Online images amplify gender bias.

Each year, people spend less time reading and more time viewing images1, which are proliferating online2-4. Images from platforms such as Google and Wikipedia are downloaded by millions every day2,5,6, and millions more are interacting through social media, such as Instagram and TikTok, that primarily consist of exchanging visual content. In parallel, news agencies and digital advertisers are increasingly capturing attention online through the use of images7,8, which people process more quickly, implicitly and memorably than text9-12. Here we show that the rise of images online significantly exacerbates gender bias, both in its statistical prevalence and its psychological impact. We examine the gender associations of 3,495 social categories (such as 'nurse' or 'banker') in more than one million images from Google, Wikipedia and Internet Movie Database (IMDb), and in billions of words from these platforms. We find that gender bias is consistently more prevalent in images than text for both female- and male-typed categories. We also show that the documented underrepresentation of women online13-18 is substantially worse in images than in text, public opinion and US census data. Finally, we conducted a nationally representative, preregistered experiment that shows that googling for images rather than textual descriptions of occupations amplifies gender bias in participants' beliefs. Addressing the societal effect of this large-scale shift towards visual communication will be essential for developing a fair and inclusive future for the internet.

Monitoring hiring discrimination through online recruitment platforms.

Women (compared to men) and individuals from minority ethnic groups (compared to the majority group) face unfavourable labour market outcomes in many economies1,2, but the extent to which discrimination is responsible for these effects, and the channels through which they occur, remain unclear3,4. Although correspondence tests5-in which researchers send fictitious CVs that are identical except for the randomized minority trait to be tested (for example, names that are deemed to sound 'Black' versus those deemed to sound 'white')-are an increasingly popular method to quantify discrimination in hiring practices6,7, they can usually consider only a few applicant characteristics in select occupations at a particular point in time. To overcome these limitations, here we develop an approach to investigate hiring discrimination that combines tracking of the search behaviour of recruiters on employment websites and supervised machine learning to control for all relevant jobseeker characteristics that are visible to recruiters. We apply this methodology to the online recruitment platform of the Swiss public employment service and find that rates of contact by recruiters are 4-19% lower for individuals from immigrant and minority ethnic groups, depending on their country of origin, than for citizens from the majority group. Women experience a penalty of 7% in professions that are dominated by men, and the opposite pattern emerges for men in professions that are dominated by women. We find no evidence that recruiters spend less time evaluating the profiles of individuals from minority ethnic groups. Our methodology provides a widely applicable, non-intrusive and cost-efficient tool that researchers and policy-makers can use to continuously monitor hiring discrimination, to identify some of the drivers of discrimination and to inform approaches to counter it.

Drug Overdose Mortality by Usual Occupation and Industry: 46 U.S. States and New York City, 2020.

Objective-This report describes deaths from drug overdoses in 2020 in U.S. residents in 46 states and New York City by usual occupation and industry. Methods-Frequencies, death rates, and proportionate mortality ratios (PMRs) are presented using the 2020 National Vital Statistics System mortality data file. Data were restricted to decedents aged 16-64 for rates and 15-64 for PMRs with usual occupations and industries in the paid civilian workforce. Age-standardized drug overdose death rates were estimated for usual occupation and industry groups overall, and age-adjusted drug overdose PMRs were estimated for each usual occupation and industry group overall and by sex, race and Hispanic-origin group, type of drug, and drug overdose intent. Age-adjusted drug overdose PMRs were also estimated for individual occupations and industries. Results-Drug overdose mortality varied by usual occupation and industry. Workers in the construction and extraction occupation group (162.6 deaths per 100,000 workers, 95% confidence interval: 155.8-169.4) and construction industry group (130.9, 126.0-135.8) had the highest drug overdose death rates. The highest group-level drug overdose PMRs were observed in decedents in the construction and extraction occupation group and the construction industry group (145.4, 143.6-147.1 and 144.9, 143.2-146.5, respectively). Differences in drug overdose PMRs by usual occupation and industry group were observed within each sex, within each race and Hispanicorigin group, by drug type, and by drug overdose intent. Among individual occupations and industries, the highest drug overdose PMRs were observed in decedents who worked as fishers and related fishing occupations and in fishing, hunting, and trapping industries (193.1, 166.8-222.4 and 186.5, 161.7-214.1, respectively). Conclusions-Variation in drug overdose death rates and PMRs by usual occupation and industry in 2020 demonstrates the disproportionate burden of the ongoing drug overdose crisis on certain sectors of the U.S. workforce.

Heterogeneity in banker culture and its influence on dishonesty.

The social sciences are going through what has been described as a 'reproducibility crisis'1,2. Highly influential findings derived from accessible populations, such as laboratories and crowd-sourced worker platforms, are not always replicated. Less attention has been given to replicating findings that are derived from inaccessible populations, and recent high-profile replication attempts explicitly excluded such populations3. Pioneering experimental work4 offered a rare glimpse into banker culture and found that bankers, in contrast to other professionals, are more dishonest when they think about their job. Given the importance of the banking sector, and before academics or policy-makers rely on these findings as an accurate diagnosis of banking culture, an exploration of their generalizability is warranted. Here we conduct the same incentivized task with bankers and non-bankers from five different populations across three continents (n = 1,282 participants). In our banker studies in the Middle East and Asia Pacific (n = 148 and n = 620, respectively), we observe some dishonesty, although-in contrast to the original study4-this was not significantly increased among bankers primed to think about their work compared to bankers who were not primed. We also find that inducing non-banking professionals to think about their job does not have a significant effect on honesty. We explore sampling and methodological differences to explain the variation in findings in relation to bankers and identify two key points. First, the expectations of the general population regarding banker behaviour vary across jurisdictions, suggesting that banking culture in the jurisdiction of the original study4 may not be consistent worldwide. Second, having approached 27 financial institutions, many of which expressed concerns of adverse findings, we expect that only banks with a sound culture participated in our study. The latter introduces possible selection bias that may undermine the generalizability of any similar field study. More broadly, our study highlights the complexity of undertaking a high-fidelity replication of sensitive, highly publicized fieldwork with largely inaccessible populations resulting from institutional and geographical barriers. For policy-makers, this work suggests that caution should be exercised in generalizing the findings of the original study4 to other populations.

The scientific value of numerical measures of human feelings.

Human feelings measured in integers (my happiness is an 8 out of 10, my pain 2 out of 6) have no objective scientific basis. They are "made-up" numbers on a scale that does not exist. Yet such data are extensively collected-despite criticism from, especially, economists-by governments and international organizations. We examine this paradox. We draw upon longitudinal information on the feelings and decisions of tens of thousands of randomly sampled citizens followed through time over four decades in three countries (n = 700,000 approximately). First, we show that a single feelings integer has greater predictive power than does a combined set of economic and social variables. Second, there is a clear inverse relationship between feelings integers and subsequent get-me-out-of-here actions (in the domain of neighborhoods, partners, jobs, and hospital visits). Third, this feelings-to-actions relationship takes a generic form, is consistently replicable, and is fairly close to linear in structure. Therefore, it seems that human beings can successfully operationalize an integer scale for feelings even though there is no true scale. How individuals are able to achieve this is not currently known. The implied scientific puzzle-an inherently cross-disciplinary one-demands attention.

The intergenerational sources of the U-turn in gender segregation.

In the early 1970s, the balkanization of the US labor market into "men's occupations" and "women's occupations" began to unravel, as women entered the professions and other male-typed sectors in record numbers. This decline in gender segregation continued on for several decades but then suddenly stalled at the turn of the century and shows no signs of resuming. Although the stall is itself undisputed, its sources remain unclear. Using nearly a half-century of data from the General Social Survey, we show that a resurgence in segregation-inducing forms of intergenerational transmission stands behind the recent stall. Far from serving as impartial conduits, fathers are now disproportionately conveying male-typed occupations to their sons, whereas mothers are effectively gender-neutral in their transmission outcomes. This segregative turn among fathers accounts for 47% of the stall in the gender segregation trend (between 2000 and 2018), while the earlier integrative turn among fathers accounts for 34% of the initial downturn in segregation (between 1972 and 1999). It follows that a U-turn in intergenerational processes lies behind the U-turn in gender segregation.

Unexpected employee location is associated with injury during robberies.

Millions of employees are victims of violent crimes at work every year, particularly those in the retail industry, who are frequent targets of robbery. Why are some employees injured while others escape from these incidents physically unharmed? Departing from prevailing models of workplace violence, which focus on the static characteristics of perpetrators, victims, and work environments, we examine why and when injuries during robberies occur. Our multimethod investigation of convenience-store robberies sought evidence from detailed coding of surveillance videos and matched archival data, preregistered experiments with formerly incarcerated individuals and customer service personnel, and a 3-y longitudinal intervention study in the field. While standard retail-industry safety protocols encourage employees to be out from behind the cash register area to be safer, we find that robbers are significantly more likely to injure or kill employees who are located there (versus behind the cash register area) when a robbery begins. A 3-y field study demonstrates that changing the safety training protocol-through providing employees with a behavioral script to follow should a robbery begin when they are on the sales floor-was associated with a significantly lower rate of injury during these robberies. Our research establishes the importance of understanding the interactive dynamics of workplace violence, crime, and conflict.

Rapid wage growth at the bottom has offset rising US inequality.

US earnings inequality has not increased in the last decade. This marks the first sustained reversal of rising earnings inequality since 1980. We document this shift across eight data sources using worker surveys, employer-reported data, and administrative data. The reversal is due to a shrinking gap between low-wage and median-wage workers. In contrast, the gap between top and median workers has persisted. Rising pay for low-wage workers is not mainly due to the changing composition of workers or jobs, minimum wage increases, or workplace-specific sources of inequality. Instead, it is due to broadly rising pay in low-wage occupations, which has particularly benefited workers in tightening labor markets. Rebounding post-Great Recession labor demand at the bottom offset enduring drivers of inequality.

Preventing soft skill decay among early-career women in STEM during COVID-19: Evidence from a longitudinal intervention.

As the workforce shifts to being predominantly hybrid and remote, how can companies help employees-particularly early-career women in science, technology, engineering, and mathematics (STEM) fields-develop greater confidence in their soft skills, shown to improve organizational retention? We evaluate the effects of an online longitudinal intervention to develop soft skills among early-career women employees at a North American biotechnology company during the height of the COVID-19 pandemic. Controlling for baseline levels collected immediately prior to nationwide lockdowns, we find that a 6-month online intervention increased early-career women's assessments of their soft skills at work by an average of 9% (P < 0.001), compared with a decrease of about 3.5% for a matched control group (P < 0.05), resulting in an average treatment effect of nearly 13% on the treated group. Furthermore, we find evidence that the intervention led to an increase in manager-assessed performance for early-career women relative to employees not in the intervention, and that overall, increased self-assessments of soft skill competencies were associated with greater odds of retention. Results show how employee soft skill development was affected by the pandemic and provide insights for a feasible and cost-effective method to train and engage a hybrid or fully remote workforce.

Trends in social mobility in postrevolution China.

In this paper, we study long-term trends in social mobility in the People's Republic of China since its inception in 1949, with two operationalizations: 1) intergenerational occupational mobility and 2) intergenerational educational mobility. We draw on an accumulation of administrative and survey data and provide comparable estimates of these measures for birth cohorts born after 1945. To help interpret the results, we compare trends in China to those in the United States for the same birth cohorts. We find an increase in intergenerational occupational mobility in China due to its rapid industrialization in recent decades. Net of industrialization, however, intergenerational occupational mobility has been declining for recent cohorts. Intergenerational educational mobility in China shows a similar declining trend. In addition, mobility patterns have differed greatly by gender, with women in earlier cohorts and from a rural origin particularly disadvantaged. We attribute the general decline in social mobility to market forces that have taken hold since China's economic reform that began in 1978. In contrast, social mobility by both measures has been relatively stable in the United States. However, while social mobility in China has trended downward, it is still higher than that in the United States, except for women's educational mobility.

Health and inclusive labour force participation.

The future of work is rapidly changing, with higher flexibility of the labour market and increasing informal employment in many countries worldwide. There is also an increased pressure to extend working careers until older age. We introduce the concept of working life expectancy as a useful metric, capturing the expected numer of years in paid employment across the working age individuals, in particular among different groups. We describe factors that determine working life expectancy. Macro-level factors focus on the socioeconomic and political context that influences labour force participation, primarily policies and legislation in specific countries. At the meso level, employment contracts and working conditions are important. The micro level shows that individual characteristics, such as education, gender, and age, influence working careers. There are three important groups with a disadvantaged position in the labour market-workers with chronic diseases, workers with impairing disabilities, and workers aged 50 years or more. Within each of these disadvantaged groups, macro-level, meso-level, and micro-level factors that influence entering and exiting paid employment are discussed. To assure that paid employment is available for everyone of working age and that work contributes to better health, specific challenges need to be addressed at the macro, meso, and micro levels. To reach inclusive labour force participation, national policies, company practices, and workplace improvements need to be aligned to ensure safe and healthy workplaces that contribute to the health and wellbeing of workers and their communities.

COVID-19 Mortality by Usual Occupation and Industry:46 States and New York City, United States, 2020.

Objectives-This report describes COVID-19 mortality in 2020 among U.S. residents in 46 states and New York City by usual occupation and industry.

Trends in occupational and work-related contact dermatitis attributed to nickel, chromium and cobalt in the UK: findings from The Health and Occupation Research network 1996-2019.

BACKGROUND: Occupational exposure to metals such as nickel, chromium and cobalt can be associated with contact dermatitis, which can adversely affect an individual's health, finances and employment. Despite this, little is known about the incidence of metal-related occupational contact dermatitis over prolonged periods of time. OBJECTIVES: To investigate the medically reported trends in the incidence of work-related contact dermatitis attributed to nickel, chromium and cobalt in the UK. METHODS: Incidence and trends in cases of occupational contact dermatitis caused by nickel, chromium or cobalt between 1996 and 2019 (inclusive), reported to the EPIDERM surveillance scheme, were investigated and compared with trends in the incidence of occupational contact dermatitis attributed to agents other than the aforementioned metals. A sensitivity analysis restricting the study cohort to cases attributed to only one type of metal was also conducted. RESULTS: Of all cases reported to EPIDERM during the study period, 2374 (12%) were attributed to nickel, chromium or cobalt. Cases predominantly comprised females (59%), with a mean (SD) age (males and females) of 38 (13) years. Cases were most frequently reported in manufacturing, construction, and human health and social activity industries. The most frequently reported occupations were hairdressing, and sales and retail (assistants, cashiers and checkout operators). The highest annual incidence rate of contact dermatitis was observed in females (2.60 per 100 000 persons employed per year), with the first and second peak seen in those aged 16-24 and ≥ 65 years, respectively. A statistically significant decrease in the incidence of occupational contact dermatitis attributed to metals over the study period was observed for all occupations (annual average change -6.9%, 95% confidence interval -7.8 to -5.9), with much of the decrease occurring between 1996 and 2007. Similar findings were obtained in the sensitivity analyses. CONCLUSIONS: Over a period of 24 years, there has been a statistically significant decline in the incidence of metal-related occupational contact dermatitis in the UK. This could be attributed not only to improvements in working conditions, which have reduced metal exposure, but could also be due to the closure of industries in the UK that might have generated cases of contact dermatitis owing to metal exposure.

Characterizing the Burden of Occupational Chemical Exposures by Sociodemographic Groups in the United States, 2021.

Objectives. To estimate the number and prevalence of workers in the United States exposed to chemical hazards available in the Canadian job-exposure matrix (CANJEM) database and examine exposure disparities across sociodemographic groups. Methods. We merged US worker demographic data from the Current Population Survey with CANJEM to characterize the burden and sociodemographic distribution of 244 chemical exposures in the United States in 2021. An interactive version of the full data set is available online (https://deohs.washington.edu/us-exposure-burden). Results. Of the chemical exposures examined, the most prevalent were cleaning and antimicrobial agents (14.7% of workforce estimated exposed), engine emissions (12.8%), organic solvents (12.1%), polycyclic aromatic hydrocarbons (10.1%), and diesel engine emissions (8.3%). Racial and ethnic minoritized groups, persons with lower educational attainment, foreign-born noncitizens, and males were generally overrepresented in exposure to work-related chemical hazards. Conclusions. In the United States, marginalized sociodemographic groups are estimated to experience an inequitable burden to many chemical exposures because of occupational segregation. Data from this analysis can inform occupational and public health research, policy, and interventions aimed at reducing the burden of disease and health inequities in the United States. (Am J Public Health. 2024;114(1):57-67. https://doi.org/10.2105/AJPH.2023.307461).

Salary Disparities in Public Health Occupations: Analysis of Federal Data, 2021‒2022.

Objectives. To assess salary differences between workers within key public health occupations in local or state government and workers in the same occupations in the private sector. Methods. We used the US Department of Labor's Occupational Employment and Wage Survey (OEWS). Referencing previous studies matching Standard Occupational Classification (SOC) codes with health department occupations, we selected 44 SOC codes. We contrasted median salaries in OEWS for workers in each occupation within state or local government with workers in the same occupations outside government. Results. Thirty of 44 occupations paid at least 5% less in government than the private sector, with 10 occupations, primarily in management, computer, and scientific or research occupations paying between 20% and 46.9% less in government. Inspection and compliance roles, technicians, and certain clinicians had disparities of 10% to 19%. Six occupations, primarily in social work or counseling, paid 24% to 38.7% more in government. Conclusions. To develop a sustainable public health workforce, health departments must consider adjusting their salaries if possible, market their strong benefits or public service mission, or use creative recruitment incentives such as student loan repayment programs for hard-to-fill roles. (Am J Public Health. 2024;114(3):329-339. https://doi.org/10.2105/AJPH.2023.307512).

Standard Occupational Classification Codes: Gaps in Federal Data on the Public Health Workforce.

Objectives. To determine whether US Department of Labor standard occupational classification (SOC) codes can be used for public health workforce research. Methods. We reviewed past attempts at SOC matching for public health occupations and then used the National Institute for Occupational Safety and Health Industry and Occupation Computerized Coding System (NIOCCS) to match the actual job titles for 26 516 respondents to the 2021 Public Health Workforce Interests and Needs Survey (PH WINS) with SOC codes, grouped by respondents' choice of job category in PH WINS. We assessed the accuracy of the NIOCCS matches and excluded matches under a cutpoint using the Youden Index. We assessed the percentage of SOC matches with insufficient information and diversity of SOC matches per PH WINS category using the Herfindahl-Hirschman Index. Results. Several key public health occupations do not have a SOC code, including disease intervention specialist, public health nurse, policy analyst, program manager, grants or contracts specialist, and peer counselor. Conclusions. Without valid SOC matches and detailed data on local and state government health departments, the US Department of Labor's data cannot be used for public health workforce enumeration. (Am J Public Health. 2024;114(1):48-56. https://doi.org/10.2105/AJPH.2023.307463).

A high-frequency mobility big-data reveals how COVID-19 spread across professions, locations and age groups.

As infected and vaccinated population increases, some countries decided not to impose non-pharmaceutical intervention measures anymore and to coexist with COVID-19. However, we do not have a comprehensive understanding of its consequence, especially for China where most population has not been infected and most Omicron transmissions are silent. This paper aims to reveal the complete silent transmission dynamics of COVID-19 by agent-based simulations overlaying a big data of more than 0.7 million real individual mobility tracks without any intervention measures throughout a week in a Chinese city, with an extent of completeness and realism not attained in existing studies. Together with the empirically inferred transmission rate of COVID-19, we find surprisingly that with only 70 citizens to be infected initially, 0.33 million becomes infected silently at last. We also reveal a characteristic daily periodic pattern of the transmission dynamics, with peaks in mornings and afternoons. In addition, by inferring individual professions, visited locations and age group, we found that retailing, catering and hotel staff are more likely to get infected than other professions, and elderly and retirees are more likely to get infected at home than outside home.

Effect of long working hours on psychological distress among young workers in different types of occupation.

BACKGROUND: We investigated how the association between long working hours and psychological distress varies across different employment and occupation types in young workers. METHODS: Examining a nationally representative sample of 7246 Korean workers (3621 women) aged 15 to 40, we analyzed 23,492 observations spanning from 2016 to 2020. Psychological distress was measured using the Brief Encounter Psychosocial Instrument. We employed a generalized estimating equation to estimate odds ratios (ORs) and 95% confidence intervals (CIs). RESULTS: Of the total observations, 5.2% worked <35 h/week, 52.9% worked 35-40 h/week, 23.5% worked 41-48 h/week, 10.3% worked 49-54 h/week, and 8.2% worked ≥55 h/week. The OR (95% CI) of the association between long working hours and psychological distress was 1.38 (1.11-1.72) for <35 h/week, 1.47 (1.32-1.65) for 41-48 h/week, 1.74 (1.49-2.04) for 49-54 h/week, and 2.11 (1.75-2.55) for ≥55 h/week compared to 35-40 h/week. The OR (95% CI) of the association between working ≥55 h/week and psychological distress was significantly higher among wage workers (OR [95% CI]: 2.37 [1.94-2.89]) compared to self-employed workers (OR [95% CI]: 0.84 [0.52-1.36]). Additionally, the OR (95% CI) of the association between working ≥55 h/week and psychological distress was significantly higher among white-collar workers (OR [95% CI]: 3.24 [2.54-4.13]) compared to service/sales workers (OR [95% CI]: 1.22 [0.86-1.72]) or blue-collar workers (OR [95% CI]: 1.71 [1.10-2.67]). No clear gender differences were observed. CONCLUSION: Psychological distress caused by long working hours can be pronounced among white-collar and wage workers.

Job strain, occupation, and bladder health among women.

OBJECTIVES: Lower urinary tract symptoms (LUTS) are common among employed women. An underexplored topic is whether characteristics of women's occupations may influence LUTS. The present study examined whether job strain and its individual components (psychological demands, decision latitude) were associated with greater LUTS and their impact and whether, compared to managerial and professional occupations, occupations characterized by manual labor, sales, service, nursing, and teaching were associated with greater LUTS and their impact. METHODS: Coronary Artery Risk Development in Young Adults cohort study data were analyzed. Job strain and occupation were assessed in 1987-88 and 1995-96. In 2012-13, LUTS and their impact were assessed. LUTS/impact category (a composite variable ranging from bladder health to mild, moderate, and severe LUTS/impact) was regressed on job strain and occupation in separate analyses, adjusting for age, race, parity, education, and financial hardship (n = 1006). RESULTS: Job strain and its individual components were not associated with LUTS/impact. In comparison to managerial and professional occupations, service occupations in 1987-88 and 1995-96 were both associated with greater odds of LUTS/impact in proportional odds logistic regression analyses. Employment as a nurse, health assistant, or health aide in 1995-96 was associated with greater odds of any LUTS/impact versus bladder health. Support positions in 1987-88 and sales positions in 1995-96 were associated with greater odds of moderate or severe LUTS/impact versus bladder health or mild LUTS/impact. CONCLUSIONS: Future research should examine characteristics of workplaces that may promote or constrain bladder health (e.g., time and autonomy to void when desired, infrastructure to void).

The chronically ill in the labour market - are they hierarchically sorted by education?

BACKGROUND: The chronically ill as a group has on average lower probability of employment compared to the general population, a situation that has persisted over time in many countries. Previous studies have shown that the prevalence of chronic diseases is higher among those with lower levels of education. We aim to quantify the double burden of low education and chronic illness comparing the differential probabilities of employment between the chronically ill with lower, medium, and high levels of education and how their employment rates develop over time. METHODS: Using merged Norwegian administrative data over a 11-year period (2008-2018), our estimations are based on multivariable regression with labour market and time fixed effects. To reduce bias due to patients' heterogeneity, we included a series of covariates that may influence the association between labour market participation and level of education. To explicitly explore the 'shielding effect' of education over time, the models include the interaction effects between chronic illness and level of education and year. RESULTS: The employment probabilities are highest for the high educated and lowest for chronically ill individuals with lower education, as expected. The differences between educational groups are changing over time, though, driven by a revealing development among the lower-educated chronically ill. That group has a significant reduction in employment probabilities both in absolute terms and relative to the other groups. The mean predicted employment probabilities for the high educated chronic patient is not changing over time indicating that the high educated as a group is able to maintain labour market participation over time. Additionally, we find remarkable differences in employment probabilities depending on diagnoses. CONCLUSION: For the chronically ill as a group, a high level of education seems to "shield" against labour market consequences. The magnitude of the shielding effect is increasing over time leaving chronically ill individuals with lower education behind. However, the shielding effect varies in size between types of chronic diseases. While musculoskeletal, cardiovascular and partly cancer patients are "sorted" hierarchically according to level of education, diabetes, respiratory and mental patients are not.