J. Lye and J. Hirschberg
Department of Economics
University of Melbourne

The good health of an individual is a combination of uncontrollable and controllable factors. Factors such as genetics and random events are uncontrollable while chosen activities such as smoking and drinking are controllable. In 1972 Grossman proposed and found a significant relationship between a person’s health and their earnings. In this article we examine the relationship between labour outcomes and activities that influence health – specifically drinking and smoking. The first part of this article summarizes the main findings in the medical literature that link these activities to health. However, these actions can also have economic consequences, such as affecting an individual’s employment opportunities and their productivity. The second part of the article discusses the consequences of engaging in these activities for workers and we examine the costs of these activities as reflected in job performance and how much people are paid.

Alcohol Consumption and Smoking: A Look at the Medical Literature

Alcohol Consumption

Moderate drinking is the only pattern of drinking that has been shown to have potential health benefits (ACSH 1999). The (US) National Institute on Alcohol Abuse and Alcoholism (NIAAA 1997) has defined moderate drinking as no more than two standard drinks per day for a man age 65 or under, and no more than one standard drink per day for men over the age of 65 and all women where a standard drink consists of one 12-ounce can of beer, one 5-ounce glass of wine, or a mixed drink containing 1.5 ounces of 80 proof spirits. The upper limit of moderate drinking is set lower for women because it has been found that the same amount of alcohol has a stronger effect in women due to differences in body size, body composition and the efficiency of alcohol metabolism (eg. NIAAA 1997a) relative to men. In Australia while per capita consumption of alcohol has stabilised over recent years Australians remain the highest consumers of alcohol per person of any English speaking population, ranking fifteenth in the world (NHMRC 1992).

Health problems associated with the excessive intake of alcohol include liver damage, heart disease, ulcers, cancers of the head and throat, alcoholism and impairment of mental functions. Overuse of alcohol has also been shown to contribute to obesity and high blood pressure. Women are less likely than men to be alcoholic, they are also more likely to abstain from consumption and have lower average alcohol consumption levels (Roman 1988). Youth are the heaviest drinking group and are also more likely to suffer from alcoholism. However, as men and women age they tend to reduce their consumption and are less likely to be alcoholics.

Multiple studies in medical research ( eg. Klatsky et al 1990) have described a U-shaped relationship between alcohol consumption and risk of death from all causes as shown in Figure 1. From this result one can infer that nondrinkers and heavy drinkers have an increased risk of death compared with moderate drinkers. This association is largely due to lower death rates from coronary heart disease among moderate drinkers. The risk of cardiovascular mortality has been estimated to be between 45 and 80 per cent lower for a moderate consumer of alcohol than for either an abstainer or an abuser (Klatsky et al 1990). This relationship has been observed in both men and women and in different age, geographic, and ethnic groups (NIAAA 1997). It is also independent of other risk factors for heart disease such as smoking and obesity (Moore and Pearson 1986). In addition, it has been found that all types of alcohol beverages are associated with reduced risk of coronary heart disease (Rimm et al. 1996).

Explanations for the moderate consumption of alcohol’s protective effect against heart ailments appears to be due to its effect in increasing the removal of cholesterol from tissue (see Baum-Baicker 1985). These researchers also report that diet changes associated with moderate drinking, the silicon content in wine and beer, decreased platelet aggregation and coagulation, and the ability to lessen stress and alter personality patterns associated with coronary heart disease also aid in lessening the risks of heart disease for moderate drinkers. In addition, they also report that heavy drinkers and abstainers suffer from higher rates of clinical depression than regular moderate drinkers.

Smoking

Smoking is now identified as a major cause of heart disease, stroke, several different forms of cancer and a wide variety of other health problems (eg. Doll et al. 1986). Peto et al. (1992) estimated that, during the 1990s tobacco use will be the largest single cause of premature death in developed countries causing approximately 30% of all deaths between the ages of 35 and 69. In 1992 the weight of cigarettes consumed per adult in the population was 1,825 grams. In 1992 the estimated number of cigarettes consumed per adult smoker was 9,962 (Winstanley et al. 1995).

Manning et al. (1991) estimate that smoking reduces the life expectancy of a 20 year old by about 4.3 years or 7 minutes per cigarette. Furthermore, half of the cigarette smokers who commence smoking as teenagers are predicted to die due to tobacco related illness. Cancers may begin to occur in people aged in their 30s if they have been smoking for 15-20 years (Peto et al. 1994). Smoking causes a build-up of fatty deposits in the artery walls and contributes to poor circulation, high blood pressure, heart disease and stroke. Smoking also paralyses and ultimately destroys the lining of the lungs and damages the air sacs through which oxygen passes into the blood in exchange for carbon monoxide. Toxic chemicals in cigarette smoking damage the lung capacity and clearance function, leading to increased coughs and respiratory infections among smokers (Brown et al. 1987 ). "Pack-a-day" smokers when compared with people who have never smoked have 10 times the risk of lung cancer and twice the risk of heart disease (Savitz et al. 1993).

Since the 1980s there have been numerous studies that demonstrate harmful health effects in adult non-smokers from the inhalation of other people‘s tobacco smoke (passive smoking). These include irritation to the eyes, nose and throat, allergy, headache, nausea, decreased lung function, chronic airways disorders, lung cancer and emphysema (Winstanley et al. 1995). A number of surveys (eg. Hill 1986) show that a majority of smokers as well as nonsmokers believe that passive smoking is harmful to health and this awareness has increased over time.

Alcohol Consumption and Smoking: The Economics Literature

Alcohol Consumption

An extensive literature indicates that problem drinking is associated with substantial adverse consequences in employment. In this literature problem drinking is generally defined as heavy drinking and also the diagnosis of an alcohol-related disorder, alcohol abuse and/or dependence. The precise definition of heavy drinking differs by author. One example is Hamilton and Hamilton (1997) who define heavy drinkers as those who drank at least once a week in the previous twelve months and drank eight or more drinks on one or more days in the previous week.

The pattern and extent of problem drinking vary between jobs and between different subgroups of workers (Plant 1979). Economic costs of problem drinking include increased absenteeism including sickness absence, increased numbers of industrial accidents, increased employee replacement costs due to alcohol related terminations and premature death and decreased worker efficiency (Blose and Holder 1991). Webb et al. (1994) using a sample of all employees of a large industrial plant manufacturing metal products show problem drinkers were 2.7 times more likely to have injury-related absences than non-problem drinkers. Jones et al. (1995) using New Zealand survey data from November 1990 to May 1992 estimate a 25% reduction in work performance among heavy alcohol users.

Lee et al. (1990) examine data on men from the Scottish Heart Health Study and conclude that the unemployed were likely to drink more and engage in binge drinking behavior than the employed. A number of other studies all using data from the US that have examined labour supply and problem drinking have produced contradictory results. Benham and Benham (1982) found no significant relationship between problem drinking and employment. Kenkel and Ribar (1994) also find little evidence that alcohol problems have a negative impact on labour supply although they only consider a sample of young adults, none of whom is older than thirty-one. On the other hand Mullahy and Sindelar (1991) who focus on the 30-59 age group find that the effects of problem drinking vary across the life cycle and by gender. In a further study of this age group Mullahy and Sindelar (1996) find that for both men and women, problem drinking results in increased possibility of unemployment.

There have also been a number of studies which have examined the consequences of drinking on wages using regression analysis. Regression analysis is concerned with the study of the relationship between one variable called the dependent variable and one or more other variables called explanatory variables. In these studies earnings is the dependent variable and the explanatory variables include: previous labour market experience, the individual’s education, gender, family status, race, occupation, health and alcohol consumption is used. These studies have only focussed on the relationship between alcohol consumption and earnings for those individuals with a relatively stable employment pattern. In these cases the results have tended to be more consistent. Berger and Leigh (1988) conclude from their US sample that drinkers earn more than non-drinkers. Using Canadian data Hamilton and Hamilton (1997) extend this research by decomposing the group of drinkers into heavy, moderate and light drinkers. They find that moderate alcohol consumption leads to increased earnings relative to abstention. However, heavy drinking leads to reduced earnings relative to moderate drinking. Heavy drinkers also receive lower returns to higher education than other drinker types. In an analysis of US workers Heien (1996) found that moderate drinkers earn more than either abstainers or abusive drinkers. Zarkin et al. (1998) use US data on workers between the ages of 30-54. Their results suggest that men who consume alcohol have approximately 7% higher wages than men who do not drink, and this estimated wage premium is approximately the same over a wide range of alcohol consumption. For women, the estimated alcohol use premium is approximately half as large as for men however this result may only be an artifact of the data.

Smoking

Smokers on average suffer poorer health than non-smokers and may be more likely to miss work due to illness. Ault (1991) using US data from 1968 show that smokers miss no more work than non-smokers. Rather smokers tend to be younger, heavier drinkers and blue collar workers and this group tend to have high absentee rates regardless of whether or not they smoke. However, more recent research has shown some relationship between smoking and absenteeism. Bush and Wooden (1995) using the Australian National Health survey data for 1989/90 show smoking is consistently found to have a large and significant impact on absence from work. The probability of a male smoker being absent from work is estimated to be 66% greater than that for a male who has never smoked. For women, the corresponding figure is just 23%. Ryan et al. (1992) examines absenteeism among Telecom Australia’s workforce. They estimated that during the 1991-1992 financial year smoking related absenteeism cost this employer $16.5 million. Leigh (1995) using US data for 1986 shows that smoking appears to make a moderate contribution to absenteeism for men but only a slight contribution for women. It is argued that the reason there is a gender difference is that women have not been smoking as long as men and also tend to have a lower smoking intensity.

There have been a number of cases in Australia in which compensation has been awarded due to discomfort or disease caused by second-hand tobacco smoke in the workplace (Winstanely et al. 1995). If employers, co-workers or customers dislike smokers, discrimination against smokers leading to lower wages could result. Using US data Levine et al. (1997) examine differences between smokers and non-smokers. After allowing for differences in education, occupation and other personal characteristics they conclude that workers who smoked earned 4-8% less than non-smokers. They conclude that discrimination against smokers may be the chief cause of the observed wage differential.

A consequence of concerns over the danger of passive smoke has been the widespread institution of workplace smoking bans in Australia. Research has shown that bans bring about a reduction in numbers of cigarettes smoked on workdays (Borland et al 1990). However, bans may also result in workers taking time out for a cigarette outside the building which can be expensive for employers. One estimate is that a ‘quick smoke’ costs up to 25 minutes of productive time (Stickels 1994). This can also cause resentment among non-smoking staff who may feel that smokers are not working and may be a reason for employers to discriminate against them.

The Combined Effects of Alcohol Consumption and Smoking

The above studies have only considered the impact on labour outcomes of one of these activities at a time. However research shows (Gulliver 1995) that "smokers drink and drinkers smoke". Moreover the heaviest alcohol consumers are also the heaviest consumers of tobacco. Postulated mechanisms for the concurrent use of alcohol and tobacco are that either drug may increase the rewarding effects of the other or that either drug may decrease the unpleasant effects of the other. This suggests that the impact of both drinking and smoking on labour outcomes should be jointly considered. It should be noted that the method most widely used for the analysis of these effects is the linear regression statistical technique. A very useful characteristic of this method is its ability to accommodate multiple effects simultaneously. Thus one can determine the effect of both smoking and drinking at the same time and in this way determine if their effects can be separated. Hirschberg and Lye (1999a, 1999b) using Australian data consider the consequences of both drinking and smoking on wages. The conclusion is that smoking has a negative impact on wages and the previously determined beneficial effects of moderate drinking only exist for the non-smoking group. Thus smoking seems to negate any of the positive aspects of moderate alcohol consumption.

References