Predictors of Canadian legislators’ support for tobacco control policies

Abstract

It is clear that regulatory strategies can be effective in reducing tobacco use. Because legislators ultimately determine whether many of these policies are enacted, they are a key focus for tobacco policy research. This study identifies political and personal predictors of Canadian legislators’ support for tobacco control policies. Data are from a 1996–97 survey of federal, provincial and territorial legislators. Multivariate regression analysis was used to assess relationships between five groups of variables (political factors including political ideology, personal characteristics, tobacco experiences, tobacco knowledge, interest group saliency) and support for tobacco control based on an 11-item scale. Support for tobacco control varied by political party. Support was higher among legislators who thought government had a duty to promote healthy lifestyles, knew second-hand smoke could cause lung cancer, knew tobacco caused more deaths than alcohol, and said they wanted more contact with medical associations about tobacco issues. Support was lower among current smokers and those with tobacco industry jobs in their ridings. The findings indicate that political party membership cannot be ignored in enlisting legislator support for tobacco control. It also appears that legislators who oppose tobacco control measures may not be opposed to tobacco control per se, but are more generally opposed to a government role in health promotion. Thus, public health professionals and tobacco control advocates need to be more attentive to the way tobacco control issues are framed for particular legislators. Further, meetings with health groups about tobacco issues would be welcomed by many legislators; non-governmental organizations and other health advocates could work to increase tobacco knowledge among legislators.

Introduction

The implementation of effective policies to reduce tobacco consumption and exposure to second hand smoke should be facilitated by knowledge of the antecedents of adoption of these policies (Davis, 1995). Because legislators are in the position to enact many of the policies that are needed to support a comprehensive tobacco control strategy, they are an important focus for research on policy adoption.

Reports of legislators’ support for tobacco control measures include studies of US state legislators in Missouri (Davis, Kern, Perry, Brownson, & Harmon, 1989), Wisconsin (Gifford & Brindley, 1994), Kentucky (Hahn, Toumey, Rayens, & McCoy, 1999), and in North Carolina, Texas and Vermont (Goldstein et al., 1997). Dutch members of Parliament have also systematically been surveyed on tobacco policy issues (Hendriks, Garretsen, & van de Goor, 1997). In Canada, we have reported on support for specific tobacco control measures among provincial and territorial legislators (de Guia et al., 1998).

Other studies have explored the determinants of legislators’ support for tobacco control. Moore et al. used secondary data sources to investigate the influences on US House and Senate members’ votes on tobacco control initiatives (Moore, Wolfe, Lindes, & Douglas, 1994). Their multivariate analysis indicated that being a Republican, representing a tobacco-producing state, and receiving more money from tobacco political action committees were associated with opposition to tobacco control initiatives. Using the Theory of Planned Behavior as a model, Flynn et al. examined data collected from state legislators in North Carolina, Texas and Vermont (Flynn et al., 1998). They found that state, political party, attitudes, and perceived norms were significantly associated with legislators’ voting intentions on cigarette tax increases. Glantz and Begay explored influences on California State legislators’ records on tobacco control issues (Glantz and Begay, 1994). Those receiving more money from tobacco political action committees, Republicans, and those rated by members of the Sacramento press corps as having lower integrity and intelligence were more likely to oppose tobacco control initiatives.

Some of the US analyses have been criticized for not including measures of ideology (Wright, 1998), and there are other constructs that have not been explicitly considered in previous research on the topic. These include personal characteristics, tobacco experiences, knowledge about tobacco, and the saliency of various interest groups. The exclusion of variables related to a dependent variable in a multivariate analysis can result in biased regression coefficients (Maddala, 1992). Further, considering a broader range of variables and constructs could provide additional strategic information on ways in which health advocates can work toward creating favorable conditions for the enactment of tobacco control policies.

The present study of Canadian legislators did include measures of political ideology, as well as a broad range of variables, and is unique in that all federal, provincial, and territorial legislators were eligible for the study. We assessed the relationship between Canadian legislators’ support for tobacco control and five groups of potentially important variables, as outlined in a recently proposed conceptual framework for studying legislator decision making (Cohen et al., 2000): (1) political factors; (2) personal characteristics; (3) experiences with tobacco; (4) knowledge of tobacco-related health effects; and (5) interest group saliency.