Young adult substance use following involuntary job loss

Nicholas A. Jolly; Gwendolyn Davis

doi:10.2478/izajole-2023-0003

Table 1

Descriptive Statistics (means) of Selected Variables

	Non-Displaced	Displaced
Total Earnings^a,b	$3,564.65	$2,218.23
Hours Worked^b	15.99	21.38
Employed	56.03%	55.29%
Age	19.11	18.77
Female	54.90%	49.93%
White^c	51.73%	36.51%
Responsible for Own Household	15.17%	15.13%
Education
< High School	15.39%	13.86%
High School Degree/GED	23.97%	29.79%
Some College/Associates Degree	57.76%	56.12%
High School Plus Another Degree (not college)	0.12%	-
Bachelor's Degree	2.51%	0.23%
Master's Degree	0.25%	-
Currently Smoke	12.72%	21.77%
Number Cigarettes/Day	7.55	8.80
Currently Drink Alcohol	48.58%	46.50%
Drink at Least Several Times/Month	52.48%	50.43%
Number Drinks/Session	3.58	3.65
Ever Used Marijuana	40.17%	45.98%
Number of Uses of Marijuana previous 12 Months^d	12.61	14.33
Ever Used Amphetamines, Cocaine, Barbiturates, or Tranquilizers	15.81%	15.95%
Ever Used Pain Relievers/Narcotics	30.32%	34.43%
Number of Uses of Pain Relievers/Narcotics previous	5.34	6.17
12 Months^d
Total Number of Individuals	2,288	747

Notes: Sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. All calculations are means/proportions of selected variables. All years are used in the calculations. However, calculations only use the first observation in which an individual is observed.

a – Real 2018 dollars;

b – equals the total across all five jobs;

c – represents the modal report for race;

d – midpoint of the range is used in the calculation

Table 2

Displacement's Effect on Labor Market Outcomes

	Earnings	Log Earnings	Employed
Panel A: Average Annual Effect
After	−4,285.556^*** (733.586)	−0.754^*** (0.144)	−0.134^*** (0.020)
R²	0.17	0.16	0.08
N	10,816	8,460	10,816
# Individuals	3,039	2,863	3,039
Panel B: Event History
At least t-6	2,083.255^* (1,242.017)	−0.213 (0.286)	−0.018 (0.041)
t-4	1,721.573^** (773.122)	0.364 (0.233)	−0.023 (0.027)
t	−2,745.970^*** (624.354)	−0.743^*** (0.166)	−0.167^*** (0.023)
t+2	−4,545.250^*** (957.543)	−0.635^*** (0.209)	−0.106^*** (0.027)
t+4	−7,065.855^*** (1,407.884)	−0.482^* (0.260)	−0.070^** (0.033)
At least t+6	−8,929.685^*** (2,181.589)	−0.399 (0.315)	−0.130^*** (0.045)
R²	0.17	0.17	0.09
N	10,816	8,460	10,816
# Individuals	3,039	2,863	3,039

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. The coefficients in panel B are estimated relative to period t-2.

* p<0.1;

** p<0.05;

*** p<0.01

Table 3

Displacement's Effect on Substance Use

Panel A: Extensive Margin
	Currently Smokes	Currently Drinks	Ever Use Marijuana	Ever Used Other Drugs	Ever Used Pain Relievers/Narcotics
After	0.032^** (0.013)	0.031^* (0.018)	−0.016 (0.015)	0.010 (0.014)	−0.003 (0.037)
R²	0.00	0.11	0.06	0.03	0.06
N	10,790	10,781	10,759	10,759	4,486
# People	3,039	3,039	3,039	3,039	2,004
Avg. Pre-Job Loss	0.22	0.55	0.47	0.18	0.40

Panel B: Intensive Margin
	# Cigarettes per Day	Drinks Several Times per Month	# Drinks per Session	# Times used Marijuana past 12 Months	# Times used Pain Relievers/Narcotics past 12 Months
After	0.341 (0.551)	−0.030 (0.026)	0.463^*** (0.128)	1.967^** (0.827)	−2.167 (1.879)
R²	0.02	0.05	0.03	0.03	0.04
N	1,721	6,883	6,825	5,320	1,419
# People	748	2,516	2,509	2,019	978
Avg. Pre-Job Loss	8.20	0.56	3.48	14.13	6.34

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Table 4

Displacement's Effect on Substance Use by Type of Support – Extensive Margin

	Cigarettes	Alcohol	Marijuana	Other Drugs	Pain Relievers/Narcotics
Panel A: Parental Support
After	0.045^** (0.019)	0.035 (0.023)	0.016 (0.020)	0.031 (0.019)	0.054 (0.054)
R²	0.01	0.11	0.07	0.03	0.06
N	9,221	9,212	9,191	9,191	4,084
# Individuals	2,656	2,656	2,656	2,656	1,843
Panel B: Government Support
After	0.084^** (0.036)	−0.016 (0.048)	−0.043 (0.040)	0.073^* (0.038)	−0.068 (0.116)
R²	0.01	0.11	0.07	0.03	0.06
N	8,138	8,131	8,111	8,111	3,786
# Individuals	2,388	2,388	2,388	2,388	1,728
Panel C: Own Support
After	0.030^** (0.014)	0.040^** (0.020)	0.006 (0.017)	0.023 (0.016)	−0.018 (0.039)
R²	0.00	0.11	0.07	0.03	0.06
N	10,084	10,076	10,053	10,053	4,313
# Individuals	2,857	2,857	2,857	2,857	1,935

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. Parental support is parents/relatives helping to pay for tuition, rent, mortgage, other expenses not defined, or providing a personal loan. Government support includes SSI, TANF, or UI. Own support includes savings/checking accounts, stocks, mutual funds, investment trusts, money market accounts, government bonds, or certificates of deposits.

* p<0.1;

** p<0.05;

*** p<0.01

Table 5

Displacement's Effect on Substance Use by Type of Support – Intensive Margin

	# Cigarettes	Drinks Several Times per Month	# Drinks per Session	# Times Used Marijuana in Previous 12 Months	# Times Used Pain Relievers in Previous 12 Months
Panel A: Parental Support
After	0.614 (0.834)	−0.067^* (0.035)	0.214 (0.152)	3.147^*** (1.129)	−1.313 (2.667)
R²	0.03	0.05	0.04	0.03	0.04
N	1,359	5,958	5,908	4,544	1,291
# Individuals	608	2,190	2,184	1,749	894
Panel B: Government Support
After	0.539 (1.009)	0.066 (0.060)	0.679^** (0.334)	2.389 (1.615)	−3.034 (2.231)
R²	0.03	0.06	0.03	0.02	0.05
N	1,097	5,181	5,135	3,877	1,180
# Individuals	506	1,941	1,934	1,543	829
Panel C: Own Support
After	0.034 (0.636)	−0.036 (0.028)	0.324^** (0.135)	2.324^** (0.903)	−2.030 (2.136)
R²	0.02	0.05	0.03	0.03	0.04
N	1,503	6,508	6,455	4,937	1,363
# Individuals	661	2,368	2,362	1,882	939

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. See table 4 and text for descriptions of types of support.

* p<0.1;

** p<0.05;

*** p<0.01

Table 6

Displacement's Effect on Substance Use - Declining Economic Activity

Panel A: Extensive Margin
	Currently Smokes	Currently Drinks	Ever Use Marijuana	Ever Use Other Drugs
After	0.032 (0.027)	0.052 (0.034)	0.018 (0.029)	0.040 (0.029)
R²	0.00	0.11	0.07	0.03
N	8,585	8,577	8,559	8,559
# Individuals	2,492	2,492	2,492	2,492

Panel B: Intensive Margin
	# Cigarettes per Day	Drinks Several Times per Month	# Drinks per Session	# Times used Marijuana past 12 Months
After	1.000 (0.949)	−0.077^* (0.046)	0.597^*** (0.230)	0.344 (1.426)
R²	0.03	0.06	0.03	0.02
N	1,225	5,503	5,460	4,115
# Individuals	551	2,041	2,035	1,620

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. The treatment group is limited to those who experience an increase in their state's unemployment rate between the period before job loss and the year of displacement.

* p<0.1;

** p<0.05;

*** p<0.01

Table 7

Heterogeneity by Gender

Panel A: Substance Use at Extensive Margin
	Cigarettes	Alcohol	Marijuana	Other Drugs	Pain Relievers
Men	0.041^** (0.020)	0.004 (0.023)	−0.054^*** (0.020)	0.043^** (0.019)	0.062 (0.052)
Women	0.023 (0.015)	0.057^** (0.025)	0.023 (0.021)	−0.024 (0.019)	−0.055 (0.048)
R²	0.00	0.11	0.07	0.03	0.06
N	10,790	10,781	10,759	10,759	4,845
# Individuals	3,039	3,039	3,039	3,039	2,194
Sig. Different	No	No	1%	1%	10%
Men Avg.	0.275	0.595	0.567	0.208	0.319
Women Avg.	0.173	0.499	0.388	0.164	0.459

Panel B: Substance Use at Intensive Margin
	Number Cigarettes	Drinks Several Times per Month	# Drinks per Session	# Times used Marijuana	# Times used Pain Relievers
Men	0.058 (0.755)	−0.023 (0.035)	0.635^*** (0.187)	2.400^** (1.097)	2.252 (3.118)
Women	0.748 (0.584)	−0.038 (0.037)	0.263^* (0.152)	1.354 (1.123)	−4.685^** (2.158)
R²	0.02	0.05	0.03	0.03	0.05
N	1,721	6,883	6,825	5,320	1,595
# Individuals	748	2,516	2,509	2,019	1,095
Sig. Different	No	No	No	No	10%
Men Avg.	8.601	0.634	3.827	16.889	3.566
Women Avg.	7.598	0.472	3.091	10.350	7.799

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Table 8

Heterogeneity by Race

Panel A: Substance Use at Extensive Margin
	Cigarettes	Alcohol	Marijuana	Other Drugs	Pain Relievers
Non-White	0.053^*** (0.015)	0.045^** (0.022)	−0.020 (0.020)	−0.010 (0.016)	−0.052 (0.045)
White	−0.005 (0.021)	0.006 (0.027)	−0.009 (0.021)	0.042^* (0.024)	0.094 (0.057)
R²	0.01	0.11	0.06	0.03	0.06
N	10,783	10,774	10,751	10,751	4,844
# Individuals	3,036	3,036	3,036	3,036	2,193
Sig. Different	5%	No	No	10%	5%
White Avg.	0.298	0.677	0.532	0.360	0.384
Non-White Avg.	0.175	0.466	0.442	0.080	0.407

Panel B: Substance use at Intensive Margin
	Number Cigarettes	Drinks Several Times per Month	# Drinks per Session	# Times used Marijuana	# Times used Pain Relievers
Non-White	0.635 (0.768)	−0.004 (0.038)	0.669^*** (0.160)	2.621^** (1.042)	−4.053^* (2.203)
White	0.059 (0.668)	−0.060^* (0.033)	0.209 (0.187)	1.057 (1.228)	0.626 (3.093)
R²	0.02	0.05	0.03	0.03	0.05
N	1,717	6,878	6,820	5,317	1,595
# Individuals	747	2,514	2,507	2,018	1,095
Sig. Different	No	No	10%	No	No
White Avg.	9.589	0.628	4.088	13.941	3.566
Non-White Avg.	6.739	0.495	2.945	14.307	7.799

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Table 9

Heterogeneity by Living Arrangements

Panel A: Substance Use at Extensive Margin
	Cigarettes	Alcohol	Marijuana	Other Drugs	Pain Relievers
With Others	0.025 (0.015)	0.058^*** (0.022)	−0.020 (0.020)	0.035^** (0.017)	0.104^** (0.051)
On Own	0.040^** (0.020)	−0.003 (0.026)	−0.011 (0.021)	−0.021 (0.021)	−0.112^** (0.047)
R²	0.00	0.11	0.06	0.03	0.06
N	10,790	10,781	10,759	10,759	4,845
# Individuals	3,039	3,039	3,039	3,039	2,194
Sig. Different	No	10%	No	5%	1%
On Own Avg.	0.245	0.573	0.498	0.203	0.464
With Others Avg.	0.199	0.518	0.452	0.167	0.333

Panel B: Substance Use at Intensive Margin
	Number Cigarettes	Drinks Several Times per Month	# Drinks per Session	# Times used Marijuana	# Times used Pain Relievers
With Others	0.769 (0.720)	−0.046 (0.034)	0.457^*** (0.155)	1.582 (1.144)	−1.253 (3.352)
On Own	−0.054 (0.696)	−0.012 (0.038)	0.470^** (0.194)	2.382^** (1.094)	−2.548 (1.971)
R²	0.02	0.05	0.03	0.03	0.04
N	1,721	6,883	6,825	5,320	1,595
# Individuals	748	2,516	2,509	2,019	1,095
Sig. Different	No	No	No	No	No
On Own Avg.	8.969	0.545	3.536	13.332	6.609
With Others Avg.	7.246	0.571	3.413	15.010	5.957

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. Living on own is defined as being a head of household or a spouse of the head of household during the survey of the reported job loss. Living with others is defined as living with friends or relatives at the time of job loss.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix A-1

Displacement's Effect on Substance Use

	Prob. of Smoking	Prob. of Drinking	# Drinks per Session	# Times Used Marijuana
≤t-6	−0.037 (0.027)	0.001 (0.037)	0.155 (0.266)	−2.604 (1.893)
t-4	−0.016 (0.018)	0.037 (0.027)	0.021 (0.184)	−1.606 (1.239)
t	0.028^** (0.013)	0.035^* (0.019)	0.529^*** (0.155)	1.486^* (0.844)
t+2	0.023 (0.018)	0.056^** (0.023)	0.380^** (0.169)	1.283 (1.131)
t+4	0.029 (0.023)	0.019 (0.031)	0.376^* (0.201)	2.987^** (1.413)
≥t+6	−0.005 (0.034)	0.030 (0.042)	0.404 (0.305)	2.708 (1.861)
R²	0.00	0.11	0.03	0.03
N	10,790	10,781	6,825	5,320
# Ind.	3,039	3,039	2,509	2,019

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. Estimates come from equation (2). Coefficients are relative to period t-2.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table B-1

The Effect of Substance Use on Employment

	Smoking	Drinking	Marijuana	Other Drugs	Pain Relievers
Panel A: Substance User at t = 0
After	−0.178^*** (0.039)	−0.107^*** (0.024)	−0.122^*** (0.026)	−0.158^*** (0.040)	−0.182^** (0.071)
R²	0.09	0.09	0.09	0.09	0.02
N	8,536	9,809	9,475	8,458	4,002
# Individuals	2,484	2,796	2,719	2,463	1,800
Panel B: Not Substance User at t = 0
After	−0.117^*** (0.023)	−0.186^*** (0.036)	−0.147^*** (0.030)	−0.126^*** (0.023)	−0.141^*** (0.049)
R²	0.09	0.09	0.09	0.09	0.03
N	10,023	8,750	9,076	10,093	4,268
# Individuals	2,845	2,533	2,608	2,864	1,911

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variable equals one if the respondent is employed and zero otherwise. Additional independent variables include age, age squared, and dummy variables for calendar year. Panel A limits the treatment group to those who are substance users at the time of the reported job loss. Panel B limits the treatment group to those who are not current substance users at the time of job loss. The substance in question is in the column headings.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table B-2

The Effect of Substance Use on Employment

	Smoking	Drinking	Marijuana	Other Drugs	Pain Reliever
Panel A: Non-users in t – 2 & Users in t = 0
After	−0.194^** (0.084)	−0.088^** (0.043)	−0.146^** (0.057)	−0.118^* (0.061)	−0.171 (0.106)
R²	0.09	0.09	0.09	0.09	0.02
N	7,962	8,325	8,100	8,056	3,809
# Individuals	2,345	2,443	2,376	2,370	1,729
Panel B: Non-users in t – 2 & t = 0
After	−0.111^*** (0.024)	−0.176^*** (0.040)	−0.128^*** (0.031)	−0.130^*** (0.024)	−0.145^** (0.057)
R²	0.09	0.09	0.09	0.09	0.02
N	9,857	8,512	8,869	9,855	4,031
# Individuals	2,806	2,474	2,557	2,808	1,815

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variable equals one if the respondent is employed and zero otherwise. Additional independent variables include age, age squared, and dummy variables for calendar year. Panel A limits the treatment group to those who started using substances between the survey before job loss and the survey of the reported job loss. Panel B limits the treatment group to those who remained non-users during the survey before job loss and the survey of job loss. The substance in question is in the column headings.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table C-1

Displacement's Effect on Labor Market Outcome - Heterogeneity

	Log Earnings	Real Earnings	Employed
Panel A: Gender
Men	−0.816^*** (0.184)	−2,988.339^*** (983.679)	−0.108^*** (0.028)
Women	−0.694^*** (0.200)	−5573.236^*** (867.409)	−0.161^*** (0.028)
R²	0.16	0.17	0.08
N	8,460	10,816	10,816
# Individuals	2,863	3,039	3,039
Sig. Different	No	5%	No
Panel B: Race
Non-White	−0.946^*** (0.178)	−5,958.624^*** (785.528)	−0.151^*** (0.025)
White	−0.492^** (0.212)	−1481.940 (1177.504)	−0.108^*** (0.031)
R²	0.17	0.17	0.08
N	8,453	10,807	10,807
# Individuals	2,861	3,036	3,036
Sig. Different	10%	1%	No
Panel C: Living Arrangements
With Others	−0.984^*** (0.184)	−5,054.904^*** (795.454)	−0.086^*** (0.026)
On Own	−0.506^** (0.202)	−3360.716^*** (1094.889)	−0.193^*** (0.030)
R²	0.17	0.17	0.09
N	8,460	10,816	10,816
# Individuals	2,863	3,039	3,039
Sig. Different	10%	No	1%

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. Living on own is defined as being a head of household or a spouse of the head of household during the survey of the reported job loss. Living with others is defined as living with friends or relatives at the time of job loss.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table D-1

Heterogeneity - Relative Differences

	# Cigarettes	# Drinks	Marijuana	Pain Relievers
Panel A: Gender
Men	−0.058 (0.206)	0.306^*** (0.100)	0.431 (0.349)	−0.089 (0.638)
Women	−0.076 (0.145)	0.106^* (0.062)	0.332 (0.507)	−0.645 (0.475)
R²	0.04	0.02	0.04	0.06
N	1,263	4,730	2,773	318
# Individuals	458	1,496	896	211
Sig. Different	No	10%	No	No
Panel B: Race
Non-White	−0.064 (0.147)	0.296^*** (0.081)	0.154 (0.311)	−0.285 (0.423)
White	−0.063 (0.231)	0.117 (0.091)	0.735 (0.559)	−0.777 (0.718)
R²	0.04	0.02	0.04	0.06
N	1,259	4,726	2,773	318
# Individuals	457	1,495	896	211
Sig. Different	No	No	No	No
Panel C: Living Arrangements
With Others	0.104 (0.220)	0.212^*** (0.071)	0.282 (0.365)	−0.422^* (0.254)
On Own	−0.230 (0.152)	0.216^** (0.104)	0.502 (0.461)	−0.478 (0.780)
R²	0.04	0.02	0.04	0.06
N	1,263	4,730	2,773	318
# Individuals	458	1,496	896	211
Sig. Different	No	No	No	No

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix E-1

Displacement's Effect on Living Arrangements

	On Own	Transition
After	−0.024 (0.015)	−0.017 (0.013)
R²	0.42	0.09
N	10,816	10,816
# Individuals	3,039	3,039

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix E-2

Non-Random Attrition

	Employed	Current Smoke	Current Drink	Marijuana	Other Drug	Pain Reliever
Attrite	0.006 (0.010)	0.008 (0.016)	−0.015 (0.012)	0.002 (0.013)	0.005 (0.014)	0.000 (0.000)
R²	0.68	0.68	0.69	0.69	0.69	1.00
N	5,725	5,718	5,716	5,707	5,707	884
# Individuals	1,417	1,417	1,417	1,417	1,417	573

The data come from the 2005–2017 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variable equals one if the individual attrites from the sample in the following period. The main independent variable of interest in each regression is indicated in the column. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix E-3

Displacement's Effect on Attrition

	Attrite
After	0.008 (0.016)
R²	0.68
N	5,725
# Individuals	1,417

The data come from the 2005–2017 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variable is noted in the column heading. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table F-1

Substance Use's Effect on Displacement

Panel A: Extensive Margin
	Current Smoke	Current Drink	Ever Used Marijuana	Ever Used Other Drugs	Ever Used Pain Relievers/Narcotics
Job Loss	0.010 (0.022)	−0.012 (0.013)	0.014 (0.014)	0.009 (0.017)	−0.029 (0.026)
R²	0.01	0.01	0.01	0.01	0.00
N	7,768	7,761	7,755	7,755	2,510
# Individuals	3,038	3,037	3,038	3,038	1,793

Panel B: Intensive Margin
	# Cigarettes Smoked per Day	Drink Several Times per Month	# Drinks per Session	# Times Used Marijuana in Past 12 Months	# Times Used Pain Relievers in Past 12 Month
Job Loss	−0.001 (0.003)	−0.008 (0.015)	−0.005 (0.003)	0.000 (0.001)	0.000 (0.004)
R²	0.01	0.01	0.01	0.01	0.02
N	1,263	4,782	4,741	3,631	920
# Individuals	658	2,191	2,179	1,754	761

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table F-2

Sensitivity Analysis - Labor Market Outcomes

	Earnings	Log Earnings	Employed
Panel A: Displaced Main Job
After	−6,673.086^*** (849.543)	−0.863^*** (0.245)	−0.406^*** (0.033)
R²	0.17	0.16	0.10
N	10,816	8,460	10,816
# Individuals	3,039	2,863	3,039
Panel B: Employed at Baseline
After	−4,616.058^*** (949.157)	−0.860^*** (0.196)	−0.141^*** (0.022)
R²	0.23	0.14	0.12
N	5,781	4,914	5,781
# Individuals	1,695	1,655	1,695
Panel C: Works at Least 25 Hours at Baseline
After	−4,896.733^*** (1,456.795)	−0.836^*** (0.203)	−0.179^*** (0.030)
R²	0.12	0.17	0.07
N	4,005	3,589	4,005
# Individuals	989	986	989
Panel D: Displaced Workers Present between t–2 and t+2
After	−4,993.206^*** (848.074)	−0.886^*** (0.167)	−0.117^*** (0.024)
R²	0.17	0.17	0.08
N	10,004	7,802	10,004
# Individuals	2,787	2,617	2,787

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table F-3

Sensitivity Analysis - Substance Use: Extensive Margin

	Currently Smokes	Currently Drinks	Ever Use Marijuana	Ever Use Other Drugs	Ever Use Pain Relievers
Panel A: Displaced from Main Job
After	0.032 (0.022)	−0.020 (0.027)	−0.075^*** (0.025)	−0.011 (0.021)	0.065 (0.068)
R²	0.00	0.11	0.07	0.03	0.06
N	10,790	10,781	10,759	10,759	4,486
# Individuals	3,039	3,039	3,039	3,039	2,004
Panel B: Employed at Baseline Interview
After	0.040^** (0.017)	0.025 (0.025)	−0.031 (0.020)	0.014 (0.020)	0.057 (0.046)
R²	0.00	0.09	0.06	0.03	0.06
N	5,767	5,763	5,758	5,758	2,523
# Individuals	1,695	1,695	1,695	1,695	1,147
Panel C: Worked at Least 25 Hours per Week at Baseline Interview
After	0.049^** (0.020)	0.020 (0.026)	−0.012 (0.023)	0.012 (0.024)	0.032 (0.070)
R²	0.01	0.07	0.06	0.04	0.05
N	3,999	3,996	3,994	3,994	1,098
# Individuals	989	989	989	989	475
Panel D: Displaced Workers Present between t – 2 and t + 2
After	0.028^* (0.015)	0.019 (0.021)	−0.034^** (0.017)	−0.001 (0.017)	0.043 (0.055)
R²	0.00	0.11	0.06	0.03	0.06
N	9,979	9,970	9,950	9,950	4,132
# Individuals	2,787	2,787	2,787	2,787	1,849

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table F-4

Sensitivity Analysis - Substance Use: Intensive Margin

	# Cigarettes Smoked per Day	Drinks Several Times per Month	# Drinks Consumed per Session	# Times Used Marijuana past 12 Months	# Times used Pain Relievers/Narcotics past 12 Months
Panel A: Displaced from Main Job
After	1.503^* (0.841)	−0.036 (0.046)	0.785^*** (0.230)	1.287 (1.371)	−1.451 (2.836)
R²	0.03	0.05	0.03	0.02	0.04
N	1,721	6,883	6,825	5,320	1,419
# Individuals	748	2,516	2,509	2,019	978
Panel B: Employed at Baseline Interview
After	0.412 (0.623)	−0.016 (0.035)	0.560^*** (0.172)	1.396 (1.134)	−4.750 (3.226)
R²	0.05	0.05	0.03	0.03	0.05
N	918	3,790	3,762	2,868	765
# Individuals	411	1,432	1,429	1,111	545
Panel C: Worked at Least 25 Hours per Week at Baseline Interview
After	−0.084 (0.701)	0.008 (0.039)	0.669^*** (0.210)	0.652 (1.171)	−3.729 (2.969)
R²	0.03	0.03	0.04	0.02	0.13
N	758	2,805	2,785	2,186	417
# Individuals	310	885	887	714	286
Panel D: Displaced Workers Present from t – 2 and t + 2
After	0.589 (0.665)	−0.055^* (0.031)	0.454^*** (0.151)	0.824 (0.950)	−1.635 (2.576)
R²	0.03	0.05	0.03	0.02	0.04
N	1,539	6,362	6,307	4,883	1,315
# Individuals	665	2,299	2,292	1,834	901

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table G-1

Displacement's Effect on Labor Market Outcomes

	Earnings	Log Earnings	Employed
After	−3,635.597^*** (1,339.037)	−0.274 (0.262)	−0.137^*** (0.040)
R²	0.17	0.18	0.09
N	8,422	6,505	8,422
# Individuals	2,463	2,297	2,463

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. The treatment group is limited to those displaced through plant closure only.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table G-2

Displacement's Effect on Substance Use

Panel A: Extensive Margin
	Currently Smokes	Currently Drinks	Ever Used Marijuana	Ever Used Other Drugs	Ever Used Pain Relievers/Narcotics
After	0.002 (0.024)	0.047 (0.032)	−0.041 (0.030)	−0.023 (0.024)	−0.039 (0.072)
R²	0.00	0.11	0.07	0.02	0.06
N	8,403	8,395	8,374	8,374	3,907
# Individuals	2,463	2,463	2,463	2,463	1,779

Panel B: Intensive Margin
	# Cigarettes Smoked Per Day	Drinks Several Times per Month	# Drinks per Session	# Times Used Marijuana in Past 12 Months	# Times Used Pain Relievers Past 12 Months
After	0.047 (0.841)	−0.026 (0.046)	0.137 (0.237)	1.770 (1.694)	−5.535^* (2.913)
R²	0.03	0.06	0.03	0.02	0.04
N	1,156	5,363	5,319	3,991	1,221
# Individuals	531	2,010	2,003	1,580	860

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. The treatment group is limited to those displaced through plant closure only.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table G-3

Heterogeneity by Gender

Panel A: Labor Market Outcomes
	Log Earnings	Real	Employed
Men	−0.908^** (0.368)	−3,174.104 (2,223.554)	−0.077 (0.055)
Women	0.225 (0.354)	−4002.388^*** (1513.980)	−0.184^*** (0.054)
R²	0.18	0.17	0.09
N	6,505	8,422	8,422
# Individuals	2,297	2,463	2,463
Sig. Different	5%	No	No
Panel B: Substance Use at Extensive Margin
	Currently Smokes	Currently Drinks	Ever Used Marijuana	Ever Used Other Drugs	Ever Used Pain Relievers/Narcotics
Men	−0.001 (0.036)	0.008 (0.041)	−0.061 (0.039)	−0.022 (0.041)	0.047 (0.099)
Women	0.004 (0.031)	0.078^* (0.045)	−0.024 (0.042)	−0.024 (0.028)	−0.099 (0.098)
R²	0.00	0.11	0.07	0.02	0.06
N	8,403	8,395	8,374	8,374	3,907
# Individuals	2,463	2,463	2,463	2,463	1,779
Sig. Different	No	No	No	No	No
Panel C: Substance Use at Intensive Margin
	# Cigarettes Smoked Per Day	Drinks Several Times per Month	# Drinks per Session	# Times Used Marijuana in Past 12 Months	# Times Used Pain Relievers Past 12 Months
Men	0.667 (1.294)	−0.054 (0.064)	0.624^* (0.365)	2.149 (2.545)	−7.610^** (3.573)
Women	−0.743 (0.807)	0.002 (0.065)	−0.347 (0.265)	1.382 (2.152)	−4.688 (3.749)
R²	0.03	0.06	0.04	0.02	0.04
N	1,156	5,363	5,319	3,991	1,221
# Individuals	531	2,010	2,003	1,580	860
Sig. Different	No	No	5%	No	No

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. The treatment group is limited to those displaced through plant closure only.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table G-4

Heterogeneity by Race

Panel A: Labor Market Outcomes
	Log Earnings	Real Earnings	Employed
Non-White	−0.504 (0.334)	−6,288.925^*** (1,392.415)	−0.145^*** (0.052)
White	−0.050 (0.406)	546.915 (2510.808)	−0.125^** (0.058)
R²	0.18	0.17	0.09
N	6,501	8,416	8,416
# Individuals	2,296	2,461	2,461
Sig. Different	No	5%	No
Panel B: Substance Use at Extensive Margin
	Currently Smokes	Currently Drinks	Ever Used Marijuana	Ever Used Other Drugs	Ever Used Pain Relievers/Narcotics
Non-White	0.038 (0.030)	0.048 (0.041)	−0.049 (0.040)	−0.018 (0.031)	−0.082 (0.085)
White	−0.060 (0.037)	0.048 (0.047)	−0.034 (0.041)	−0.041 (0.038)	0.060 (0.129)
R²	0.00	0.11	0.07	0.02	0.06
N	8,398	8,390	8,369	8,369	3,906
# Individuals	2,461	2,461	2,461	2,461	1,778
Sig. Different	5%	No	No	No	No
Panel C: Substance Use at Intensive Margin
	# Cigarettes Smoked Per Day	Drinks Several Times per Month	# Drinks per Session	# Times Used Marijuana in Past 12 Months	# Times Used Pain Relievers Past 12 Months
Non-White	−0.013 (1.351)	−0.012 (0.067)	0.468 (0.325)	5.123^** (2.483)	−5.342 (4.714)
White	0.223 (1.032)	−0.039 (0.063)	−0.294 (0.318)	−2.184 (1.988)	−5.776^** (2.575)
R²	0.03	0.06	0.04	0.03	0.04
N	1,152	5,359	5,315	3,988	1,221
# Individuals	530	2,009	2,002	1,579	860
Sig. Different	No	No	10%	5%	No

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. The treatment group is limited to those displaced through plant closure only.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table G-5

Heterogeneity by Living Arrangements

Panel A: Labor Market Outcomes
	Log Earnings	Real Earnings	Employed
With Others	−0.838^** (0.338)	−5,882.592^*** (1,195.762)	−0.127^*** (0.047)
On Own	0.669^* (0.385)	201.614 (2794.817)	−0.153^** (0.069)
R²	0.18	0.17	0.09
N	6,505	8,422	8,422
# Individuals	2,297	2,463	2,463
Sig. Different	1%	5%	No
Panel B: Substance Use at Extensive Margin
	Currently Smokes	Currently Drinks	Ever Used Marijuana	Ever Used Other Drugs	Ever Used Pain Relievers/Narcotics
With Others	0.004 (0.028)	0.053 (0.040)	−0.072^** (0.034)	0.003 (0.030)	0.068 (0.101)
On Own	−0.001 (0.042)	0.036 (0.049)	0.012 (0.054)	−0.067^* (0.040)	−0.161^* (0.097)
R²	0.00	0.11	0.07	0.02	0.06
N	8,403	8,395	8,374	8,374	3,907
# Individuals	2,463	2,463	2,463	2,463	1,779
Sig. Different	No	No	No	No	10%
Panel C: Substance Use at Intensive Margin
	# Cigarettes Smoked Per Day	Drinks Several Times per Month	# Drinks per Session	# Times Used Marijuana in Past 12 Months	# Times Used Pain Relievers Past 12 Months
With Others	0.879 (1.199)	−0.067 (0.059)	0.084 (0.268)	1.696 (2.148)	−0.860 (1.539)
On Own	−1.190 (0.795)	0.041 (0.068)	0.225 (0.431)	1.901 (2.649)	−13.398^** (5.556)
R²	0.03	0.06	0.03	0.02	0.05
N	1,156	5,363	5,319	3,991	1,221
# Individuals	531	2,010	2,003	1,580	860
Sig. Different	No	No	No	No	5%

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. The treatment group is limited to those displaced through plant closure only.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table H-1

Descriptive Statistics (means) of Selected Variables

	Non-Displaced (matched)	Displaced (matched)
Total Earnings^a,b	$3,094.56	$2,214.53
Hours Worked^b	18.43	21.14
Employed	54.13%	55.35%
Age	18.93	18.78
Female	53.59%	50.20%
White^c	35.86%	36.40%
Responsible for Own Household	15.02%	14.88%
Education
< High School	13.86%	13.69%
High School Degree/GED	23.41%	29.700%
Some College/Associates Degree	61.14%	56.38%
Bachelor's Degree	1.36%	0.23%
Master's Degree	0.23%	-
Currently Smoke	18.94%	21.65%
Number Cigarettes/Day	8.49	8.72
Currently Drink Alcohol	44.25%	46.55%
Drink at Least Several Times/Month	50.15%	50.29%
Number Drinks/Session	3.71	3.64
Ever Used Marijuana	43.55%	46.07%
Number of Uses of Marijuana previous 12 Months^d	11.82	14.4
Ever Used Amphetamines, Cocaine, Barbiturates, or Tranquilizers	16.42%	15.86%
Ever Used Pain Relievers/Narcotics	27.93%	34.43%
Number of Uses of Pain Relievers/Narcotics previous	9.39	6.17
12 Months^d
Total Number of Individuals	739	739

Notes: Sample is the matched treatment and control units. See main text for description of matching procedure. All years are used in the calculations. However, calculations only use the first observation in which an individual is observed.

a – Real 2018 dollars;

b – equals the total across all five jobs;

c – represents the modal report for race;

d – midpoint of the range is used in the calculation

Appendix Table H-2

Displacement's Effect on Labor Market Outcomes

	Earnings	Log Earnings	Employed
After	−1,991.898^*** (663.265)	−0.647^*** (0.138)	−0.112^*** (0.020)
R²	0.20	0.15	0.08
N	6,043	4,760	6,043
# Individuals	1,307	1,263	1,307

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. See the main text for a description of the matching estimator.

* p<0.1;

** p<0.05;

*** p<0.01

Appendix Table H-3

Displacement's Effect on Substance Use

Panel A: Substance Use at Extensive Margin
	Current Smokes	Current Drink	Ever Used Marijuana	Ever Used Other Drugs	Ever Used Pain Reliever/Narcotics
After	0.024^* (0.013)	0.020 (0.018)	0.002 (0.015)	0.024^* (0.014)	−0.007 (0.030)
R²	0.00	0.10	0.05	0.02	0.07
N	6,034	6,028	6,015	6,015	1,883
# Individuals	1,307	1,307	1,307	1,307	695

Panel B: Substance Use at Intensive Margin
	# Cigarettes Smoked per Day	Drinks Several Times per Month	# Drinks per Session	# Times Used Marijuana Past 12 Months	# Times Used Pain Relievers Past 12 Months
After	0.462 (0.548)	−0.002 (0.024)	0.389^*** (0.116)	1.097 (0.756)	−1.776 (1.153)
R²	0.03	0.03	0.03	0.03	0.04
N	1,285	3,773	3,741	3,183	646
# Individuals	432	1,114	1,114	942	361

The data come from the 2005–2019 waves of the TAS from the PSID. The sample includes individuals who respond to at least two surveys. Displaced workers must respond to the survey prior to job loss. Standard errors clustered at the individual level shown in parentheses. The dependent variables are noted in the column headings. Additional independent variables include age, age squared, and dummy variables for calendar year. See the main text for a description of the matching estimator.

* p<0.1;

** p<0.05;

*** p<0.01

Young adult substance use following involuntary job loss

Figures & Tables

Table 1

Table 2

Figure 1

Table 3

Figure 2

Table 4

Table 5

Table 6

Table 7

Table 8

Table 9

Appendix A-1

Appendix Table B-1

Appendix Table B-2

Appendix Table C-1

Appendix Table D-1

Appendix E-1

Appendix E-2

Appendix E-3

Appendix Table F-1

Appendix Table F-2

Appendix Table F-3

Appendix Table F-4

Appendix Table G-1

Appendix Table G-2

Appendix Table G-3

Appendix Table G-4

Appendix Table G-5

Appendix Table H-1

Appendix Table H-2

Appendix Table H-3

Figure H-1

Figure H-2

Paradigm

My account