%7B%22colId%22:%22Location%22,%22sort%22:%22asc%22%7D.

[18] “Opioids prescribed per capita, in morphine mg equivalents (MME),” Opioid & Health Indicators Database, amfAR, http://opioid.amfar.org/indicator/mme_percap.

[19] “Expanded State Employment Status Demographic Data,” Local Area Unemployment Statistics, Bureau of Labor Statistics, U.S. Department of Labor, https://www.bls.gov/lau/ex14tables.htm.

[20] In some states, prescription opioid-involved overdose fatality data do not go back to 1999. For instance, some states were missing data on prescription opioid-involved overdose deaths specifically for 1999 or for 1999 and 2000. In those cases, this study calculates the annual growth rate in overdose fatalities involving prescription opioids using the 4 to 6 years following the years without the necessary data (the years that the growth rates are calculated from depends on which years missed data). The analysis assumes that in the years with missing data, overdose fatalities grew at the same rate, and projects overdose fatalities back to 1999.

[21] Interactive Data, Bureau of Economic Analysis, U.S. Department of Commerce, https://www.bea.gov/itable/index.cfm.

[22] Ibid.