Economics is a social science and, as such, follows the scientific method. We isolate a problem and try to solve it, or observe a particular behaviour and we attempt to explain it. That entails a series of definitions and assumptions, the construction of a model and the formulation of a postulate or hypothesis. Once this is complete, we then undertake a statistical analysis and see if the model supports or rejects our hypothesis.
Not all disciplines follow such a methodology. For example, Reuters published an article citing a University of Victoria study that suggests that the privatization of BC liquor stores is responsible for an increase in alcohol related deaths in that province. (Click here for article) According to the article, a statistician from George Mason University takes issue with the way the study was done and the interpretation of the data. If, as the UVIC study suggests, the decline in government liquor stores causes more alcohol related deaths, then opening more government run liquor stores should result in fewer deaths.
What the study apparently omitted was a hypothesis of causation. That is, did the privatization of liquor stores cause the increase in deaths or was there another factor that happened to be related to the timing of the privatization and also to the deaths? We did a quick and dirty survey of some literature to find out what may affect the demand for alcohol and the increase in alcohol abuse.
Most studies suggest that alcohol is a normal good. When income rises, alcohol consumption rises. (See the BBC article). Another study tests this hypothesis for causation and finds that increased alcohol consumption can lead to higher incomes due to social networking. (See report on San Jose State U study) However, during a recession, when many people lose their jobs, depression can lead to an increase in alcohol abuse. (See University of Maryland study) This would appear to be conflicting. However, the first case deals with alcohol consumption and the second with substance abuse which are entirely different. (See article in the Irish Medical Times) In addition to income and depression being factors, genetics may also play a role. An article from the Scotsman suggests that Scottish residents born in Scotland or Ireland are more likely to suffer from alcohol abuse than those born outside Scotland. Celts are more susceptible than Anglos, Saxons or Normans? (Click here for article)
We would suggest that the authors of this particular study redo their analysis correcting for the recession. Our suspicion is that the opening of the private liquor stores coincides with the recession and that the “cause” of the alcohol related deaths has more to do with depression, induces by job losses and falling income, than it does with privatization. This would show a correlation, not causation.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment