Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, ECONOMICS AND FINANCE (oxfordre.com/economics). (c) Oxford University Press USA, 2019. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

date: 10 December 2019

Health Economics of the Workplace: Workplace Accidents and Effects of Job Loss and Retirement

Summary and Keywords

There are three main topics in research on the effects of work on health.

The first topic is workplace accidents where the main issues are reporting behavior and workplace safety policies. A worker seems to be less inclined to report a workplace accident for fear of job loss when unemployment is high or when the worker has a temporary contract that may not be renewed. Workplace safety legislation has intended to reduce the incidence and severity of workplace accidents but empirical evidence on this result is unclear.

The second topic is employment and health where the focus is on how job characteristics and job loss affect health, in particular mental health. Physically demanding jobs have negative health effects. The effects of working hours vary and the effects of job loss on physical and mental health are not uniform. Job loss seems to increase mortality.

The third topic concerns retirement and health. Retirement seems to have a negative effect on cognitive skills and short-term positive effects on overall health. Other than that, the effects are very inconsistent, that is, even with as clear a measure as mortality, it is not clear whether life expectancy goes up, goes down, or remains constant due to retirement.

Keywords: work, health, workplace accidents, job loss, retirement, mortality, health economics

Introduction

Both work and health are multidimensional. For an economist, at the minimum, work implies an exchange of leisure for income but workers may also derive utility from having a job beyond monetary benefits or experience disutility related to specific types of work. Health includes physical health and mental health, with happiness as one particular indicator. Various relationships exist between work and health. On the one hand, working involves running the risk of being confronted with a workplace accident. Work can cause physical problems or stress. Workers may become ill because of an unpleasant work environment and fear of job loss, or even disabled through a workplace accident caused by a dangerous work setting. On the other hand, work can also have positive effects on physical and mental health (e.g., by contributing to self-esteem, confidence, social interactions, and happiness). Health may also affect work, since unhealthy workers may earn less or have difficulties in finding a job when unemployed or in keeping a job when employed. This article considers only the possible effects in one direction, that is, from work to health and not the other way around, which implies no discussion of disability as a separate health state because almost all economic research on disability deals with the effects of being disabled or of receiving disability benefits on the difficulties of finding a job (Koning & Lindeboom, 2015).

Table 1. Indicators of Self-Assessed Health

Country

All ages

Age 25–54

Job

No Job

Australia

3.1

3.2

2.9

0.3

Canada

3.2

3.3

3.0

0.3

Chile

2.9

3.0

2.8

0.2

Estonia

2.5

2.7

2.6

0.1

Finland

2.8

3.1

2.8

0.3

France

3.0

3.2

2.8

0.4

Germany

2.9

3.2

2.7

0.5

Hungary

2.7

3.0

2.7

0.3

Italy

2.9

3.0

3.0

0.0

Japan

2.6

2.7

2.6

0.1

Korea

3.0

3.0

2.9

0.1

Mexico

3.0

3.1

2.9

0.2

Netherlands

2.9

3.1

2.6

0.5

New Zealand

3.1

3.2

3.0

0.2

Norway

3.2

3.4

2.4

1.0

Poland

2.7

3.1

2.7

0.4

Slovenia

2.8

3.1

2.7

0.4

Spain

2.9

3.1

2.9

0.2

Sweden

3.0

3.1

2.9

0.2

Switzerland

3.1

3.3

3.0

0.3

Turkey

2.8

3.0

2.7

0.3

United Kingdom

3.0

3.2

2.9

0.3

United States

3.1

3.2

2.9

0.3

Note. Health as a subjective measure, with a scale from 1 (poor) to 4 (very good). Adapted from World Value Survey Wave 5 (2005–2009) and Wave 6 (2010–2014).

Table 1 shows differences across Organisation for Economic Cooperation and Development (OECD) countries in self-assessed health where the measure is from 1 to 4 (poor—fair—good—very good). The average self-reported health over all ages in the first column is quite high in many countries. The highest average reported health over all ages is in Canada and Norway (3.2), and the lowest in Estonia (2.5). On a scale from 1 to 4, the cross-country difference of 0.7 is quite high. However, it is not clear to what extent these differences are related to actual cross-country differences in objective health or related to cultural differences in the perception of health. The remaining columns show the average reported health for prime age individuals, distinguishing between those who had a job at the time of the survey and those who did not. In every country except Italy, prime age individuals with a job report on average better health than those without a job. It is unclear to what extent this difference is an effect of individuals with worse health finding it more difficult to be employed or whether the difference is related to a justification bias caused by non-employed individuals.1 The highest self-reported health for prime age workers is in Norway, while for prime age nonworkers the highest average health is in Canada, Italy, New Zealand, and Switzerland. Workers have the lowest average health in Estonia and Japan, while nonworkers have the lowest average health in Norway.

Focusing on the effects of work on health, three main topics are distinguished. First, workplace accidents and workplace safety policies are discussed. For these, there are clear relationships between work and health. Workplace accidents are only possible when working; workplace safety policies aim to intervene regarding the effects of working on health.

Second, the relationships between job characteristics, job loss, and health are discussed. Third, the relationship between retirement and health are examined. Both job loss and retirement are transitions from work to nonwork, of which the health effects may depend on whether this transition is voluntary or involuntary (Bassanini & Caroli, 2015). By studying the health effects of such transitions, something can be learned about the effects of work on health. If health deteriorates because of job loss, this implies that working has beneficial effects on health. If health improves after retirement, this implies that early retirement is good for health.

The effects of work on health are not easily established because of measurement issues, both related to work and related to health, and because of difficulties in establishing causal effects rather than associations. Concerning workplace accidents, causality is less of an issue than misreporting. Workers who fear that they may lose their job because of reporting that they experienced a workplace accident may be reluctant to inform their employer or proper authorities. Although it may seem straightforward that an accident causally affects work, this is not always the case. For example, it may be that less healthy workers are exposed to more dangerous work than healthier workers. In the relationship between job loss and health as well as between job insecurity and health, causality is an important issue. If unhealthy workers are more likely to lose their job, then comparing workers who lost their job with those who did not is not informative regarding a causal relationship. Also, when investigating the effects of retirement on health, one needs to address causality issues. It could be that less healthy workers retire earlier; thus, comparing retired workers with employees of the same age does not reveal the causal effects of retirement on health.

Causality can be addressed in various ways, that is, to account for possible selectivity through correlated observed and unobserved characteristics affecting both work and health, and to account for reverse causality, that is, health having consequences on the labor market position of individuals. Some studies use instrumental variables. A firm closure is an example of a variable that affects health through job loss but not health directly. Early retirement eligibility is another example of a variable affecting health through a labor market transition (retirement) but not health directly. To account for correlations through time-invariant unobserved heterogeneity, some studies use panel data that allow for individual fixed effects. Some studies base identification of causal effects on propensity score matching based on the assumption that after including a large number of observables, unobservables are no longer relevant. Identification of the relevant effects may also come from policy changes that affect one group of workers but not another. And, identification may be based on discontinuities in, for example, the earliest retirement age.

Workplace Accidents

An accident at work is defined as a discrete occurrence in the course of work which leads to physical or mental harm. Fatal accidents at work are those that lead to the death of the victim within 1 year. According to Eurostat (2017), in 2012 there were 3.2 million non-fatal and 4,000 fatal workplace accidents in the European Union. Male workers experienced about 2.2 million non-fatal accidents, female workers about 1 million. Construction, manufacturing, transportation, agriculture, forestry, and fishing sectors together accounted for just over two thirds of all fatal accidents at work and over half of all non-fatal workplace accidents. There are two types of common injuries, namely wounds and superficial injuries (about 30% of the total) and dislocations, sprains, and strains (about 25%). Around one in 10 accidents resulted in concussion and internal injuries, while a similar proportion of accidents concerned bone fractures (Eurostat, 2017).

Table 2. Workplace Safety Indicators, 2015

Workplace accidents

Includes commuting

Inspectors

Country

Fatal

Non-fatal

Source

Number

Source

Australia

1.7

10.5

I

Austria

2.3

14.8

I

Yes

0.8

L

Belgium

1.6

14.0

I

Yes

0.7

O

Canada

2.0

11.9

R

0.1

O

Czech Republic

2.6

8.9

O

1.0

L

Denmark

1.0

17.9

O

Estonia

2.5

10.8

L

0.6

L

Finland

1.4

17.3

O

Yes

1.5

O

France

2.6

31.6

I

0.8

O

Germany

1.6

18.1

I

1.5

I

Greece

1.2

1.6

I

Yes

Hungary

2.3

5.5

L

0.9

L

Ireland

2.5

8.5

L

0.3

O

Israel

1.7

22.6

I

Yes

0.8

L

Italy

2.4

14.1

I

Yes

Latvia

3.7

2.2

L

1.2

L

Lithuania

4.0

2.8

L

1.5

L

Luxembourg

3.3

18.7

I

Mexico

8.2

31.3

I

0.2

O

Netherlands

0.5

10.3

I

Yes

Norway

1.5

4.0

L

Yes

1.3

L

Poland

2.5

5.1

E

1.0

L

Portugal

3.5

29.5

I

Yes

0.8

L

Slovak Republic

2.8

4.5

L

1.3

L

Slovenia

2.8

15.1

O

Yes

0.8

L

Spain

2.1

27.7

I

1.0

L

Sweden

1.0

7.8

O

Yes

0.5

O

Switzerland

1.3

19.0

O

1.1

L

Turkey

6.9

13.2

I

0.3

L

United Kingdom

0.4

7.6

I

0.3

I

United States

3.4

10.7

E

0.1

O

Note. Fatal accident rate is the average number of new cases of fatal occupational injury during the calendar year per 100,000 workers. Non-fatal accident rate is the average number of new cases of non-fatal occupational injury during the calendar year per 1,000 workers. Inspectors number = per 10,000 employed; for some countries, the reference group is employees; for other countries, it is insured persons; for some countries, the information is from the calendar year 2013 or 2014.

Source information: E: Establishment survey; I: Insurance records; L: Labor inspectorate records; O: Other source; R: Records of employers’ organizations.

General Source: ILOSTAT.

Table 2 shows differences across OECD countries in workplace accident rates where a distinction is made between fatal and non-fatal accidents. The numbers shown are indicative of the seriousness of workplace accidents but cross-country comparisons are not easy to make because of differences in definitions and measurement issues. Some countries include commuting accidents in the statistics while others do not; some countries include occupational diseases while others do not; some countries count workplace accidents as a share of all workers, while other countries have insured persons in the denominator. Furthermore, cross-country differences exist in the industrial structure. Countries with a high employment share in manufacturing are more likely to have a high workplace accident rate than countries with a high employment share in services. The highest rate of fatal accidents is 6.9 per 100,000 workers in Turkey, while the lowest is 0.4 per 100,000 workers in the United Kingdom. For non-fatal accident rates, the highest number is in France, with 31.6 per 1,000 workers, and the lowest rate is in Greece with 1.6 per 1,000 workers.

An important topic in economic research on workplace accidents is reporting behavior and the way this is related to the state of the labor market or the types of labor contracts. Another important topic is the relationship between workplace safety and workplace accidents.

Reporting Behavior

Workplace accidents are often found to be cyclical. An important issue is whether this is a real phenomenon due to cyclical variations in workplace safety or whether it is a reporting phenomenon. Reporting of an accident may be influenced by the state of the labor market. When unemployment is high, in fear of job loss workers may be reluctant to report a minor workplace accident. Alternatively, when unemployment is high, economic activity within firms is low and workers are more relaxed and less likely to be involved in hazardous activities, thus increasing workplace safety. Identification of the reporting effect comes from a comparison of fatal and non-fatal accident rates. If workplace safety would show cyclical fluctuations, this would materialize in both types of accidents. If only non-fatal workplace accidents show cyclical fluctuations, reporting is the more likely to blame. Underreporting occurs if workers are afraid that reporting an accident may lead to job loss, and this is more likely to occur in times of high unemployment when the consequences of job loss are more severe.

Some studies use time-series cross-country data. Based on data from 16 OECD countries, Boone and van Ours (2006) concluded that cyclical fluctuations were present in non-fatal but not in fatal workplace accident rates. Thus, cyclical fluctuations were not caused by changes in workplace safety but instead were related to reporting behavior. Palali and van Ours (2017) analyzed cross-country time-series data on fatal and non-fatal workplace accidents and road accidents. Since neither fatal workplace accidents nor road accidents have cyclical fluctuations, the variation in non-fatal workplace accidents must have been related to reporting behavior. Cyclical fluctuations in workplace accidents have not only been investigated using aggregate data. Boone et al. (2011) used Austrian matched worker-firm data, finding that workers who reported an accident were more likely to be fired later on. Furthermore, reporting behavior was found to be an important source of cyclical fluctuations in workplace accidents.

Underreporting may also be more likely by workers who have temporary jobs. The evidence for this is not straightforward. Amuedo-Dorantes (2002), using Spanish data, found that temporary workers were more likely to be confronted with a workplace accident than workers with a permanent contract. However, once differences in working conditions were accounted for, temporary workers were not more likely to have a workplace accident. Guadalupe (2003), using Spanish data, found that workplace accident rates were higher among temporary workers than among permanent contract workers, which could be related to their lack of work experience or because temporary workers put more effort into the work to increase the probability that their temporary contract would be extended. García-Serrano et al. (2010), analyzing Spanish data, found that temporary help agency workers had a lower workplace accident rate than regular temporary or permanent contract workers. Bena et al. (2013) analyzed Italian data and found that workplace accident rates fell with job tenure. Probst et al. (2013) found that accident underreporting was more relevant when workers’ perception of job insecurity was greater. Bender, Green, and Heywood (2012) analyzed data from 30 European countries and showed that piece-rate workers had a higher workplace accident rate.

From the point of view of workplace safety, it is important to know whether temporary workers are more likely to suffer from severe workplace accidents. Previous studies on contract type and workplace accidents were focused on the incidence of workplace accidents. Using administrative Italian data on workplace accidents, Picchio and van Ours (2017) found that if confronted with a workplace accident, compared to workers with a permanent contract, workers with a temporary contract were more likely to have a severe injury. Although part of the effect may have been due to temporary workers being more likely to be assigned dangerous tasks because they had less bargaining power, most of the difference as related to the nature of the employment contract was due to underreporting by temporary workers.

Pouliakas and Theodossiou (2013), in their overview of the literature on occupational safety and health, mentioned that the nature of the job contract may be important not only because of reporting behavior. In times of high aggregate demand, firms have no incentive to provide workplace safety training to workers on temporary and casual contracts. Temporary workers may be unfamiliar with characteristics of the workplace and their job, and they were therefore more susceptible to workplace accidents. Underreporting may have arisen from pressure by the employer through direct intimidation, through the threat of job loss, reassignment to a lower-paying job, denial of overtime, denial of promotions, or harassment. Firms may want to underreport workplace accidents because inspections by safety authorities are often based on reported accidents (Workplace Safety and Insurance Board, 2013).

All in all, statistics on workplace accidents suffer from cyclical underreporting. When unemployment is high, workers are reluctant to report small workplace accidents. Whether the nature of the contract matters is not clear. Some studies find that temporary jobs or casual work are more accident related. Also, underreporting may be an issue but in addition, it may be that temporary or casual workers are assigned to different types of work with a higher risk profile.

Workplace Safety Policies

Cross-country differences exist in monitoring workplace safety policies (e.g., in terms of numbers of labor inspectors).2 The last two columns of Table 2 show that the number of workplace labor inspectors per 10,000 employed range from a low of 0.1 in Canada and the United States to a high of 1.5 in Finland, Germany, and Lithuania. These differences are not necessarily indicative of differences in the intensity of monitoring of workplace safety policies but do indicate that resources regarding monitoring vary.

Many studies on the relationship between workplace safety policies and workplace accidents are based on U.S. data but European studies also exist. In the United States, the Occupational Safety and Health Act (OSHA) was signed in 1970 “to save lives, prevent injuries, and protect the health of American workers.” In 1989, the EU adopted Directive 89/391/EEC “to encourage and enhance the protection of workers through measures concerning the prevention of work-related risks, the protection of safety and health, the elimination of risk and accident factors and also the informing, consultation, balanced participation and training of workers.”

Several studies have been performed on the effectiveness of OSHA. Ruser and Butler (2009) found weak linkages between noncompliance and workplace accidents and significant effects of OSHA enforcement on industry violation rates. Gray and Scholz (1993) using U.S. firm panel data on injuries and OSHA inspections, pointed to the “regression to the mean” phenomenon. Firms with an abnormally high injury rate were selected for inspection. Because of this, accident rates tended to be lower in the year after inspection, which was then erroneously attributed to a causal effect of the inspection.

Lanoie (1992) studied the effect of the occupational safety and health regulation in Canada (Quebec), finding no significant effects on the risk of workplace accidents. Arocena and Nunez (2009) analyzed the effects of the occupational safety and health legislation in Spanish manufacturing, finding that the number of serious injuries was reduced in advanced manufacturing sectors, but such an effect did not occur in traditional industries. Bande and López-Mourelo (2014) studied Spanish data, finding that regional differences in severe and fatal accidents were explained by differences in workplace inspection regimes. Aires et al. (2010) analyzed the effects of workplace safety regulations in the construction sector using data from 15 EU countries and found no conclusive evidence of such effects.

All in all, it is not clear whether and, if so, how workplace safety regulations affect workplace accidents. Institutions do not change very often, so it is hard to identify causal effects. The analysis of the effects of workplace inspections may be hampered by a regression to the mean, and firm management may focus more on compliance with the law than on an increase in overall workplace safety.

Employment and Health

How employment affects health may be related to the nature of the job, job insecurity, and job loss. Table 3 shows differences across OECD countries in terms of various indicators of job stability (i.e., job separation rates, the share of temporary jobs in employment, and perception of job insecurity). The first column shows the annual separation rates in 2012, which range from a low of 6.2% in the Slovak Republic to a high of 25.6% in Turkey. The second column shows the share of workers with a temporary contract in 2016. Also, here the cross-country differences are immense, from a low of 2.5% in Estonia to a high of 26.3% in Spain. The third column presents an indication of perceived job insecurity as measured in the World Value Survey, ranging from 1 to 4 (not at all to very much). Workers in the Netherlands on average have the lowest job insecurity (2.0), while workers in Japan and Korea have the highest job insecurity (3.2). Clearly, the various indicators are not in line with each other. While the feeling of insecurity is the same in Japan and Korea, the share of temporary jobs is low in Japan (4.7%) and rather high in Korea (15.1%). In the Netherlands, with one of the highest shares of temporary workers, workers report the lowest job insecurity.

Table 3. Job Separation, Temporary Work, and Job Insecurity (Workers’ Ages 25–54)

Country

Job separation rate (%)

Temporary job (%)

Job insecurity (%)

Australia

5.5

2.3

Austria

12.5

5.9

Belgium

10.4

8.4

Chile

17.0

2.7

Czech Republic

9.3

8.4

Denmark

17.7

9.0

Estonia

15.4

2.5

3.1

Finland

16.8

13.3

France

12.0

13.5

Germany

12.6

9.6

2.2

Greece

14.8

10.8

Hungary

10.7

8.0

Ireland

12.1

5.9

Italy

9.2

14.5

Japan

4.7

3.2

Korea

15.1

3.2

Luxembourg

6.3

7.0

Netherlands

11.9

16.3

2.0

Norway

12.3

6.8

Poland

15.3

23.6

2.4

Portugal

15.2

20.3

Slovak Republic

6.2

8.2

Slovenia

9.9

13.9

2.1

Spain

18.8

26.3

2.8

Sweden

15.9

12.1

2.0

Switzerland

13.3

7.9

Turkey

25.6

10.4

2.9

United Kingdom

12.4

4.2

Note. Separation rate is the difference between the net employment change rate and the hiring rate. Adapted from OECD estimates, based on the European Union Labor Force Survey (2012). Temporary work relates to jobs that have a predetermined termination date as opposed to permanent jobs with unlimited duration. Source: OECD Employment Database (2018). Job insecurity is the degree to which people are worried about losing their job or not finding a job, on a scale from 1 (not at all) to 4 (very much). Source: World Value Survey, Wave 6 (2010–2014).

Nature of the Job

A strong association exists between jobs and health. Case and Deaton (2005), for example, showed that U.S. manual workers were less healthy than nonmanual workers. Similar to other relationships between work and health, a causal effect of type of job to health was only one possible explanation for such an association. Alternatively, reverse causality existed such that workers with poor health had access to only certain types of jobs, or the association was caused by joint observed or unobserved characteristics, for example, by educational attainment, with low education level workers being more likely to both have a manual job and poor health. Not many economic studies established a causal relationship between the nature of a job and health. An exception was Ravesteijn, van Kippersluis, and van Doorslaer (2018), who, using German data on occupation and self-reported health, found that selection effects explained a large part of the association between the two. Furthermore, they found that high physical demands and low job control had negative health effects, which increased with age.

Health may also be affected by characteristics such as working hours or job insecurity. Some studies exist regarding the relationship between working hours and health. Establishing a causal effect from working hours to health is challenging as selectivity through unobserved correlated variables need to be taken into account, while reverse causality, that is, the effect of health on working hours, may also be an issue. Bassanini and Caroli (2015) concluded, on the basis of an overview of previous studies, that long working hours had negative physical and mental health effects whereby many studies used an individual fixed effects approach. Berniell and Bietenbeck (2017) exploited the 1998 reduction of the workweek in France from 39 to 35 hours at constant earnings. They found that the working hours reduction increased self-reported health and reduced smoking. Booth and van Ours (2008) used British data on life satisfaction of partnered individuals, finding that neither for men nor for women was life satisfaction affected by their hours of work. Booth and van Ours (2009) used Australian data, finding that part-time working women were more satisfied with their hours of work than full-time working women. By contrast, male partners’ life satisfaction was unaffected by their partners’ market hours but was significantly increased if they themselves were working full-time. Analyzing Dutch data, Booth and van Ours (2013) concluded that men were happiest if they worked in an extensive part-time or full-time job. They were also happier if their partner worked in a part-time job, although once household income was accounted for, their life satisfaction was unaffected by their partners’ hours.

The health of workers may also be affected by job insecurity (i.e., in anticipation or fear of job loss). Job insecurity may occur if workers realize or fear that their firm is losing money and will have to cut costs by firing workers. It may also be that workers have a temporary contract which may or may not be renewed at the end of a predefined employment period. Caroli and Godard (2014) investigated the relationship between perceived job insecurity and a range of health outcomes, using information about male workers in 22 European countries. Their main finding was that perceived job insecurity had limited effects on health outcomes (e.g., headaches, eyestrain, and skin problems). Pirani and Salvini (2015) found that in Italy, workers with a temporary contract were more likely to suffer from poor self-reported health than workers with a permanent contract. Cottini and Ghinetti (2018) used Danish matched worker-firm data, finding that employment insecurity had negative health effects as measured through self-reported feelings of mental health and general well-being. In their analysis, they use changes in the workforce of the firm as an instrumental variable for employment insecurity.

All in all, some job characteristics such as physically demanding jobs, jobs with low worker control, as well as jobs that are insecure have negative health effects. The effects of working hours are less clear, although long working hours seem to have negative health effects. The difference may be due to the time needed for adjustments. The nature of the job in terms of occupation and type is the result of a long-term process, while adjusting hours of work can be done rather quickly within the same job or by changing jobs.

Job Loss

Job loss is a recurrent phenomenon. It may occur if an individual worker does not perform sufficiently to cover his or her wage costs and the employer decides to fire the worker. Job loss may also occur if, for whatever reason, a firm realizes losses and the employer cuts wage costs by firing one or more workers. In the extreme, a firm goes bankrupt and all workers lose their jobs. There are several reasons why job loss may have a negative effect on health. The drop in earnings may cause stress or a drop in self-esteem. A negative income shock induced by job loss may affect health through a change in consumption behavior, although it is not clear whether this behavior change has a positive or negative effect on health. A negative health change may occur if workers who lost their job choose less healthy, cheaper foods, but, as a positive effect, they may decide to cut expenditures on expensive luxury goods such as cigarettes and increase physical exercise.

Many studies focus on short-term effects of job loss but there are also studies that investigate the effect of job loss on mortality, which may partly be driven by experiencing long-term unemployment. The negative association between job loss and health has been known for quite some time. However, at the level of individual workers there are three alternative explanations for this association. First, it could be that the negative association is driven by other personal characteristics that affect both job loss and health. These other characteristics may be known or unknown to the researcher. It is possible that older workers are more likely to lose their jobs and, since older workers often have worse health, those who lose their job are on average less healthy. The second reason for the negative association may be a causal effect from health on job loss, that is, less healthy workers are more likely to lose their jobs. From the perspective of the researcher interested in the effects of job loss, this would be reverse causality. The third reason may be that job loss has a causal negative effect on health whereby these effects are at least partly related to the loss of income. Information on firm closures is helpful to establish a causal effect. The likelihood that a firm closes because of the bad health of its workers is not very high, thus ruling out the possibility of selective dismissals or reverse causality. Within the studies based on firm closures and mass layoffs, two types can be distinguished: studies focusing on physical and mental health indicators and studies focusing on mortality.

Browning, Moller Dano, and Heinesen (2006) analyzed Danish data, finding no effect on hospitalization for stress-related diseases of the circulatory system and diseases of the digestive system. Salm (2009) used U.S. data, finding no causal effect of job loss on various measures of physical and mental health. Schaller and Stevens (2015) analyzed U.S. data, finding that job loss was associated with a decline in self-reported mental and physical health and increases in depression and anxiety, but no relationship was found between job loss and a variety of diseases such as diabetes, arthritis, hypertension, heart disease, or high cholesterol. Black, Devereux, and Salvanes (2015) analyzed Norwegian data on workers in their early 40s, finding few short-term health effects. Michaud, Crimmins, and Hurd (2016) used U.S. biomarker data, finding no negative health effects for workers who lost their job due to a firm closure.

Eliason and Storrie (2009) used Swedish data, finding increased mortality for men but not for women in the first 4 years following job loss. The increase in mortality was mainly due to alcohol-related conditions and suicide. Browning and Heinesen (2012) studied the health effects for Danish full-time working males with a strong labor market attachment, finding that the effect on overall mortality was highest in the first year after job loss, in particular due to circulatory diseases. The authors also found that the health consequences of job loss were more serious if it occurred in a labor market with high unemployment. Bloemen, Hochguertel, and Zweerink (2018) studied the effect of job loss on mortality for older Dutch male workers with strong labor force attachment, finding an increase in mortality after job loss, suggesting that this was related to stress and changes in lifestyle.

Related to studies on job loss are investigations of the effects of transitions to unemployment. When workers became unemployed, on average their happiness dropped substantially. This drop in happiness went beyond the loss of income that most individuals experienced when they became unemployed. This finding was common in many studies (e.g., Clark & Oswald, 1994; Winkelmann & Winkelmann, 1998; Kassenboehmer & Haisken-Denew, 2009; Clark, 2003). However, analyzing German data, Gielen and van Ours (2014) found no drop in happiness across the board, but there was substantial variation. Workers who were unhappy with their job, who had sufficient alternative household income sources, or who had previously been unemployed did not become unhappy after a job loss.

All in all, it is not clear whether job loss has short-run effects in terms of mental or physical health, while entering unemployment does seem to have a negative effect on happiness. There seem to be long-term effects related to job loss in terms of increased mortality. It could be that health effects need some time to materialize, in which case short-term effects may be absent while long-term effects are easier to establish.

Retirement and Health

The health effects of retirement are interesting because average retirement ages are going up due to increasing standard retirement ages and because of disappearing early retirement programs (Boeri & van Ours, 2013).3

Table 4. Life Expectancy at Age 65 and Effective Retirement Ages

Life expectancy at age 65

Effective retirement age

Life expectancy after effective retirement age

Country

Males

Females

Males

Females

Males

Females

Australia

19.6

22.3

65.2

63.6

19.4

23.7

Austria

18.5

21.7

62.0

60.6

21.5

26.1

Belgium

18.4

21.9

61.3

59.7

22.1

27.2

Canada

19.2

22.0

65.9

63.1

18.3

23.9

Chile

17.6

21.3

71.0

67.2

11.6

19.1

Czech Republic

16.2

20.0

62.5

60.8

18.7

24.2

Denmark

18.2

20.8

63.7

63.1

19.5

22.7

Estonia

15.6

20.9

64.8

65.3

15.8

20.6

Finland

18.2

21.9

63.2

62.5

20.0

24.4

France

19.4

23.5

60.0

60.3

24.4

28.2

Germany

18.1

21.3

63.3

63.2

19.8

23.1

Greece

18.9

21.7

62.0

60.2

21.9

26.5

Hungary

14.6

18.7

63.6

60.7

16.0

23.0

Iceland

18.7

21.3

69.7

67.2

14.0

19.1

Ireland

18.6

21.1

66.9

63.5

16.7

22.6

Israel

19.5

21.6

69.3

66.5

15.2

20.1

Italy

19.4

22.9

62.1

61.3

22.3

26.6

Japan

19.6

24.4

70.2

68.8

14.4

20.6

Korea

18.4

22.6

72.0

72.2

11.4

15.4

Latvia

14.0

19.0

62.0

61.2

17.0

22.8

Luxembourg

18.9

22.7

61.2

61.0

22.7

26.7

Mexico

16.8

18.7

71.6

67.5

10.2

16.2

Netherlands

18.5

21.1

63.5

62.3

20.0

23.8

New Zealand

19.4

21.5

68.4

66.4

16.0

20.1

Norway

19.1

21.6

66.2

64.4

17.9

22.2

Poland

16.0

20.5

62.6

59.8

18.4

25.7

Portugal

18.0

21.8

69.0

64.9

14.0

21.9

Slovak Republic

15.3

19.2

60.8

59.5

19.5

24.7

Slovenia

17.9

21.8

62.3

60.9

20.6

25.9

Spain

19.4

23.6

62.2

62.6

22.2

26.0

Sweden

19.1

21.5

65.8

64.6

18.3

21.9

Switzerland

20.0

22.9

66.0

64.3

19.0

23.6

Turkey

16.1

19.3

66.1

66.5

15.0

17.8

United Kingdom

18.8

21.1

64.6

63.2

19.2

22.9

United States

18.0

20.6

66.8

65.4

16.2

20.2

Note. Life expectancy at age 65 is the average number of years that a person of that age can be expected to live, assuming that age-specific mortality levels remain constant. Effective retirement age is the average age at which older workers withdraw from the labor force. Adapted from OECD estimates.

Table 4 shows differences across OECD countries in life expectancy at age 65. In every country, life expectancy at age 65 is higher for women than for men. The cross-country differences in life expectancy at age 65 are substantial. Males at age 65 have the lowest life expectancy in Latvia (14 years) and the highest in Switzerland (20 years), a difference of 6 years. For women, an age 65 life expectancy is highest in Japan (24.4 years) and lowest in Hungary and Mexico (18.7 years), a difference of 5.7 years. The effective retirement age, that is, the average age at which workers retire, is usually not very far from age 65 but also here there are multiple cross-country differences. For males, the lowest effective retirement age is in France (60), the highest in Korea (72), a difference of 12 years. For females, the highest effective retirement age is also in Korea (72.2) while the lowest is in the Slovak Republic (59.5), a difference of 12.7 years. These differences in life expectancy at age 65 and effective retirement ages imply that the remaining duration of life after retirement is very different across countries. In France, male workers on average live 24.4 years after retirement and female workers live 28.2 years. Male workers in Mexico have a remaining life duration of only 10.2 years, while female workers in Korea live 15.4 years after retirement.

As with other health-related topics, to establish the causal effect of retirement on health one needs to take into account that the association between these variables may also be caused by joint observed and unobserved determinants and by reverse causality. Association through joint time-invariant unobserved determinants can be removed by introducing individual fixed effects. To deal with potential reverse causality, two methods are generally used: instrumental variables and a regression discontinuity design (RDD). Eligibility ages for early retirement or retirement-related social security benefits are popular instrumental variables. RDD analysis typically exploits the sudden increase in retirement probability as soon as an individual attains the age for pension eligibility. Sometimes an increase in standard retiring age is exploited to establish causality.

As with job loss, for the health effects of retirement a distinction can be made between studies focusing on mental and physical health and those focusing on mortality. In addition to this distinction, there is also a difference between single-country studies and multi-country studies. In single-country studies, identification is usually based on discontinuities in the eligibility to retire or on changes in the earliest retirement age. In multi-country studies, identification is usually based on cross-country differences in early retirement ages.

In single-country studies, one of the first is by Kerkhofs and Lindeboom (1997) who used Dutch panel data, finding that work had negative health effects such that early retirees and the unemployed were healthier than people at work. Bonsang, Adam, and Perelman (2012) investigated the effects of retirement on cognitive functioning of older Americans, finding significant negative, though not instantaneous, effects on cognitive functioning as measured by word learning and recall. Bonsang and Klein (2012) studied well-being effects of retirement in Germany, distinguishing between voluntary and involuntary retirement. They found that voluntary retirement had no effect on life satisfaction, while involuntary retirement had a negative effect. Insler (2014) found a positive health effect of retirement in the United States and attributed this to a behavioral change of retirees, who, for example, were more likely to stop smoking when they retired. Eibich (2015) found positive effects of retirement on the health of German workers, which he attributed to relief from work-related stress and strain, to an increase in sleep duration, and to a more active lifestyle. Gorry, Gorry, and Slavov (2015), studying the retirement of U.S. workers, found effects on happiness that were immediate, while health effects showed up later on. Kesavayuth, Rosenman, and Zikos (2016), studying retiring workers in the United Kingdom, found on average no effects on well-being. However, the effect was heterogeneous and related to personality traits. For females, the well-being effect of retiring was high if they scored high in openness or low in conscientiousness. Fé and Hollingsworth (2016) investigated the retirement effects on health and health-care utilization for U.K. males, finding neither short-term nor long-term effects. Shai (2018) found for Israeli workers that employment at older ages had a negative health effect, in particular for low education level workers. Picchio and van Ours (2018) studied the effects of retirement on health and happiness in the Netherlands, finding that these were heterogeneous by gender and marital status. Retirement of partnered men had positive effects on self-perceived health and happiness of both themselves and their partner. Single men retiring experienced a drop in self-assessed health. Irrespective of whether they were partnered or single, retirement of women had hardly any effect on their health. Partnered women experienced an increase in happiness if their partner retired, but if they themselves retired their happiness dropped. For single women, the drop in happiness at retirement was significantly negative. The health and happiness effects of retirement were also heterogeneous by educational attainment.

Some single-country studies on the health effects of retirement focused on mortality rates. Hernaes, Markussen, Piggott, and Vestad (2013), using a retirement reform in Norway, found that the mortality of workers who retired early was not different from those who did not retire early. Hallberg, Johansson, and Josephson (2015) analyzed the effects of an early retirement offer to Swedish army personnel, finding that the mortality of early retirees was lower. Fitzpatrick and Moore (2018) used a change in eligibility for social security retirement insurance in the United States to establish the effects of retirement, finding an increase in male mortality, which the authors attributed to retirement-associated changes in unhealthy behaviors. The increase was largest for unmarried males and males with low education levels. For retiring females, there was no significant increase in mortality. Kuhn et al. (2018) used Austrian data for blue collar workers, showing that early retirement increased mortality for men but not for women. Bloemen, Hochguertel, and Zweerink (2017), analyzing Dutch data, found that early retired workers had lower mortality rates than those who did not retire early.

Multi-country studies have been made on the effects of retirement on health. Using data from the United States, England, and 11 European countries, Rohwedder and Willis (2010) found that early retirement had negative effects on cognitive skills of people in their early 60s. Using similar cross-country data, Horner (2014) concluded that well-being improved through retirement but this was a temporary effect. Fonseca et al. (2014) analyzed data from various European countries, concluding that there was weak evidence of retirement reducing depression. Belloni, Meschi, and Pasini (2016), using data from 10 European countries, concluded that retirement had a positive effect on mental health of men while women were unaffected. The positive mental health effect was stronger for blue-collar men in areas that were strongly hit by the Great Recession. From a study based on 10 European countries, Mazzonna and Peracchi (2017) concluded that the retirement effects on health and cognitive skills were negative and increased with years after retiring. The effects were also heterogeneous in the sense that for physically demanding occupations, retiring had a positive and immediate effect on both health and cognitive skills. Kolodziej and García-Gómez (2017) used cross-country data, finding that the mental health effects of retirement were larger for those in poor mental health. Mu¨ller and Shaikh (2018) used data from various European countries to investigate the causal health effects of the retirement of a partner. Health was negatively affected by the retirement of the partner and positively affected by the person’s own retirement. These effects were heterogeneous: Male health was not affected by the retirement of his spouse, while female health was negatively affected by the retirement of her partner.

Finally, some overview studies have been made. Focusing on longitudinal studies, Van der Heide, van Rijn, Robroek, Burdorf, and Propper (2013) concluded that the effects on general health and physical health were unclear, while there seemed to be beneficial effects on mental health. Nishimura, Oikawa, and Motegie (2018) investigated the differences in the retirement effects across various studies, concluding that the choice of the estimation method was the key factor in explaining these differences. Banks, Chandola, and Matthews (2015) provided an overview of previous studies, concluding that there was no consensus about how retirement affected health, in particular physical health and mortality.

All in all, often, but not always, retirement was found to improve mental health; some studies found an improvement in physical health and a negative effect on cognitive skills. The effect of retirement on mortality varied from negative to positive, with some studies finding no effects.

Conclusions

Several relationships can be found between work and health. Some workers are confronted with a workplace accident such that there is a direct effect of working on being exposed to a negative health shock. As with many other labor market institutions, it is difficult to establish how workplace safety affects the occurrence of workplace accidents. An important issue in the study of the economics of workplace accidents is how important misreporting behavior is, (i.e., whether workplace accidents reported are a reflection of workplace safety). Studies that investigate workplace accidents usually conclude that reporting behavior is an important issue. To avoid being held accountable, workers are less inclined to report a workplace accident in situations where job loss would have serious effects, such as in a labor market with high unemployment or if the worker has a temporary contract that may not be renewed.

There are other relations between work and health that are less straightforward. Job characteristics in terms of occupation or working hours may also affect health. Job insecurity (i.e., fear of job loss) may affect health. To establish the effects of work on health, one can also investigate what happens if a worker loses her job. This can be because of involuntary job loss or because of voluntary or involuntary retirement. Although the transition in labor market status is the same (i.e., from work to nonwork), there are clear differences between job loss and retirement. Job loss usually occurs at a younger age than retirement. Job loss is usually unexpected, with consequences in terms of earnings, prestige, and psychological harm. Retirement, however, is often expected, with well-anticipated effects in terms of income. Retirement is a socially acceptable reason to stop working whereas job loss may be socially not appreciated.

As for job loss, the effects on health are not clear in the sense that different studies come to different conclusions. In terms of mental health and physical health, some studies find no effects whereas other studies find a deterioration of mental health, in particular in terms of life satisfaction. Also, mortality seems to increase after job loss.

The empirical evidence on the effects of retirement on health is mixed. Some studies find a positive effect, while other studies conclude there is no effect or a negative effect. The health effects of retirement may be beneficial in cases of voluntary retirement, while it may be detrimental if the retiree is cut off from social contacts. The health effects of retirement may be related to changes in health-related behavior such as smoking, drinking, and physical exercise. Whether retirement has positive or negative effects may depend on the type of work performed up to retirement. If the work was stressful or placed the worker in unhealthy situations, retirement may benefit health. However, it is possible that the work was satisfying because it was interesting and meaningful. Then retirement may be harmful for one’s health. Retirement always comes with an increase in leisure time and often with a substantial drop in income. Thus, the health effects of retirement may also depend on labor market institutions, in particular retirement benefit replacement rates. Having more leisure time could have a positive effect on health, but not necessarily. A drop in income in itself may be harmful to one’s health as higher income is usually associated with a healthy lifestyle and low income with an unhealthy lifestyle. Whether this also holds for negative income shock at retirement is open for discussion.

In terms of the effects of work on health, there is heterogeneity all over the (work)place. Part of the heterogeneity in the effects may be related to cross-country differences in labor market institutions. These vary a great deal with differences in employment protection legislation, unemployment benefits, retirement ages, and old age benefits and pensions. Workers may respond differently in terms of health, depending on particular labor market institutions or the situation in the labor market. The reporting of a workplace accident, the health consequences of job loss, or perceived job insecurity may depend on the level of unemployment, the level and duration of unemployment, or sickness benefits. The health consequences of job loss not only depend on the benefit levels but also on the opportunities to find a new job quickly. Perception about job insecurity is not only driven by the stability of the job in regard to specific employment protection rules but also by the subjective probability to find another job quickly.

Nevertheless, heterogeneity in the empirical findings relating work to health may also relate to differences in terms of identification methods, the nature of the data, or the dependent variable of interest. In particular, establishing causal effects is not always straightforward as often this causality relies on assumptions which are not always easy to test. In addition, health effects may be long in materializing, which make them even harder to correlate to work or a transition in labor market status.

References

Aires, M. D. M., Gamez, M. C. R., & Gibb, A. (2010). Prevention through design: The effect of European directives on construction workplace accidents. Safety Science, 48, 248–258.Find this resource:

Amuedo-Dorantes, C. (2002). Work safety in the context of temporary employment: The Spanish experience. Industrial and Labor Relations Review, 55(2), 262–272.Find this resource:

Arocena, P., & Nunez, I. (2009). The effect of occupational safety legislation in preventing accidents at work: Traditional versus advanced manufacturing industries. Environment and Planning C: Government and Policy, 27, 159–174.Find this resource:

Bande, R., & López-Mourelo, E. (2014). The spatial distribution of workplace accidents in Spain: Assessing the role of workplace inspections. MPRA Paper No. 56767, Munich Personal RePEc Archive.Find this resource:

Banks, J., Chandola, T., & Matthews, K. (2015). International encyclopedia of the social & behavioral sciences (2nd ed., Vol. 20, pp. 598–601). Oxford, U.K.: Elsevier.Find this resource:

Bassanini, A., & Caroli, E. (2015). Is work bad for health? The role of constraint versus choice. Annals of Economics and Statistics, 119/120, 13–37.Find this resource:

Belloni, M., Meschi, E., & Pasini, G. (2016). The effect on mental health of retiring during the economic crisis. Health Economics, 25, 126–140.Find this resource:

Bena, A., Giraudo, M., Leombruni, R., & Costa, G. (2013). Job tenure and work injuries: A multivariate analysis of the relation with previous experience and differences by age. BMC Public Health, 139(869), 1–9.Find this resource:

Bender, K. A., Green, C. P., & Heywood, J. S. (2012). Piece rates and workplace injury: Does survey evidence support Adam Smith? Journal of Population Economics, 25, 569–590.Find this resource:

Berniell, M. I., & Bietenbeck, J. (2017). The effect of working hours on health. IZA Discussion Paper No. 10524.Find this resource:

Black, N., Johnston, D. W., & Suziedelyte, A. (2017). Justification bias in self-reported disability: New evidence from panel data. Journal of Health Economics, 54, 124–134.Find this resource:

Black, S. E., Devereux, P. J., & Salvanes, K. G. (2015). Losing heart? The effect of job displacement on health. International Labor Relations Review, 68(4), 833–861.Find this resource:

Bloemen, H., Hochguertel, S., & Zweerink, J. (2017). The causal effect of retirement on mortality: Evidence from targeted incentives to retire early. Health Economics, 26(12), e204–e218.Find this resource:

Bloemen, H., Hochguertel, S., & Zweerink, J. (2018). Job loss, firm-level heterogeneity and mortality: Evidence from administrative data. Journal of Health Economics, 59, 78–90.Find this resource:

Boeri, T., & van Ours, J. C. (2013). The economics of imperfect labor markets. Princeton, NJ: Princeton University Press.Find this resource:

Bonsang, E., Adam, S., & Perelman, S. (2012). Does retirement affect cognitive functioning? Journal of Health Economics, 31(3), 490–501.Find this resource:

Bonsang, E., & Klein, T. (2012). Retirement and subjective well-being. Journal of Economic Behavior & Organization, 83(3), 311–329.Find this resource:

Boone, J., & van Ours, J. C. (2006). Are recessions good for workplace safety? Journal of Health Economics, 25, 1069–1093.Find this resource:

Boone, J., van Ours, J. C., Wuellrich, J. P., & Zweimüller, J. (2011). Recessions are bad for workplace safety. Journal of Health Economics, 30, 764–773.Find this resource:

Booth, A., & van Ours, J. C. (2008). Job satisfaction and family happiness: The part-time work puzzle. Economic Journal, 118, F77–F99.Find this resource:

Booth, A., & van Ours, J. C. (2009). Hours of work and gender identity: Does part-time work make the family happier? Economica, 76, 176–196.Find this resource:

Booth, A., & van Ours, J. C. (2013). Part-time jobs: What women want? Journal of Population Economics, 26, 263–283.Find this resource:

Browning, M., & Heinesen, E. (2012). Effect of job loss due to plant closure on mortality and hospitalization. Journal of Health Economics, 31(4), 599–616.Find this resource:

Browning, M., Moller Dano, A., & Heinesen, E. (2006). Job displacement and stress-related health outcomes. Health Economics, 15(10), 1061–1075.Find this resource:

Caroli, E., & Godard, M. (2014). Does job insecurity deteriorate health? Health Economics, 25(2), 131–147.Find this resource:

Case, A., & Deaton, A. (2005). Broken down by work and sex: How our health declines. In D. A. Wise (Ed.), Analyses in the economics of aging (pp. 185–212). Chicago, IL: University of Chicago Press.Find this resource:

Clark, A. E. (2003). Unemployment and social norms: Psychological evidence from panel data. Journal of Labor Economics, 21, 323–351.Find this resource:

Clark, A. E., & Oswald, A. J. (1994). Unhappiness and unemployment. Economic Journal, 104, 648–659.Find this resource:

Cottini, E., & Ghinetti, P. (2018). Employment insecurity and employees’ health in Denmark. Health Economics, 27(2), 426–439.Find this resource:

De Grip, A., Lindeboom, M., & Montizaan, R. (2011). Shattered dreams: The effects of changing the pension system late in the game. Economic Journal, 122, 1–25.Find this resource:

Eibich, P. (2015). Understanding the effect of retirement on health: Mechanisms and heterogeneity. Journal of Health Economics, 43, 1–12.Find this resource:

Eliason, M., & Storrie, D. (2009). Does job loss shorten life? Journal of Human Resources, 44(2), 277–302.Find this resource:

Eurostat. (2017). Key figures on Europe. Luxembourg: European Union.Find this resource:

Fé, E., & Hollingsworth, B. (2016). Short- and long-run estimates of the local effects of retirement on health. Journal of the Royal Statistical Society Series A, 179(4), 1051–1067.Find this resource:

Fitzpatrick, M., & Moore, T. (2018). The mortality effects of retirement: Evidence from social security eligibility at age 62. Journal of Public Economics, 157, 121–137.Find this resource:

Fonseca, R., Kapteyn, A., Lee, J., Zamarro, G., & Feeney, K. (2014). A longitudinal study of well-being of older Europeans: Does retirement matter? Journal of Population Ageing, 7(1), 21–41.Find this resource:

García-Serrano, C., Hernanz, V., & Toharia, L. (2010). Mind the gap, please! The effect of temporary help agencies on the consequences of work accidents. Journal of Labor Research, 31(2), 162–182.Find this resource:

Gielen, A., & van Ours, J. C. (2014). Unhappiness and job finding. Economica, 81, 544–565.Find this resource:

Gorry, A., Gorry, D., & Slavov, S. N. (2015). Does retirement improve health and life satisfaction? NBER Working Paper No. 21326.Find this resource:

Gray, W. B., & Scholz, J. T. (1993). Does regulatory enforcement work? A panel analysis of OSHA enforcement. Law and Society Review, 27, 177–213.Find this resource:

Guadalupe, M. (2003). The hidden costs of fixed term contracts: The impact on work accidents. Labour Economics, 10(3), 339–357.Find this resource:

Hallberg, D., Johansson, P., & Josephson, M. (2015). Is an early retirement offer good for your health? Quasi-experimental evidence from the army. Journal of Health Economics, 44, 274–285.Find this resource:

Hernaes, E., Markussen, S., Piggott, J., & Vestad, O. L. (2013). Does retirement age impact mortality? Journal of Health Economics, 32(3), 586–598.Find this resource:

Horner, E. M. (2014). Subjective well-being and retirement: Analysis and policy recommendations. Journal of Happiness Studies, 15, 125–144.Find this resource:

Insler, M. (2014). The health consequences of retirement. Journal of Human Resources, 49, 195–233.Find this resource:

Kassenboehmer, S. C., & Haisken-Denew, J. P. (2009). You’re fired! The causal negative effect of entry unemployment on life satisfaction. Economic Journal, 119, 867–889.Find this resource:

Kerkhofs, M., & Lindeboom, M. (1997). Age related health dynamics and changes in labour market status. Health Economics, 6(4), 407–423.Find this resource:

Kesavayuth, D., Rosenman, R. E., & Zikos, V. (2016). Retirement, personality, and well-being. Economic Inquiry, 54(2), 733–750.Find this resource:

Kolodziej, I. W., & García-Gómez, P. (2017). The causal effects of retirement on mental health: Looking beyond the mean effects. Ruhr Economic Papers No. 668.Find this resource:

Koning P., & Lindeboom, M. (2015). The rise and fall of disability insurance enrollment in the Netherlands. Journal of Economic Perspectives, 29(2), 151–172.Find this resource:

Kuhn, A., Wuellrich, Staubli, S., J.-P., & Zweimüller, J. (2018). Fatal attraction? Extended unemployment benefits, labor force exits, and mortality, NBER Working Paper No. 25124.Find this resource:

Lanoie, P. (1992). The impact of occupational safety and health regulation on the risk of workplace accidents: Quebec, 1983–1987. Journal of Human Resources, 27, 643–660.Find this resource:

Mazzonna, F., & Peracchi, F. (2017). Unhealthy retirement. Journal of Human Resources, 52(1), 128–151.Find this resource:

Michaud, P.-C., Crimmins, E. M., & Hurd, M. D. (2016). The effect of job loss on health: Evidence from biomarkers. Labour Economics, 41, 194–203.Find this resource:

Müller, T., & Shaikh, M. (2018). Your retirement and my health behavior: Evidence on retirement externalities from a fuzzy regression discontinuity design. Journal of Health Economics, 57, 45–59.Find this resource:

Nishimura, Y., Oikawa, M., & Motegie, H. (2018). What explains the difference in the effect of retirement on health? Evidence from global aging data. Journal of Economic Surveys, 32(3), 792–847.Find this resource:

Palali, A., & van Ours, J. C. (2017). Workplace accidents and workplace safety: On underreporting and temporary jobs. Labour, 31(1), 1–14.Find this resource:

Picchio, M., & van Ours, J. C. (2017). Temporary jobs and the severity of workplace accidents. Journal of Safety Research, 61, 41–51.Find this resource:

Picchio, M., & van Ours, J. C. (2018). The causal effects of retirement on health and happiness, Tinbergen Institute Discussion Paper, Rotterdam.Find this resource:

Pirani, E., & Salvini, S. (2015). Is temporary employment damaging to health? A longitudinal study on Italian workers. Social Science and Medicine, 124, 121–131.Find this resource:

Pouliakas, K., & Theodossiou, I. (2013). The economics of health and safety at work: An interdisciplinary review of the theory and policy. Journal of Economic Surveys, 27(1), 167–208.Find this resource:

Probst, T. M., Barbaranelli, C., & Petitta, L. (2013). The relationship between job insecurity and accident under-reporting: A test in two countries. Work & Stress, 27(4), 383–402.Find this resource:

Ravesteijn, B., van Kippersluis, H., & van Doorslaer, E. (2018). The wear and tear on health: What is the role of occupation? Health Economics, 27(2), e69–e86.Find this resource:

Rohwedder, S., & Willis, R. J. (2010). Mental retirement. Journal of Economic Perspectives, 24(1), 119–138.Find this resource:

Ruser, J., & Butler, R. (2009). The economics of occupational safety and health. Foundations and Trends in Microeconomics, 5(5), 301–354.Find this resource:

Salm, M. (2009). Does job loss cause ill health? Health Economics, 18(9), 1075–1089.Find this resource:

Schaller, J., & Stevens, A. H. (2015). Short-run effects of job loss on health conditions, health insurance, and health care utilization. Journal of Health Economics, 43, 190–203.Find this resource:

Shai, O. (2018). Is retirement good for men’s health? Evidence using a change in the retirement age in Israel. Journal of Health Economics, 57, 15–30.Find this resource:

Van der Heide, I., van Rijn, R., Robroek, S., Burdorf, A., & Propper, K. (2013). Is retirement good for your health? A systematic review of longitudinal studies. BMC Public Health, 13(1), 1180.Find this resource:

Winkelmann, L., & Winkelmann, R. (1998). Why are the unemployed so unhappy? Evidence from panel data. Economica, 65, 1–15.Find this resource:

Workplace Safety and Insurance Board. (2013). Workplace injury claim suppression: Final report. Technical Report, Prism Economics and Analysis.Find this resource:

Notes:

(1.) Individuals may report a health status to justify their labor market position. For example, disabled persons may report a worse health status to justify them receiving disability benefits. Using Australian data, Black, Johnston, and Suziedelyte (2017) concluded that non-employed individuals and disability benefit recipients were more likely to misreport their health status or exaggerate their level of disability.

(2.) Labor inspectors are public officials responsible for enforcement of labor standards who have the authority to initiate processes that may lead to legal action.

(3.) Not only actual retirement may affect health. A change in perspectives related to future pensions or early retirement may affect the health of workers. De Grip, Lindeboom, and Montizaan (2011) exploited a policy change in retirement rules in the Netherlands. In 2006, early retirement was made less attractive for civil servants born from January 1, 1950 onward. The person who was born 1 day too late had to work 13 months longer to reach the same level of pension benefits. The authors showed that this reduced attractiveness of early retirement and caused depression, which was partly related to the perceived unfairness of the policy change.