TECHNICAL SUMMARY

Connecticut’s number of alcohol-related fatalities remained essentially constant from 1992 to 2002.  In addition, the percentage of alcohol-related fatalities (50.6%) out of all crash fatalities (312) during 2001 was higher than the national percentage (41.4%) and was also higher than other New England States (45.9%).  Connecticut and the National Highway Traffic Safety Administration (NHTSA) focused their efforts to reduce drinking and driving, with media efforts focusing particularly on men 21 to 34 because of their overrepresentation in alcohol-related fatal crashes.

Background

Federal Highway Administration funds were transferred by the Connecticut Department of Transportation and committed to the Division of Highway Safety for a statewide impaired-driving publicity and enforcement campaign.  The campaign represents the first time that Connecticut has expended such a substantial amount of money for both media and enforcement in its efforts to reduce impaired driving and, ultimately, alcohol-related injury and fatal crashes.

The campaign consisted of three components: (1) media with an enforcement message, (2) enhanced periods of enforcement surrounding the July 4th and winter holidays focusing on the use of sobriety checkpoints, and (3) sustained enforcement between holiday enforcement periods. 

Connecticut’s efforts followed the NHTSA  impaired-driving high-visibility enforcement model and was a test of NHTSA’s model, which includes (1) paid and earned media in support of (2) statewide high-intensity enforcement crackdowns, and (3) planned, sustained enforcement efforts between crackdowns.  The enforcement component involved commitment to sustained DWI enforcement throughout the year and two enhanced enforcement crackdowns covering the Independence Day and Thanksgiving/Christmas holiday periods.  NHTSA’s model focuses on crackdowns that cover 85 percent of the States’ populations and involve high-visibility sobriety checkpoints and/or saturation patrols during three weekends (16 days) of these holiday periods.  It also encompasses public awareness efforts involving State earned media, State funded paid media, and NHTSA paid media.

Media 

Congress appropriated $11 million for paid media.  Of the total, $5.5 million was spent to purchase air time on national TV, and the remainder was used to develop the ad used and for paid media in the 13 strategic evaluation States chosen because of their high alcohol-related fatality trend or high number of alcohol-related fatalities.  Although Connecticut was not a strategic evaluation State, it followed NHTSA’s impaired-driving high-visibility enforcement model including using NHTSA’s ads.  The TV ad was targeted especially to young men 18 to 34 and was placed on TV programs often viewed by this group.  The ad ran during the June 20 – July 13, 2003, period.

Connecticut’s media campaign spanned 11 months beginning in March 2003 at a cost of $2,199,533.  The enforcement grant funds totaled $1,582,568.  Connecticut spent a total of $3,782,101 on its 2003 impaired-driving publicity and enforcement program.  For each of the State’s campaign periods, a four- to six-week paid and earned media (media coverage a program “earns” whenever it makes the news in print or broadcast) campaign with a strong enforcement message was implemented.  The focus of the media campaign was generally the same as that of NHTSA, primarily young men 21 to 34 years old due to their high rate of involvement in alcohol-related crashes.  The media campaigns focused on two holiday periods during 2003 and were also designed to create the perception of sustained enforcement between these two holiday periods.  Some media components continued throughout the campaign.

Sobriety Checkpoints

Sobriety checkpoints constituted the main focus of the enforcement effort. In total, the State funded 24 sobriety checkpoints during the July 4th holiday period.  Fifteen towns held at least one sobriety checkpoint during the July 4th holiday period.  Under the expanded grants, a total of 18 towns and the State Police conducted a total of 89 sobriety checkpoints.  Some of the sobriety checkpoints conducted under the expanded grants were conducted during the two holiday periods, but there is no information on exactly how many of them were held during the holiday enforcement periods.  Twenty-eight police agencies and the State Police conducted a total of 51 sobriety checkpoints during the winter holiday period, more than twice as many as conducted during the July 4th holiday period.  Twenty additional sobriety checkpoints included an evaluation research component after the normal portion of the sobriety checkpoint in which researchers collected direct observations of drinking and driving by obtaining breath alcohol concentration (BAC) information from a random sample of drivers passing through the sobriety checkpoint.

In summary, the program began during the July 4th crackdown period, was sustained during the next few months, and then peaked during the Thanksgiving to Christmas holiday period.

Program Evaluation

The evaluation included statewide telephone surveys, direct observations of drinking and driving at sobriety checkpoints before and after each holiday enhanced enforcement effort, DWI arrest data, and alcohol-related fatality data.

Statewide Telephone Survey

Connecticut conducted telephone surveys before and after each holiday period.  The sampling plan was designed to ensure a representative sample of Connecticut drivers and used a random digit dialing approach to interview a total of 2,430 drivers over the course of the campaign.

Direct Observations of Driver BACs

Driver BACs were collected at sobriety checkpoints in nine towns before and after the holiday enforcement periods as a direct measure of the effect of enforcement and publicity on drinking and driving.  A geographically diverse set of sites in the State, focusing on towns with particularly high alcohol-related fatal and injury crash rates, were chosen.  The research team obtained voluntary, “blind,” anonymous BACs from randomly selected drivers on handheld breath-testing devices.  Generally, about 92  percent or more drivers agreed to the BAC test.   The team collected 1,249 BAC samples from drivers before and after the July 4th holiday enforcement period and 2,115 BAC samples from drivers before and after the winter holiday enforcement period. 

Alcohol-Related Fatalities

Alcohol-related fatality data was taken from NHTSA’s Fatality Analysis Reporting System (FARS) for 2000 through the preliminary 2004 data.  The overall alcohol-related fatality trend for the State and the alcohol-related fatality trend for fatalities involving men 21 to 34 years old were analyzed using the Autoregressive Integrated Moving Average (ARIMA) technique.

Results and Discussion

Exposure to Enforcement Message

There was a statistically significant increase in the number of telephone survey respondents who reported hearing or seeing something about alcohol-impaired driving in Connecticut after each holiday period compared to responses by those asked before each holiday period.  There was an 8.6-percentage-point increase after the July 4th holiday period from 55.6 percent to 64.2 percent (χ²(1)=9.42, p<.01) and a similar 8.5-percentage-point increase from 53.3 percent to 61.8 percent after the winter 2003 holiday period (χ²(1)=8.97, p<.01).

Perceptions of Enforcement

The number of telephone survey respondents indicating that State Police “very strictly” enforce the drinking and driving laws increased significantly compared to the other options combined after both the July 4th holiday enforcement period from 39.2 percent to 48.8 percent (χ²(1)=10.21, p=.001) and the winter 2003 holiday enforcement period from 38.3 percent to 48.1 percent (χ²(1)=10.29, p=.001).  Respondents indicated that local police “very strictly” enforce the drinking and driving laws significantly more often compared to the other options combined after both the July 4th (χ² (1)=3.94, p<.05) and winter (χ²(1)=11.54, p=.001) holiday periods compared to the period before each holiday.

When all telephone survey respondents were asked about the chances of being stopped if a driver had been drinking, there was a significant increase after both the July 4th holiday enforcement period from 53.8 percent to 62.5 percent (χ²(1)=7.683, p<.01) and winter 2003 holiday period from  53.5 percent to 64.6 percent (χ²(1)=13.548, p<.001) in the number of respondents indicating that they thought a driver who had been drinking was more likely to be stopped by police compared to all the other options combined.

Roadside Survey Driver BACs

More than three times as many sobriety checkpoints were held during the winter holiday period and were preceded by many more sobriety checkpoints during the sustained enforcement period that extended from summer to fall, so the cumulative effect of the campaign on driver BACs requires comparing the pre-July 4th data to the post-winter holiday period data.  There are differences in the locations and times of night when comparing the pre-July 4th holiday period BACs to the post-winter holiday period BACs, but in looking at the overall impact of the program from its inception to its peak, the comparison indicated that there was a significant decrease in the proportion of drivers with a positive BAC from the pre-July 4th period to the post-winter holiday enforcement period (χ²(2)=7.015, p<.01).  The pre-July 4th holiday sobriety checkpoints and post-winter holiday sobriety checkpoints were both conducted during colder, non-holiday times of the year.  The breath test refusal rates for both periods were very similar.

The proportion of male drivers with a positive BAC at sobriety checkpoints decreased significantly from 17.8 percent for the pre-July 4th to 10.6 percent for the post-winter holiday period (χ²(1)=10.42, p<.01).  Female drivers had a positive BAC about the same percentage of time from the pre-July 4th holiday period (9%) to the post-winter holiday period (9.3%).  There were generally more male drivers going through the sobriety checkpoint locations during all survey periods and they generally were more likely to have a positive BAC, but the proportion of men drinking and driving decreased almost to the same rate as the women by the post-winter holiday period.

Figure 2.*  Connecticut Predicted Alcohol-Related Fatalities 2000-2004 After Contiguous County Data and Modeling Applied to Remove Noise and the Effects of Regionwide Efforts to Combat Drinking and Driving as Well as Seasonal and Economic Variations

Alcohol-Related Fatalities

ARIMA analyses indicated that there was a significant decrease in the alcohol-related fatality trend for the 18-month period following the beginning of the impaired-driving high-visibility enforcement campaign.  The intervention period trend from July 2003 through December 2004 was evaluated in comparison to the trend from 2000 through the first six months of 2003 (p=.042).  When alcohol-related fatalities from contiguous counties in New York, Rhode Island, and Massachusetts were used as a covariate, the significant decrease in the estimated monthly average number of alcohol-related fatalities in Connecticut during the second half of 2003 through December 2004 was stronger (p=.01).  Figure 2 shows graphically the significant reduction in the estimated alcohol-related fatality trend in Connecticut after extraneous influences such as noise, bordering county drinking and driving trends, and seasonal and economic variation was removed from the series by using contiguous county monthly alcohol-related fatality totals.  The estimated reduction in the number of alcohol related fatalities determined by the ARIMA analysis was about 2.6 lives each month for the 18 months following June 2003.  Thus, if there were no campaign, there would have been an estimated 47 additional alcohol-related fatalities.

Summary and Conclusions

Surveys indicated that the paid media had reached a substantial number of Connecticut drivers.  Law enforcement agencies conducted a large number of sobriety checkpoints throughout the campaign period, with a particularly large number of sobriety checkpoints during the winter 2003 holiday period.  The combined publicity and highly visible enforcement campaign achieved its ultimate goal: significantly reducing the alcohol-related fatality trend, saving an estimated 2.6 lives a month, or a total of 47 lives, for the 18 months following the start of the first campaign crackdown in July 2003.  The campaign also resulted in a significant decrease in the alcohol-related fatality trend for men 21 to 34 years old, saving an estimated 1.6 lives each month for a total of 29 lives for the 18-month period following the first crackdown.