Valuations of Life and Health Risks: Evidence from the Indian Labor Market
S.Madheswaran* and K.R.Shanmugam**
*Gokhale Institute of Politics and Economics,
Pune 411 004, India. Email:
**Madras School of Economics, Chennai.
Abstract
Development in environment can have a significant impact on it. This is true of projects in major sectors, including power and energy, industry, transportation, sanitation and sewage. Exposure to environment contaminants may cause risks to human life and health. In order to regulate these risks, the government undertakes various projects that impose costs on society in exchange for reducing the risks. To determine whether a project is socially desirable, one needs to compare the value of reducing risks to the cost of such reductions. Several methods have been proposed in the literature for estimating the implicit prices for life and health. The willingness to pay approach (WTP), however, has been increasingly considered as the relevant method. The approaches to estimating WTP to reduce the risks of death and injury fall into four basic categories: wage differentials between alternative occupations with different statistical risks, contingent valuation studies, consumer market studies and foregone earnings. This paper deals with the first category. There are considerable empirical works to analyze the workers employment decisions in the market with potentially hazardous work. However, most of the work is related to developed countries. Empirical study on this problem is practically non-existent in India and perhaps in developing countries, mainly due to data constraints. In this backdrop, this paper attempts to estimate the values of life and health risk based on compensating wage differential framework using the primary data from Chennai, southern part of India. The empirical part of this paper examines the worker’s behavior in choosing their job risks and the role trade unions in influencing the wage-risk trade-off. It also analyses the problem of sample selection and its effects on the estimated compensation for job risks. The empirical results provide a strong support for the efficiency of labor market in deriving the optimal risk level. They also imply that the unionists receive a higher compensation for work related risks than the non-unionists. Allowing for sample selectivity produces a down ward bias in the estimated union wage differentials for deadly hazards. The value of statistical life is also calculated. The empirical results show that the calculated value of statistical life is Rs.15.55 million and Rs.5.49 million and the estimated value of injury is Rs.5598 and Rs.2059 for the union and non-union sector workers respectively. Comparison of our estimated value of life with those from developed nations indicates that as expected, our value is lower than the values from developed nations. The estimated values life and injury can be used to value reductions in risk of death achieved by industrial safety programs or environmental health programs. Hence, the study may be useful to policy makers, international agencies and researchers evaluating health projects in developing nations
Key Words: Compensating wage differentials, hedonic price, Value of life and injury
JEL Classification: J17, J28, J31
Valuations of Life and Health Risk: Evidence from the Indian Labor Market
S.Madheswaran* and K.R.Shanmugam**
*Gokhale Institute of Politics and Economics,
Pune 411 004, India. Email:
**Madras School of Economics, Chennai.
1.Introduction
Environmental development can have significant impacts in areas such as power and energy, industry, transportation, sanitation and sewage. Exposure to environment contaminants may cause risks to human life and health. In order to regulate these risks, the government undertakes various projects that impose costs on society in exchange for reducing the risks. To determine whether a project is socially desirable, one needs to compare the value of reducing risks to the cost of such reductions.
Several methods have been proposed in the literature for estimating the implicit prices for life and health. They include cost of illness approach, human capital approach, insurance approach, court awards and compensation approach and portfolio approach (Linnerooth, 1979). The willingness to pay (WTP) approach, however, has been increasingly considered as the relevant method. WTP is typically measured by analyzing prices paid for goods and services. Prices paid for preventing health and death risks cannot directly be obtained because prevention of these risks is not directly purchased in the market. However, there are instances when these prices can be observed or measured.
There are two principal methodologies for measuring these prices. The first known as the contingent valuation approach, rests on data generated through questionnaires ( Alberini et.al 1977, Gerking et.al 1988). In this approach, individuals are directly asked how much they would be willing to pay to reduce, for instance, their death risks at work or in traffic accidents. The second is the revealed preference approach. This method infers the hedonic (that is quality adjusted) value of an environmental good affecting the value of a market such as air quality. This method relies on property values and wage data. The approach using wages is popular because the availability of information on work related environmental risks and associated wages that workers receive enable the estimation of the market generated wage-risk trade off.
Several empirical studies emerged to estimate the value of life or injury, but most of them concern developed nations (viscusi, 1993). A study on this topic is practically scanty in developing countries, mainly due to data constraints. Those valuing the health impacts of projects in developing countries have two options. First, they can develop monetary value estimates based on data from developed nations by making appropriate adjustments (using per capita GDP). This simple transfer mechanism does not take into account differences in WTP values between different countries due to differences in factors like living standards, culture and educational attainment. Second, they can rely on the human capital approach such as loss of earnings. But this approach provides no guide to action when there are a variety of impacts on unknown value and ignores the quality of life lengthened. New research on valuing health and death risks in developing countries is the only way to resolve this problem (Asian Development Bank, 1996).
In the past, India had neglected the environmental consequences of economic growth. However, in recent years, its approach on environmental problems has changed. It implements many environmental programs and spends huge amounts on health and safety programs. Since resources are scarce, it is essential to evaluate these programs and reallocate funds to achieve maximum net benefit to society. In this context, we estimate the compensating wage differentials for job related fatal and injury risk using the data from Indian labor market.. The problem posed by sample selection is also addressed and its impact on the estimated wage differential is investigated.
The organization of the paper is as follows. Section 2 deals with theoretical and empirical studies existed in this area. Section 3 gives brief outlines the source of data and econometric methodology. Empirical results are discussed in Section 4 followed by a conclusion.
2.Review of Literature
2.1. An Economic Settings for Job Risks and Wages
The causes of accidents (or job risks) are related to technical and human factor; it has attracted the attention of Psychologists, Sociologists and Engineers for a long period. In recent years, economists have used their analytical tools to investigate the problem of employment accidents or risks in a different context. They consider job hazards as a component of job evaluation system that gives each job rating, which in turn affects the wages, that is, job risk is taken as one of the determinants of wage differentials. They use the theory of compensating or equalizing differentials, developed in the context of hedonic reconstruction of demand theory to analyze the wage-risk relationship. The theory posits that jobs with more risks require a wage premium to attract workers, other things being equal. This extra wage or premium is called compensating wage differentials. However, the required risk premium may be quite different for different people. Firms attempting to attract workers to hazardous jobs will only offer an adequate premium to staff those positions with capable employees. This premium will provide a sufficient inducement for workers most willing to accept risks, while those requiring a very large wage premium will tend to select safe occupational pursuits. Thus, the heterogeneity of workers behavior and firms form the basis for labor market equilibrium and the compensating wage differentials acts as an equalizing or a balancing factor in keeping the workers in the risky jobs. That is, compensating differentials play a vital role in allocating labor and resources amongst their various uses.
Suppose there is no compensating differentials, then no worker will choose a dangerous or an unpleasant work; choosing instead to accept equal wages for employment in relatively clean, safe and pleasant job. Further, the need to pay compensation for risky job provides financial incentive for the firms to invest in safety equipment and other risk reducing controls within the workplace in order to reduce the wage costs necessary to attract workers. However, making the work place safer typically involves substantial costs. Hence, a firm has two principal choices. It can simply pay the workers for incurring or it can reduce its wage costs by making larger safety investments in the work site. Typically these mechanisms are inter-related. The firm increases its expenditures on safety until the incremental wage reductions generated by improved safety no longer exceed the added costs of these improvements. Workers welfare enters the firm’s calculations through the influence of the risk level on wages. The worker’s own valuations of risk in effect determine the price the firm must pay if it does not financially worthwhile to diminish the risk.
Thus, the economic analysis of wage-risk trade off reflected in decisions of worker and firm suggests that these choices produce powerful incentive for safety. To the extent that expected costs of industrial accidents to firms reflected in the prevailing compensating wage differentials for hazardous work, the firm has an incentive to invest in safety equipment, supervision and training to the socially optimal extent, as the firm must bear the full social costs of accidents. In recent years, the notion of compensating wage differentials due to work accident has been generalized in terms of probability. It considers a situation in which workers face some lotteries on their health status. Thus the problem of the worker is to choose job risks to maximize his expected utility. Thus, the problem turns to be one of testing the economic rationality of workers in the markets with uncertainty. To estimate the compensating differentials for job risks, economists during the last fifteen years, have, thus, approached the problem from a microanalysis taking an individual worker as unit of analysis. They have used cross-section data pertain to individual workers’ wage, job risks and other factors which affect wages. The present project proposal follows this economic tradition.
One interesting application of this approach is that it can be used to place a money value on fatal and non-fatal accidents. By accepting wage premium for job hazards, workers implicitly reveal their value of life and limb. Thus, the estimated wage premium for job related risk can be used to determine the economic value of saving a human life and limb.
Since the firm can compensate the workers for risks either through ex ante compensation (i.e., compensating wage differentials) or ex post compensation (i.e., compensation benefits etc.), economists have recently analyzed the role of this expost component of compensation in affecting the wage-risk trade-off through its financial mechanism. In India, both ESI scheme and WCA (Workmen Compensation Act) play vital role in providing compensation for industrial accidents. However, these partial replacement benefit amounts can’t eliminate all losses in welfare resulting from work accidents. Hence, the problem of assessing whether these benefits are optimal is important.
The life-cycle issues such as variations across workers in the potential losses resulting from death and the discounting problem are also important. Three different approaches are available in the existing literature to analyze the issues. They are: discounted life-years-lost approach, life cycle and markov model of the lifetime job choices. Since no single modeling is dominant in terms of its theoretical and empirical properties, three approaches are equally plausible. Hence these 3 approaches will be utilized in the proposed study to assess whether worker’s choice with respect to job risk is consistent with rational behavior in the market that combines the elements of both intertemporal choice and uncertainty.
The alternative way is interview or contingent valuation method (CVM). In this approach, the individuals are asked directly what their trade-offs or how much they would be willing to pay for a particular risk reduction. In other words, the interviewer asks individual quite directly what their subjective trade-offs are. The CVM receives attention in valuing workers life. Viscusi and O’connor (1984) have interviewed the workers in chemical industry. They ask a series of questions pertaining to risk perception of the workers. They elicited willingness to accept for non-fatal job risks and estimated value ranges from $10,000 to $13000. Gerking, Hann and Schulze (1988) are the first study to examine contingent values of job related fatal accidents and willingness to pay for avoiding those risks. They have estimated the marginal value of safety by directly asking respondents about their willingness to substitute money for changes in job related fatal accidents. They adopt both WTP and WTA principles. He estimated value of life is $2.66 million in WTP measures and $6.82 million in WTA measures. This method is free from measurement error in risk-measures and failure to control personal and job characteristics. However, this method is not free from criticism. It is criticized that: (1) respondents have no incentive to give thoughtful or honest answers and the process of thinking about choices involving small probabilities is notoriously difficult; (2) individuals may misrepresent their preferences if they believe that their responses will affect the benefits they will receive or taxes they must pay to support a public program to save lives. Therefore Blomquist (1982) has remarked that contingent values may be unreliable due to hypothetical or strategic bias.
2.2. Theoretical Framework
The hedonic approach views that the labor market transactions as a tied sale. On the one hand, workers sell their services for wages. At the same time, they purchase non-pecuniary work attributes. These attributes may vary from job to job. Hence, the workers exercise a choice over preferred job attributes by choosing an appropriate job. On the other hand, firms simultaneously buy the services and characteristics of workers and sell their attributes of jobs offered to the market. Since the characteristics may differ among workers, firms have a choice to exercise. An acceptable match occurs when the preferred choices of both the employer and the employee are mutually consistent. The actual wage, therefore, embodies a series of implicit prices for both workers characteristics and job attributes such as pace of work and probability of injury/death accidents. Controlling for other aspects of a job, one can estimate the wage premium that workers receive for job related risks. Thus, the observed distribution of wages clear both markets over all worker characteristics and job attributes. In this sense labor market is viewed as an implicit market in job and worker attributes.
The purpose of this section is to illustrate the properties of the optimal job choice of a worker who is choosing from a set of job opportunities that involve the same number of work hours but have different probabilities of an adverse consequence. Consider a state dependent utility model. Let Y be the initial asset and W(p) be the schedule of earnings for the jobs with probability p of an event that leads to ones death or injury. Therefore, (1-p) is the probability that the worker remains healthy. Suppose U(x) denotes the utility of being healthy state and V(x) denotes the utility of being injured or dead, for any given consumption level x which is equal to Y+W(p). It is assumed that the wage and utility functions are continuous and twice differentiable. Let be the shadow price of the good constraint. It is also assumed that the worker would rather be healthy than not (i.e. U(x) > V(x) >0); the marginal utility of consumption is positive and greater in health state than in ill health state (i.e., Ux > Vx > 0) and the marginal utility of consumption is diminishing or the workers are either risk averse or risk neutral (i.e., Uxx , Vxx < 0). The workers optimal choice among hazardous job alternatives is determined by maximizing the lagrangian given by