Lucky Strike?

Arkansas Gets Satisfaction from Rare State Investments in Smoking Prevention and Cessation

By John Engberg, Donna Farley, and Dana Schultz

John Engberg is a RAND senior economist with expertise in program and policy evaluation. Donna Farley is a RAND senior policy researcher with expertise in health policy analysis, program evaluation, and quality improvement theory and practice. Dana Schultz is a RAND policy analyst with expertise in social policy.

In November 2000, Arkansas voters passed the Tobacco Settlement Proceeds Act, a referendum to invest the state’s share of a massive multistate tobacco settlement in seven health-related programs. Together, these programs have been a force behind the state’s downward trend in smoking rates, especially for the most vulnerable populations.

Arkansas has been unique among the states by investing all its funds from the settlement in programs that focus on smoking prevention and cessation and other health-related endeavors.

Smoking rates among these populations — youth, young adults, and pregnant women of all ages — show conclusively that these groups are smoking less than would be expected absent the tobacco settlement programming. And in 2007, for the first time, smoking rates for all adults fell significantly below what they had been prior to the initiation of the programs. Arkansas has been unique among the states by investing all its funds from the settlement in programs that focus on smoking prevention and cessation and other health-related endeavors. These programs deserve continued support and time to fulfill their missions.

True to the Intent

In November 1998, nearly every U.S. state signed the historic Tobacco Master Settlement Agreement, which ended years of legal battles between the states and the major tobacco companies. Under the terms of the agreement, the tobacco companies are now paying the participating states more than $206 billion over a period of 25 years. All states except Florida, Minnesota, Mississippi, and Texas are participants in the agreement, as are the District of Columbia and several U.S. territories. The agreement also constrains tobacco company business practices, particularly the marketing of tobacco products to youth.

Figure 1 — Smoking Rates Among Adults in Arkansas Are Significantly Lower Than Before

Smoking Rates Among Adults in Arkansas Are Significantly Lower Than Before
SOURCE: Evaluation of the Arkansas Tobacco Settlement Program: Progress During 2006 and 2007.
NOTE: Rates are not adjusted for changes in survey sample demographic characteristics. The confidence intervals define a range within which the estimated values would fall 95 percent of the time for survey samples if the survey were repeated over and over again; in other words, there is 95 percent confidence that the true values lie within the range.

Figure 1 — Smoking Reductions Among Adults in Arkansas Are Not as Steep as Those in California

Smoking Reductions Among Adults in Arkansas Are Not as Steep as Those in California
SOURCE: Evaluation of the Arkansas Tobacco Settlement Program: Progress During 2006 and 2007.
NOTE: Rates are adjusted for changes in survey sample demographic characteristics.

Arkansas receives only 0.828 percent of all payments. But among the 46 participating states, Arkansas is unique in the commitment being made by both elected officials and the general public to invest the state’s share of the settlement in health-related programs. Just one of the state’s seven programs, the Tobacco Prevention and Cessation Program, is completely dedicated to smoking prevention and cessation; this program, however, receives about 30 percent of the state’s settlement funds. Four other state programs target the short-term health needs of disadvantaged Arkansas residents, while the remaining two are long-term investments in the public health and health research infrastructure.

The 2000 state referendum also created the Arkansas Tobacco Settlement Commission, giving it the responsibility for monitoring and evaluating the performance of the seven public health programs. The commission contracted with RAND to evaluate both the progress made by the programs in fulfilling their missions and the effects of the programs on smoking and other health-related outcomes.

Vantage Point

From the beginning of the decade to 2007, the smoking rate among all Arkansas adults fell from 26 percent to 22 percent, a rate that is significantly below what it had been prior to the initiation of the state efforts in 2002. The decline in smoking rates is equivalent to 16 percent fewer adult smokers. Although we cannot rule out that this decline is a continuation of a preexisting trend, it nonetheless represents a major milestone for the health of Arkansans.

As shown in Figure 1, the adult smoking rates for 2005–2007 are lower than the estimates for any of the preceding years. Careful examination reveals that the upper confidence limit for 2007 is less than the lower confidence limit for several of the years up until 2002, suggesting a statistically significant decline in adult smoking.

The 2007 rate, however, is only slightly below what would have been expected based on a trend that started before 2002, and the rate has not fallen as much as in other states with comprehensive smoking control programs, suggesting that there is still work to be done. Figure 2 includes a hypothetical trend line that indicates how far the rates would fall if Arkansas’ antismoking programs and policies were as successful as those in California, which has one of the most successful statewide tobacco control programs in the United States. Although the Arkansas rates from 2005 to 2007 were slightly lower than the pre-program trend, they were significantly higher than what would have been observed if Arkansas had experienced decreases similar to those in California.

In Arkansas, there is better news for women than for men. Figures 3 and 4 show the difference. Women are smoking significantly less than would be predicted by their baseline trend, while men are not. This difference is due, in part, to a downward trend for men prior to program initiation, whereas smoking was previously level for women. The change in the trend for women, however, also suggests that tobacco control programming is more effective for women than for men.

Figure 3 and 4 — Arkansas Women Are Smoking Less Than Would Be Predicted by Their Baseline Trend, While Arkansas Men Are Not

Arkansas Women Are Smoking Less Than Would Be Predicted by Their Baseline Trend, While Arkansas Men Are Not
SOURCE: Evaluation of the Arkansas Tobacco Settlement Program: Progress During 2006 and 2007.
NOTES: Rates are adjusted for changes in survey sample demographic characteristics.

Smoking prevalence among pregnant women in particular has been decreasing, albeit very slowly. Their smoking rate was slightly below the expected rates in the years from 2003 through 2007. The lowered rates of around 15 percent are slightly more than one percentage point below the trend and are statistically significant, but the drop from the expected rate has remained steady since 2003, suggesting that no additional gains have been made since shortly after the program initiation.

Several data sources confirm that young people in Arkansas are smoking less than would be expected based on trends prior to the tobacco settlement programs. In fact, substantial decreases in smoking have been found among all five of the following groups: middle school students, high school students, young adults (age 18 to 25), pregnant teenagers, and young pregnant women (age 20 to 29). In overseeing the implementation of legislation that requires states to have laws that prohibit the sale and distribution of tobacco products to minors, the federal government has also verified a dramatic improvement in Arkansas’ compliance with such laws.

Reductions in smoking among young people are particularly advantageous because the reductions will provide health dividends to the state for years to come. This conclusion is based on the assumption — which is supported by evidence in the research literature — that young people will not initiate or resume smoking when they are older.

Many Vulnerable Populations in Arkansas Are Smoking Less Than Would Be Predicted by Their Prior Trends

Population 2000 Rate
(percentage)
2005 Rate
(percentage)
2007 Rate
(percentage)
Percentage Decrease Between 2000 and 2007
Middle school students a 15.8 9.3 9.5 39.9
High school students a 35.8 26.3 20.4 43.0
High school students b 34.7 25.9 20.7 40.3
Pregnant teenagers (14–19) c 21.5 6.1 6.1 25.1
Young adults (18–25) d 31.2 28.9 26.1 16.3
Young pregnant women (20–29) 15.9 15.2 15.3 3.8
SOURCES: a Arkansas Department of Health and Human Services. b Arkansas Youth Risk Behavior Survey. c RAND calculations based on birth certificates, adjusted for changes in population demographics. d RAND calculations based on the Behavioral Risk Factor Surveillance System, adjusted for changes in population demographics.
NOTE: The estimated decreases are significant at a 95-percent confidence level for all populations.

As shown in the first two rows of the table, smoking rates for middle and high school students have dropped dramatically since 2000. Of note, though, the data for some youth groups did not exhibit further declines beyond 2005. These rates should be monitored to make sure that continued progress is made among these vulnerable populations.

The decreases reported in the table do not account for pre-program trends in smoking rates. For example, smoking was increasing for young adults before the initiation of tobacco settlement programming. If this trend had continued, the 2007 rate would have been higher than the 2000 rate. Therefore, the impact of the program is the difference between the estimated 2007 point on the trend line and the observed 2007 smoking rate, which is larger than the difference reported in the table. A similar story holds true for pregnant teenagers (see Figures 5 and 6).

Figure 5 and 6 — For Young Adults and Pregnant Teenagers in Arkansas, the Declines in Smoking Since 2001 Counteracted Their Previous Trends

For Young Adults and Pregnant Teenagers in Arkansas, the Declines in Smoking Since 2001 Counteracted Their Previous Trends
SOURCE: Evaluation of the Arkansas Tobacco Settlement Program: Progress During 2006 and 2007.
NOTES: Rates are adjusted for changes in survey sample demographic characteristics.

The story differs for high school students. Smoking appears to have been on the decline for these youth since at least 1997. The observed youth smoking rates for 2005 and 2007 fell even lower than the ongoing trend would predict. But the prior downward trend makes the percentage decrease reported in the table an overestimate of the program effect.

Beyond smoking rates, there have been reduced hospitalization rates for a variety of diseases that are associated with smoking and secondhand smoke. The strongest evidence is of significantly fewer hospitalizations for strokes and heart attacks since the start of the tobacco settlement funding. While the hospitalization rate for strokes has remained stable nationwide, it has fallen in Arkansas since 2001, even though it had been previously on the rise in the state. Similar rates for heart attacks in Arkansas and the nation have made comparable declines. Hospitalization trends for pneumonia and asthma also show weak evidence of improvement in Arkansas.

Beyond smoking rates, there have been reduced hospitalization rates for a variety of diseases that are associated with smoking and secondhand smoke.

Other measures of health will respond to decreases in smoking only after a long time. High rates of cancer and emphysema are the result of many years of high smoking rates and will show substantial declines only after smoking has been reduced for many years.

Arkansas House Speaker Robbie Wills is joined by Senator Tracy Steele and others during a February 2, 2009, news conference called by the supporters of increased tobacco tax legislation to fund state health programs.
AP IMAGES/DANNY JOHNSTON
Arkansas House Speaker Robbie Wills, D-Conway, left, is joined by Senator Tracy Steele, D-North Little Rock, center foreground, and others during a February 2, 2009, news conference called by the supporters of increased tobacco tax legislation to fund state health programs. The next day, four Republicans broke ranks to help supporters obtain the three-fourths majorities needed in both of the state’s legislative chambers to pass the tax increase. Two weeks later, the tobacco tax hike became law.

Merit Deserved

The seven Arkansas programs supported by the tobacco settlement funds have continued to expand their reach in improving the health of Arkansans. The Tobacco Prevention and Cessation Program includes community prevention programs, school education programs, enforcement of youth tobacco control laws, public awareness campaigns, minority initiatives, and other efforts, all of which follow the guidelines of the U.S. Centers for Disease Control and Prevention (CDC). The four state programs that target the short-term health-related needs of disadvantaged Arkansans are the Arkansas Aging Initiative, the Delta Area Health Education Center, the Minority Health Initiative, and Medicaid Expansion Programs. The two state programs that expand public health education and the state infrastructure for public health research are the Arkansas Biosciences Institute and the College of Public Health.

Despite the progress of these programs, there is still plenty of room for improvement. Although Arkansas has been a national leader in spending a considerable portion of its tobacco settlement money on smoking prevention, the state still spends only about half of the amount recommended by the CDC for prevention efforts. Increasing the funding to CDC-recommended levels would help Arkansas extend its gains in smoking reduction.

Another concern is that the work of the Tobacco Prevention and Cessation Program has been distributed unevenly throughout the state, with some areas receiving substantially more services than others. Our analysis provides weak evidence that counties with greater prevention efforts have experienced larger decreases in smoking rates, yet the variation in program spending appears unrelated to the need for smoking programs. The state should investigate how grants and technical assistance are being awarded to the counties. The state should also help communities that have high smoking rates prepare high-quality grant applications for the prevention and cessation programs funded through the tobacco settlement.

Medicaid Expansion Programs should intensify their outreach efforts and increase enrollment.

The Medicaid Expansion Programs have not been fully spending their funds to extend health services to low-income populations. The unspent funds mean not only that the programs may be underutilized but also that federal funds that would otherwise be available for these services are not being tapped. Thus, the Medicaid Expansion Programs should intensify their outreach efforts and increase enrollment to meet their spending targets.

Most important, we encourage Arkansas policymakers to reaffirm their commitment to dedicate the tobacco settlement funds to health-related programming. To do justice to the services, education, and research that these programs are now delivering, they should be given the continued support and time necessary to fulfill their mission of improving the health of Arkansas residents. We believe that additional progress can be made toward reaching this goal. square

Related Reading

Evaluation of the Arkansas Tobacco Settlement Program: Progress During 2004 and 2005, Donna O. Farley, John Engberg, Brian Carroll, Matthew Chinman, Elizabeth D’Amico, Sarah Hunter, Susan Lovejoy, Lisa R. Shugarman, Hao Yu, James P. Kahan, RAND/TR-397-ATSC, 2007, 378 pp., ISBN 978-0-8330-4046-6.
Evaluation of the Arkansas Tobacco Settlement Program: Progress During 2006 and 2007, Dana Schultz, Tamara Dubowitz, Susan Lovejoy, Shannah Tharp-Taylor, Hao Yu, John Engberg, RAND/TR-611-ATSC, 2008, 242 pp., ISBN 978-0-8330-4626-0.
Evaluation of the Arkansas Tobacco Settlement Program: Progress from Program Inception to 2004, Donna O. Farley, Matthew Chinman, Elizabeth D’Amico, David J. Dausey, John Engberg, Sarah Hunter, Lisa R. Shugarman, Melony E. Sorbero, RAND/TR-221-ATSC, 2004, 306 pp., ISBN 978-0-8330-3748-0.