Aims: To estimate changes in liquor sales occurring in Washington, USA and bordering states following the privatization of government controlled liquor stores.
Design: Trend analyses of data from January 2009 to October 2014 of a natural experiment beginning 1 June 2012, when liquor prices increased and the number of stores selling liquor increased in the state of Washington. Difference-in-differences (DID) models and interrupted time-series methods were used.
Setting: Washington and bordering counties in Oregon and Idaho.
Measurements: Monthly liquor sales in 9-l cases.
Findings: DID model estimates of adjusted change in liquor sales as a result of privatization produced a cross-model average increase of 10.1% in Oregon and 8.2% in Idaho (both P < 0.001). Similar results were found using interrupted time-series. This represents a total loss to Washington of 89 865 l of liquor, 0.226% of total Washington sales, for June 2012 to May 2013. Adding these sales to Washington totals for fiscal years 2013 and 2014, we find that per-capita spirits sales were 5.80 l in both 2012 and 2013, declining slightly to 5.76 l in 2014.
Conclusions: The privatization of liquor sales in the state of Washington, USA in 2012 and the price increases associated with this resulted in a significant increase in sales in bordering counties in the states of Oregon and Idaho. However, the amount of alcohol sales and revenue lost by Washington was relatively small. Per-capita liquor sales in Washington appear to have remained flat after privatization.
Keywords: Border; monopoly; policy; privatization; spirits; tax; time-series.
© 2016 Society for the Study of Addiction.