Listener 28 April, 2001.

Keywords: Health.

Some social sciences – demography, economics, geography and psychology – started off well because they had could measure the concepts they were dealing with. Others – anthropology, politics, sociology – have never been as successful. But that something cannot be measured does not mean it is unimportant. We cannot quantify culture and related behaviour and institutions. Yet they seem to be a key elements in economic performance. Contrariwise, well-constructed measures of economic performance, such as per capita GDP, may not be good indicators of our social objectives.

The previous paragraph is a reflection as I travel to an international working party on the social costs of substance abuse. What are we quantifying those costs for? Suppose you knew the exact cost of substance abuse in New Zealand. You might say ‘gee-whiz’ but is there a practical use? How does one encapsulate in any number the heartache that the abuse causes?

There are prosaic reasons The costing exercise becomes a collecting together of what is known. It identifies the gaps in the data (and the accuracy of what is there). Consider the morbidity costs of alcohol misuse – that is the damage to people’s health and welfare which does not appear as death or reduction in material possessions. It includes the side effects of excessive drinking on the drinker, the physical damage done by drunks to their partners, children and others, and the anguish that the drinking causes to the drinker’s close ones. How big a problem is it? Most individuals are distressed by at least one serious alcohol-precipitated incident in their life, over which they had no influence – usually an ongoing one involving a friend or relative. It is also a public policy issue insofar as public policy can reduce these morbidity incidents and the pain they cause.

When I tried to measure the morbidity costs from drinking, I found lots of anecdotes but not much solid information. To my astonishment, the material I scrambled together suggested that morbidity costs are not too different from (actually it was slightly bigger than) the mortality costs of alcohol misuse. I would not go to the gallows defending the calculated figure, but practically it indicates that here is a social problem which could be easily overlooked because of the lack of data. The policy consequence could be that we put effort into, say, preventing road accidents caused by drunk drivers, for which there is some reasonable data, while we insufficiently respond women and kids being terrified and beaten up by drunks. We desperately need to improve the data here.

One of our biggest lacunas is the costs of substance abuse induced crime. There are severe technical problems. What is causation? Just because the criminal was drunk – or smoking, or on a high, or wearing red socks – when the crime was committed, does not prove that drinking – or smoking or illicit drugs or red socks – caused the crime. (Moreover they may be stone cold sober when they burgle or rob to fund an addiction.) Not long ago the technical problems seemed intractable, but some recent innovative overseas research hints that the social costs from crime and substance abuse are very significant. That seems consistent with the anecdotal evidence here in New Zealand. I worry that the omitting the costs of crime from my figures, has contribute to our ignoring the interaction between drugs and criminality.

Data also makes comparisons possible. For instance, about five times as many people die from smoking than drinking. However, it turns out that smokers die in older age groups, but many of the deaths from alcohol are of young adults. Counted in years of life lost, the relativity between smoking and drinking mortality is only about double. Policy-wise, the result draws attention to young drinkers, asking whether we should be putting as much effort into them as we do for their smoking.

Another important (overseas) finding is that the licit drugs of alcohol and tobacco do far more damage than the illicit drugs such as cannabis and heroin. The data to test it does not exist in New Zealand, but the difference is so big in Australia it must apply in New Zealand. (That their social costs are smaller need not make the policy problem less urgent, nor the anguish that the families with addicts face less stark.)

Notice how in the social sciences the scientific curiosity about what the world is like is continually interacting with a policy interest of how to make it better. Careful measurement helps pursue both objectives. But that measurement can only take us so far. For a scientist the limitations may not be important. There are so many things to investigate, he or she can confine research to that what can be measured. (There is a compelling argument that only that which can be measured – in some sense – can be investigated by a scientist.) But there are things we cannot really measure – anguish, love, hope, misery, joy – which are an integral part of our humanity, more so than anything an economist scientist works on.