Whose drinking does the liberalization of alcohol policy increase? Change in alcohol consumption by the initial level in the Finnish panel survey in 1968 and 1969.

AIMS: To study how the annual volume of alcohol consumed changed among population groups in different categories of initial consumption in a situation where per capita consumption strongly increased due to a new alcohol law.
DESIGN: A post-hoc quasi-experiment with a two-wave panel study. The regression to the mean effect was accounted for by using reference data from two auxiliary panel studies.
SETTING: Finland in 1968 and 1969, when per capita consumption increased by 46% in a year. Reference data come from Norway in 1975 and 1976, and the United States in 1993 and 1994. PARTICIPANTS: General population sample of 1720 (1292 men) aged 15-69 (response rate 91.4%). REFERENCE DATA: General population surveys in a Norwegian town (n=254) and in the United States (n=658). MEASUREMENTS: Annual volume of alcohol consumed.
FINDINGS: In Finland, the greatest change in observed consumption levels occurred among those with moderate consumption levels, but the same phenomenon was observed in the reference data. A comparison between the Finnish and the reference data indicated that in Finland consumption increased at all levels, and the higher the initial consumption level, the greater the increase.
CONCLUSIONS: A best estimate for the differential impact of future increases in alcohol consumption is that moderate and particularly heavy drinkers will be affected to a greater extent than will light drinkers or abstainers.