Value of Working Conditions: Maestas et al. (2023)

What this is. This is a distilled skeleton of Maestas et al. (2023). Read the original at https://doi.org/10.1257/aer.20190846 to replicate or extend.

TL;DR

Maestas et al. (2023) field the American Working Conditions Survey (AWCS), a new nationally representative survey covering 1,738 employed Americans, and use embedded stated-preference experiments to estimate how much workers are willing to pay for nine nonwage job amenities. Across all amenities, switching from the worst to the best job is equivalent to a 55 percent wage increase, confirming that nonwage job attributes are a central component of total compensation. Workers differ widely in their valuations by gender, race, education, and age: older workers and women place especially high value on physical job demands and paid time off. Incorporating both the incidence and the valuation of amenities into standard wage differentials attenuates the gender gap (24 percent) but widens the race and education gaps, and raises overall wage inequality. The paper builds on the compensating differentials framework of Rosen (1986) and extends the experimental approach of Mas and Pallais (2017) to a nationally representative sample and a broader set of amenities. Related prior work includes Wiswall and Zafar (2018) on stated-preference evidence for job attributes among students, and Pierce (2001) on compensation inequality once fringe benefits are included.

Core results

#	Result	Locator	Magnitude as reported
R1	Best vs worst amenity bundle: total WTP	Table 2 col 5, p. 2025	55.0% wage equivalent
R2	Paid time off: WTP for 10 days vs none	Table 2 col 5, p. 2025	16.4% wage equivalent
R3	Paid time off: WTP for 20 days vs none	Table 2 col 5, p. 2025	23.0% wage equivalent
R4	Physical demands: WTP for moderate vs heavy activity	Table 2 col 5, p. 2025	14.5% wage equivalent
R5	Schedule flexibility: WTP for setting own schedule	Table 2 col 5, p. 2025	8.9% wage equivalent
R6	Work arrangement: WTP for working alone (vs team-evaluated team)	Table 2 col 5, p. 2025	8.6% wage equivalent
R7	Gender log compensation gap with preference heterogeneity	Table 8 Panel A col 3, p. 2040	-0.142 log pts (vs -0.192 unadjusted); 24% reduction
R8	Race log compensation gap with preference heterogeneity	Table 8 Panel A col 3, p. 2040	-0.274 log pts (vs -0.208 unadjusted); 27% widening
R9	Education log compensation gap (HS or less vs college)	Table 8 Panel A col 3, p. 2040	-0.667 log pts (vs -0.559 unadjusted); 19% widening
R10	Overall wage inequality: 90-10 log wage gap	Table 8 Panel C col 3, p. 2040	1.769 (vs 1.664 unadjusted); +10.5 log pts
R11	Other amenity valuations: telecommuting, sitting, relaxed pace, own-performance evaluation, work autonomy, training, and positive community impact	Table 2 col 5, p. 2025	4.2%, 11.6%, 4.3%, 6.5%, 4.0%, 5.4%, and 3.6% of wage, respectively
R12	Revealed-preference illustration: choosing a job with dominating amenities as its relative wage changes	Figure 1, p. 2023	20% choose the amenity-dominant job at a 40% lower wage; all attentive respondents choose it when it pays 40% more
R13	Preference sorting by current amenity status and subsequent job transition	Table 3, pp. 2031-2032	Current holders value amenities 6.1 pp more on average; among initial holders, those retaining an amenity value it 4.1 pp more (p < 0.01); among initial nonholders, difference is 1.7 pp and not significant
R14	Gender heterogeneity in amenity valuations	Table 4, p. 2034	Best-to-worst WTP: 58.8% of wage for women vs 51.7% for men; moderate activity 18.4% vs 11.4%; 10 days PTO 18.7% vs 14.6%
R15	Racial heterogeneity in amenity valuations	Table 5, p. 2035	Best-to-worst WTP: 45.5% for non-White workers vs 57.0% for White workers; own schedule 3.8% vs 10.1%; autonomy -0.8% vs 5.1%
R16	Education gradient in amenity valuations	Table 6, p. 2036	Best-to-worst WTP: 48.0% (high school or less), 54.3% (some college), and 60.0% (college degree); telecommuting: -0.5%, 4.6%, and 6.9%
R17	Age heterogeneity in amenity valuations	Table 7, pp. 2037-2038	Best-to-worst WTP: 47.5%, 53.5%, 58.9%, and 74.5% across ages 25-34, 35-49, 50-61, and 62+; moderate-activity WTP rises from 8.8% to 30.5%
R18	Age compensation differential after preference adjustment	Table 8 Panel A col 3, p. 2040	Under-35 vs age 62+: -0.314 log points adjusted vs -0.143 log points in wages; age 35-49: -0.217 vs -0.081; age 50-61: -0.140 vs -0.038
R19	Interindustry compensation dispersion after amenity adjustment	Table 8 Panel B cols 1 and 3, p. 2040	Employment-weighted standard deviation rises from 0.130 (wages) to 0.155 (compensation allowing valuations to vary)
R20	Wage inequality below and above the median after amenity adjustment	Table 8 Panel C, p. 2040	90-50 gap rises 0.963 to 1.011; 50-10 gap rises 0.701 to 0.759 log points when valuations vary
R21	Valuation of weekly hours relative to total earnings	Figure 3, pp. 2029-2030	Workers require 40.4% lower earnings to accept a 50% reduction in hours (40 to 20); require a 57.7% increase for a 50% increase in hours (40 to 60)
R22	Incidence and distribution of working conditions in the 2015 US sample	Table 1, pp. 2013-2014	56.5% set their own schedule, 36.4% telecommute, 42.9% mostly sit, 59.7% have 15+ PTO days; mean hourly wage $30.30 (2015 dollars), N = 1,738
R23	Some-college compensation differential after preference adjustment	Table 8 Panel A col 3, p. 2040	-0.564 log points adjusted vs -0.502 wage
R24	Robustness of full-bundle and paid-time-off WTP across logit specifications	Table 2 cols 3-7, p. 2025	Best-to-worst: 50.7% mixed-logit mean, 56.9% median, 55.0% standard-logit full sample, 60.3% attentive, 49.9% with interactions; 10-day PTO: 17.3%, 17.8%, 16.4%, 17.4%, 16.3%; 20-day PTO: 23.7%, 24.9%, 23.0%, 25.6%, 22.9%

Overall (paper’s conclusion). Working conditions vary widely across demographic groups and throughout the wage distribution. Workers have measurable willingness to pay for most job amenities studied. Accounting for both the incidence of amenities and heterogeneity in valuations changes standard measures of the wage structure: the gender gap narrows, the race and education gaps widen, and overall wage inequality increases. Contrary to the conclusion of Krueger and Summers (1988), accounting for the value of working conditions widens rather than narrows interindustry wage differentials.

Theory / model

The paper applies the competitive compensating-differentials framework of Rosen (1986): workers sort among jobs by trading wages against nonwage working conditions, and observed wages alone therefore omit amenity value. The paper does not estimate a general-equilibrium sorting model. Its individual-level indirect utility model for respondent i, alternative j in choice pair t is $V_{ijt}=\alpha+A_{ijt}'\beta_i+\delta_i\ln w_{ijt}+\varepsilon_{ijt},$ where the attribute and wage coefficients can vary across workers. With iid Type-I extreme-value errors, the implied binary logit choice probability is $\Pr(V_{ijt}>V_{ikt})=\frac{\exp\{(A_{ijt}'-A_{ikt}')\beta_i+\delta_i(\ln w_{ijt}-\ln w_{ikt})\}}{1+\exp\{(A_{ijt}'-A_{ikt}')\beta_i+\delta_i(\ln w_{ijt}-\ln w_{ikt})\}}.$ These model expressions are printed in Section III, p. 2021. Stated-preference experiments randomly vary wages and amenities within choice pairs, identifying the wage-amenity tradeoff from respondent choices.

Method

For an amenity r, the paper defines WTP by equating utility at the current wage without that amenity to utility with the amenity and a lower wage. The three numbered main-text equations are reproduced below as printed in Section III, p. 2022. The signs and grouping in the exponents follow the PDF.

\delta_i\ln w_i=\beta_i^r+\delta_i\ln[w_i-WTP_i^r] \tag{1}

WTP_i^r=w_i\left[1-e^{-\beta_i^r/\delta_i}\right] \tag{2}

WTP_i^{\text{FULL}}=w_i\left[1-e^{-\sum_{r=1}^{R}\beta_i^r/\delta_i}\right] \tag{3}

For attributes with multiple possible values, the full-bundle calculation uses the coefficient for the most preferred value to avoid double counting. The standard logit restricts coefficients to be common across respondents; the mixed logit allows random coefficients. WTP standard errors use the delta method and are clustered by respondent unless noted otherwise.

Empirical specifications

Main stated-preference estimates. Respondents completed 10 choices each, with two nonwage attributes and wages randomized within pairs. The estimating choice probability is the logit equation stated above (Section III, p. 2021). The binary dependent variable records preference for Job A. The main estimates use survey weights and cluster standard errors by respondent; the standard logit sample is 17,380 choice pairs (Table 2, p. 2025). The paper also estimates a mixed logit, an attentive-respondent subsample, and a two-way amenity interaction specification. Online Appendix Table 4 reports probit, common-baseline, unweighted, and hours-control checks; the authors state that WTP estimates are robust across these alternatives (text p. 2022 and note 28, p. 2028).

Observational hedonic comparison. Table 2 columns 1-2 compare the experimental estimates with cross-sectional and first-difference log-wage regressions. Written out from Section IV A, p. 2025:

$\ln w_{it}=\alpha+\sum_r\gamma_r A_{ir}+X_i'\theta+u_i.$

For the linked panel, the first-difference estimator removes time-invariant worker heterogeneity; written out from the same specification:

$\Delta\ln w_{it}=\sum_r\gamma_r\Delta A_{ir}+\Delta X_i'\theta+\Delta u_i.$

where X includes age-group, race, education, and citizenship indicators in the cross section; the panel specification first-differences the matched AWCS observations to remove time-invariant worker attributes. It uses 1,737 cross-sectional respondents and 977 matched panel respondents (Table 2, p. 2025). These are observational comparisons, not the primary identification design.

Demographic heterogeneity and sorting. The standard logit is re-estimated within gender, race, education, and age groups (Tables 4-7, pp. 2034-2038). Table 3 interacts experimental WTP with 2015 amenity status, and then with 2018 amenity status for the linked follow-up sample (N = 9,770 choice-pair observations, 977 respondents; pp. 2031-2032). WTP standard errors are clustered by respondent; average differences use the delta method. Written out from the interacted utility specification described in Section IV B, p. 2031:

$V_{ijt}=\alpha+A_{ijt}'\beta+H_{i,2015}A_{ijt}'\gamma+\delta\ln w_{ijt}+\varepsilon_{ijt},$

where H records whether the respondent has the amenity in the 2015 wave. The follow-up tests condition these interactions on the 2018 amenity status as shown in Table 3 (pp. 2031-2032).

Compensation regressions. For the compensation analysis, log wage or log total compensation is regressed on demographic indicators and industry indicators. Written out from Section V, pp. 2039-2040, the demographic specification is

$\ln C_i=\sum_k\phi_kD_{ik}+e_i,$

where C is observed wages or adjusted total compensation and D are the demographic group indicators. For the industry analysis, log compensation is demeaned and regressed on the 11 industry supersector indicators with no constant:

$\widetilde{\ln C_i}=\sum_{s=1}^{11}\phi_sD_{is}+e_i.$

Total compensation adds the estimated value of a worker’s current amenities to wages, using equation below (p. 2039; the PDF does not number it):

$\ln\left(w_i+w_i\left[1-e^{-\sum_{r=1}^{R}A_{ir}\beta_i^r/\delta_i}\right]\right).$

The demographic and wage-percentile analyses use N = 1,738; the industry analysis uses N = 1,528. Standard errors and confidence intervals come from a 500-iteration respondent block bootstrap (Table 8 note, p. 2040). There are no fixed effects in the demographic regressions; the industry regression demeans within supersector.

Work-hours extension. The hours analysis replaces hourly wage with total earnings and adds hours indicators from 5 to 60, with 40 hours omitted. Figure 3 reports valuations relative to earnings with 95% confidence intervals clustered by respondent (text pp. 2029-2030). The hours-control utility model is stated in footnote 29, p. 2030:

$V_{ijt}=\alpha+A_{ijt}'\beta+\delta\ln e_{ijt}+H_{ijt}'\theta+u_{ijt},$

where e is offered total earnings and H is the vector of hours indicators.

Datasets used

Dataset	Role in paper	Wiki page
American Working Conditions Survey (AWCS), waves 2015 and 2016	Primary data source: incidence of 9 job attributes by demographic group (wave 1, July-October 2015); stated-preference experiments for WTP estimation (wave 2, December 2015-February 2016); N = 1,738 workers	no page yet
American Working Conditions Survey (AWCS), follow-up wave 2018	Longitudinal follow-up for sorting validation and preference-transition analysis; N = 977 matched respondents	no page yet
RAND American Life Panel (ALP)	Nationally representative probability-based panel that served as the sampling frame for all AWCS waves	no page yet
Current Population Survey (CPS)	Used to generate survey weights matching AWCS to US working population demographics	no page yet

Sample: 1,738 employed workers ages 25-71, from the RAND ALP, weighted to match the US working population via CPS. The AWCS data are available publicly at https://www.rand.org/pubs/tools/TL269.html. Replication data are archived at ICPSR (Maestas et al. 2023, DOI 10.3886/E184378V1).

When to read the full paper

Read the source if you are:

• Estimating compensating wage differentials or the value of specific job amenities (Tables 2-7 provide WTP estimates by amenity and demographic group with standard errors).
• Adjusting wage gaps (gender, race, education, interindustry) for nonwage job attributes; Table 8 and Section V detail the methodology and results.
• Designing stated-preference experiments for labor market research; Sections III-IV provide the experimental design, logit estimation, and robustness checks including attention screens, probit alternatives, and common-baseline variants.
• Studying heterogeneity in labor market preferences; Tables 4-7 present results by gender, race, education, and age with cross-group p-values.

Attribution and rights

This page is a distilled extract. The source paper is:

Maestas, Nicole, Kathleen J. Mullen, David Powell, Till von Wachter, and Jeffrey B. Wenger. 2023. “The Value of Working Conditions in the United States and Implications for the Structure of Wages.” American Economic Review 113(7): 2007-2047. https://doi.org/10.1257/aer.20190846

Copyright American Economic Association; reproduced with permission. Extract only; full text available at aeaweb.org. This summary is LLM-distilled by IAR, not human-verified, and not reproduced.