DOES SPENDING MORE ON TOBACCO CONTROL PROGRAMS MAKE ECONOMIC SENSE? AN INCREMENTAL BENEFIT-COST ANALYSIS USING PANEL DATA

Abstract

I. INTRODUCTION State-level tobacco control programs gath-ered momentum in the United States during the early 1990s and have since been continuing with a varying degree of success in lowering smoking prevalence. Recently, Centers for Disease Control and Prevention (CDC) modified its 1999 guidelines with a new set of guidelines on recommended funding levels for comprehensive tobacco control programs for each state (Best Practices 2007). While states decide the annual funding levels (costs) for the control programs, assessment of the economic benefits of the programs, based on their past performance, is rarely undertaken. Unless benefits of the recommended funding levels are shown to clearly outweigh the costs of the programs, states may deviate from the CDC-recommended funding levels and divert the tobacco-related revenues away from these programs. This is, in fact, a recent trend in many states and is a concern often expressed in policy debates (American Lung Association 2009). This paper carries out a benefit-cost analysis of the ongoing, state-level tobacco control programs for the first time in light of the CDC's Best Practices (2007) guide-lines. The analysis is based on several variants of econometric approaches for modeling state-level tobacco demand, using panel data on tobacco sales from all 50 states for the period 1991-2007. There is a large body of research that shows that state-level tobacco control programs along with federally and privately funded initiatives have played a significant role in tobacco prevention and control during the last two decades (Family, Pechacek, and Chaloupka 2003; Farrelly et al. 2008; Hu, Sung, and Keeler 1995b; Tauras et al. 2005). Farrelly, Pechacek, and Chaloupka (2003) investigate the joint effects of the funding level and the duration of the control program on tobacco demand and report that the longer states invest in comprehensive tobacco prevention and control programs the greater and faster the reduction in smoking. A significant amount of additional annual revenue being generated as a result of the 1998 Tobacco Master Settlement Agreement (MSA) has also opened up further opportunities to help states continue to invest in these programs. Unfortunately, in recent years funding levels for such programs have shown a steady downturn. For example, total state spending on tobacco control at 2008 prices has dropped from a high of $930.8 million in 2002 to $694.3 million in 2007. Moreover, as states face record budget deficits due to recession, many are turning to cigarette taxes to raise revenue for spending in other programs. In 2009, only North Dakota funded tobacco prevention and control programs at the level recommended by the Best Practices (2007) guidelines. Forty states and the District of Columbia spent less than 50% of the Best Practices recommended level in 2009 despite receiving millions of dollars from tobacco tax revenues and the MSA funds. Funding in the state of Washington that has been able to bring down adult smoking rate by 30% and youth smoking rate by 50% since 2002, has been reduced by half in the fiscal years 2010 and 2011 (American Lung Association 2009). This trend may adversely affect the long-run sustainability of the tobacco control efforts and the CDC's long-term target to reduce adult smoking preva-lence rate down to 10% by 2025 (Best Practices 2007). A benefit-cost analysis of the comprehensive tobacco control programs in light of the CDC's Best Practices (2007) guidelines involves, as a first step, a systematic assessment of the performance of the existing programs over a long time horizon. Specifically, we ask the following questions in the present research: (1) Does state-level spending on the tobacco control programs during the past two decades show any independent impact on tobacco demand? (2) If so, does this impact grow over time? The second question is more important, as it takes a long time to quit smoking and, as such, an immediate impact may not be discernable (Chaloupka and Warner 2000; Gruber 2001). …