Valuing psychological and behavioural effects of alcohol dependence : A contingent valuation study
Claude Jeanrenaud 1,2, Sonia Pellegrini1, Nicolas Marmagne1
1Institute for Economic and Regional Research (IRER), University of Neuchâtel, Switzerland
2Center for Studies in Public Sector Economics, Bern, Fribourg and Neuchâtel, Switzerland.
Aim of the study Method: The Contingent Valuation survey Statistical analysis
The aim of the paper is to assess the intangible costs of
alcohol dependency. It is part of a broader study A direct approach Choosing the right approach Internal validity
commissioned by the Swiss Federal Office of Public Health
aimed at valuing the social cost of alcohol abuse in The direct, or expressed preference, approach – Applying the CV method to alcohol dependency runs The internal validity of the CV survey was checked The variable CULTURE reveals the attitude of the
Switzerland. Until now there has been no attempt to value contingent valuation (CV) method or conjoint analysis into a number of challenges. using a multiple regression model. As predictor of the respondent to the disease : 1 means that the disease
the health implications of alcohol dependency. – seems to be the only feasible way to assess the WTP, we used the traditional economic and socio- is perceived as a fatality or a punishment, 0 means
human (or intangible) costs of alcohol because it The health and social implications of alcohol abuse demographic variables, such as adjusted income, that the disease is viewed as a hardship to overcome.
The usual way of valuing the cost of drugs – licit or illicit – would hardly be feasible to observe the pain and are numerous and complex; and the cognitive gender, age, social status, size of household and area of
is the so-called “cost-of-illness” approach, which is based suffering of an alcoholic on a market substitute (the limitation of the respondents had to be taken into residence (urban v rural). If the respondent considers that the quality of life of
on the human capital method. The major shortcoming of money-risk trade-off of a bartender requiring a higher account when designing the questionnaire. the other family member is greatly reduced (the
this approach is that it only values any implications of wage because he is at risk of becoming an alcoholic). A proxy for the marginal utility of income (MUI) has variable QUALITY is 1).
alcohol which result, directly or indirectly, in loss of Several consequences of alcoholism, such as been introduced into the model. The standard
The CV method is flexible and can therefore be used violence in the family, social isolation, disgrace to
production. The psychological and behavioural implications hypothesis is that all individuals have the same income- Finally, we expect the WTP to be lower if the relative
to value the social cost of the disease - resource and the family, etc. are not just facts but also highly
of alcohol abuse are ignored. utility function, thus variable income reflects both whom the respondent has imagined being an alcoholic
non-resource costs - or only the loss in quality-of-life. emotive issues. ability-to-pay and differences in the marginal utility of is not among the closest relatives (RELATIVE = 0).
We have used the latter, the so-called restricted
The second aim of the paper is to test the following the money transfer. The coefficient of the variable MUI
approach. The CV method relies on a hypothetical For practical and ethical reasons, the questionnaire
hypothesis: the inclusion of new exploratory variables is significant and has the expected sign.
market to elicit a person’s willingness-to-pay for a cannot be administered to alcoholics. The usual
reflecting the attitudes and values of respondents will
specific change in their health state. A scenario is way of dealing with this problem – the ex ante
significantly improve the capacity of the regression model
proposed to respondents, who have to state how approach – is not practicable either. Individuals
to predict willingness-to-pay (WTP). OLS estimate of WTP equation
much they would be willing to pay to benefit from a from among the general population are incapable of
better health state (ex post approach) or from a imagining that they might one day become an Linear model Logarithmic model
Alcohol dependency reduced risk of contracting the disease (ex ante alcoholic.
approach). Variable Coefficient p- value Coefficient p-value
In Switzerland there are 300,000 people suffering from Constant -109.78 0.457 0.98 0.301
alcohol dependency - 5% of the population. The disease INCOME 0.091*** 0.000 0.51*** 0.000
has both behavioural and physiological features. Physical Scenario and elicitation question MARGINAL UTILITY OF
218.89** 0.011 0.36*** 0.001
dependency is measured by degree of tolerance – the
The interviewee, having previously been informed of the health, behavioural and psychosocial consequences of alcohol CATEGORIES 161.16** 0.043 0.18* 0.081
need for increased amounts of the product to achieve
dependency, is first asked to imagine that a member of their household is an alcoholic. Then they are informed that a RESIDENCE 167.00** 0.031 0.24** 0.017
drunkenness – and withdrawal symptoms. Psychological HOUSEHOLD 367.19*** 0.000 0.54*** 0.000
new, very effective treatment has been approved. To remain effective, the treatment must be administered for the rest
dependency focuses on the user’s need for the product in EXPERIENCE 206.05*** 0.007 0.27** 0.048
of the patient’s life. The respondent is asked if they wish their relative to receive the treatment. To elicit their WTP, they
order to face the problems of everyday life and to feel CULTURE -268.85*** 0.003 -0.23** 0.049
then have to state the maximum (monthly) amount they would be prepared to pay to have their relative receive the
well. Alcohol dependency leads to progressive loss of RELATIVE 0.32*** 0.001
physical and mental capacities. The development of n 184 184
alcohol dependency may occur over 5-25 years of R2 0.310 0.376
R2 adjusted 0.282 0.347
repeated, excessive consumption. Prob (F-stat) 0.000 0.000
Survey result Significant at 1% (***), 5% (**), and 10% (*).
Cost categories 240 individuals aged 18+ from the general population in French-speaking Switzerland were interviewed in July 2002.
Respondents were selected using the following criteria: gender, age and social stratum.
The social cost of substance abuse can be divided into
three parts. The first part corresponds to the resources Four respondents refused to opt for the treatment even if it were provided at no cost. Eight individuals wished their Discussion
that are used for medical treatment, repair of damage to relative to receive the treatment, but there was no amount they were willing to pay for this. Follow-up questions have The main aim of the paper was to assess in monetary terms the intangible costs of alcohol dependency. It is done by
property, court decisions and law enforcement, and thus shown that the well-being of these 8 people would have improved, and therefore we considered these bids as “false valuing the loss in quality-of-life for both the alcoholic and those closest to them.
diverted from other beneficial use. These are called direct zero bids”. The outliers – abnormally high bids – were dropped. The respondents were considered as victims of a Contrary to the opinion often expressed by authors of works on cost-of-illness, it is possible to measure the intangible
costs. The second part corresponds to the loss of hypothetical bias. costs of illness in monetary terms: the contingent valuation method can be used to achieve this.
resources resulting from the rate of increased impairment
The inclusion of variables that reflect a respondent’s attitudes significantly increases the model’s predictive capacity.
of individuals with alcohol abuse – reduced participation in
the labour force, increased absenteeism – and from
50 Omitting intangible costs leads to a far-from-realistic picture of the social burden of alcohol dependency. The cost in
alcohol-related premature death. These are called indirect Mean WTP 40
terms of loss of quality-of-life (CHF1.7bn) exceeds the indirect costs of mortality and morbidity (CHF1.5bn).
costs. The third part of the social cost consists of the
intangible costs – sometimes called human costs – which The mean WTP is of CHF486 per month or CHF5832
are due to loss of quality-of-life . These intangible costs annually. The lowest monthly bid was CHF20, the 30
are real in that they lead to a reduced well-being of the highest bid CHF4000. On average, the burden of
population, but they do not result in a diversion or a alcohol dependency, for both the alcoholic and the 20 Institute for Economic and Regional Research (IRER)
sacrifice of productive resources. If intangible costs are other members of the household, is close to CHF6000 University of Neuchâtel
reduced, no additional resources become available to meet per case. The prevalence of alcohol dependency in 10 Pierre-à-Mazel 7, 2000 Neuchâtel, Switzerland
the needs of society. The paper focuses on the non- Switzerland is 300,000. The annual social burden to Tel. 0041 32 718 14 71
resource costs only. the country of the disease amounts to CHF1.7 billion. 0 e-mail: email@example.com ; firstname.lastname@example.org
0 500 1000 1500 2000 2500 3000 3500 4000