IV. RESULTS

Observational Surveys of Belt Use
Forty seven states plus the District of Columbia and Puerto Rico reported observed statewide safety belt usage rates for June 2003. These rates have been verified or certified by NHTSA’s National Center for Statistics and Analysis. The graph in Figure 2a shows the rates for these states. Figure 2b shows 2002 and 2003 safety belt rates. The number of states that increased in belt use far exceeded the number that decreased (40 versus 6). Two states rates remained unchanged. Safety belt rates changed from a 4 percentage point decrease to a 13 percentage point increase. Of the 18 primary states and the District of Columbia and Puerto Rico who reported their 2003 usage rates, compared to 2002 rates, 15 had their rates increase, two stayed the same and three had lower rates. For the 28 secondary states for which 2002 and 2003 rates were known, 25 increased and three dropped compared to 2002 rates.

Figure 2a. Statewide Safety Belt Use Rates 2003

d

It should be noted that historically primary states, on average, have higher belt usage rates than do secondary states (85 versus 73 percent in 2003). Theoretically, it is harder for primary states to increase their belt use compared to secondary states, because they are starting at higher rates. For example, increasing belt usage 5 percentage points from 85 to 90 percent is more difficult than going from 60 to 65 percent.
* No 2002 rate available. New Hampshire, which is the only state without an adult safety belt law, did not report rates in 2003. However, under a contract jointly funded by NHTSA and the New Hampshire Highway Safety Agency, Preusser Research Group conducted an observational survey of safety belt use in accordance with the national uniform methodology in New Hampshire before and after the May 2003 Mobilization. The result of that post survey appears in the graph above.

Figure 2b. Observed Changes in the Statewide Safety Belt Use Rate by State

d

* No 2002 rate available. New Hampshire, which is the only state without an adult safety belt law, did not report rates in 2003. However, under a contract jointly funded by NHTSA and the New Hampshire Highway Safety Agency, Preusser Research Group conducted an observational survey of safety belt use in accordance with the national uniform methodology in New Hampshire before and after the May 2003 Mobilization. The result of that post survey appears in the graph above.

One measure of safety belt usage rate change that seeks to account for this is the measure of conversion rates. A conversion rate looks at the percentage of non-users who were “converted” to users. Using this measure, on average, primary states converted 8 percent of their non-users while secondary states converted 14 percent of their non-users. The average conversion rate was 13 percent. Conversion rates ranged from 48 percent (i.e. that percentage of non-users who now are users) to 36 percent (where the state had a very small percentage of non-users that grew by that percentage). Table 4 shows the conversion rates for primary and secondary law states.

Table 4. State Reported Safety Belt Use

State	2002	2003	Change	Conversion
PRIMARY LAW
Washington	92.6%	94.8%	2	30%
Hawaii	90.4%	91.8%	1	15%
California	91.1%	91.2%	0	1%
Oregon	88.2%	90.4%	2	19%
Maryland	85.8%	87.9%	2	15%
New Mexico	87.6%	87.2%	0	-3%
Puerto Rico	90.5%	87.1%	-3	-36%
Iowa	82.4%	86.2%	4	22%
North Carolina	84.1%	86.1%	2	13%
DC	84.6%	84.9%	0	2%
New York	82.8%	84.6%	2	10%
Georgia	77.0%	84.5%	8	33%
Texas	81.1%	84.3%	3	17%
Michigan	82.9%	83.9%	1	6%
Indiana	72.2%	82.3%	10	36%
New Jersey	80.5%	81.2%	1	4%
Connecticut	78.0%	78.0%	0	0%
Alabama	78.7%	77.4%	-1	-6%
Oklahoma	70.1%	76.7%	7	22%
Louisiana	68.6%	73.8%	5	17%
SECONDARY LAW
Arizona	73.7%	86.2%	13	48%
Utah	80.1%	85.2%	5	26%
Vermont	84.9%	82.4%	-3	-17%
Montana	78.4%	79.5%	1	5%
Pennsylvania	75.7%	79.0%	3	14%
Alaska	65.8%	78.9%	13	38%
Nevada	74.9%	78.7%	4	15%
Colorado	73.2%	77.7%	5	17%
Illinois	73.8%	76.2%	2	9%
Nebraska	69.7%	76.1%	6	21%
Delaware	71.2%	74.9%	4	13%
Ohio	70.3%	74.7%	4	15%
Virginia	70.4%	74.6%	4	14%
Rhode Island	70.8%	74.2%	3	12%
West Virginia	71.6%	73.6%	2	7%
Missouri	69.4%	72.9%	4	11%
South Carolina	66.3%	72.8%	6	19%
Florida	75.1%	72.6%	-3	-10%
Idaho	62.9%	71.7%	9	24%
South Dakota	64.0%	69.9%	6	16%
Wisconsin	66.1%	69.8%	4	11%
Tennessee	66.7%	68.5%	2	5%
Kentucky	62.0%	65.5%	4	9%
North Dakota	63.4%	63.7%	0	1%
Kansas	61.3%	63.6%	2	6%
Arkansas	63.7%	62.8%	-1	-2%
Mississippi	62.0%	62.2%	0	0%
Massachusetts	51.0%	61.7%	11	22%
NO ADULT BELT LAW
New Hampshire*		49.6%

* No 2002 rate available. New Hampshire, which is the only state without an adult safety belt law, did not report rates in 2003. However, under a contract jointly funded by NHTSA and the New Hampshire Highway Safety Agency, Preusser Research Group conducted an observational survey of safety belt use in accordance with the national uniform methodology in New Hampshire before and after the May, 2003 mobilization.

Of the states that reported both 2002 and 2003 full statewide belt use rates, 28 also reported a belt use rate measured just prior to the May 2003 Mobilization. Most of the states used a “mini survey” for the pre-measurement. Mini surveys are designed to be representative of the entire state by using a sample of observation sites from the larger statewide observational survey of belt use. One should be hesitant to draw strong conclusions regarding statewide belt use from a mini survey. Comparisons of the pre to post measures for this mobilization provide strong evidence of an increase in belt use immediately following the mobilization. Specifically, the fact that all 28 states showed an increase in belt use from pre to post suggests that even if we allow for any lack of the generalizability of the mini survey, there is strong evidence that the increase is real. Exploring the pattern of belt use across all three observational surveys shows that, since 2002, belt use generally dropped. But, after the re-treatment via the sTEP program, belt-use not only reached its prior level, but exceeded it. This pattern is similar to what has been described in previous literature as a “saw-blade pattern” (NHTSA, 1996). This implies that while belt use does indeed drop over time following a mobilization, the reintroduction of a mobilization restores the lost ground and then surpasses previous highs.

A repeated measures one-way ANOVA demonstrates that this pattern of an initial decrease followed by anincrease of greater magnitude is significant (F (2,54) = 32.07, p < .001) with all the simple effects being significantly different from each other according to follow-up paired t-tests (all p’s < .004).

Figure 3. Observed Change in Statewide Safety Belt Use Rate

d

2003 National Observation Survey (NOPUS)
The National Occupant Protection Use Survey (NOPUS) found that safety belt use reached 79 percent in 2003, a 4 percentage point increase over the 75 percent observed in 2002 (Glassbrenner, 2003). Approximately 17 percent of belt nonusers were converted to users, twice the rate seen in previous years. Use continues to vary in different parts of the country, with higher rates in states that can enforce their belt laws more stringently. States with primary safety belt laws reached 83 percent; those with secondary laws reached 75 percent. Belt use continues to lag behind for drivers of pickup trucks as compared to passenger cars , SUVs and vans.

Table 5.  National Occupant Protection Use Survey (NOPUS); Results 2000 - 2003

Source:  National Center for Statistics and Analysis, NHTSA, NOPUS 2000-2003

Pre/Post Telephone Survey
The national telephone surveys included a total of 2,446 respondents; 1,201 respondents in the pre-wave, 1,245 in the post-wave. Each survey sample had equal proportions by respondent gender, age, race and ethnicity, education level, type of vehicle driven most often, and (mean) number of adults in household. Comprehensive results are presented in Appendix B.

The figures that follow show results for selected questions from the baseline and post telephone surveys. Telephone survey results are explained categorically as they pertain to: 1) exposure to program message; 2) perception of law enforcement; 3) self reported usage, and; 4) awareness/opinion of safety belt law.

Exposure to Program Message
The telephone survey included questions asking respondents about recent exposure to safety belt messages and specifically messages concerning safety belt enforcement. Respondents who indicated exposure to messages were then asked to identify sources of those messages.

Respondents were asked if in the past 30 days they had seen or heard any messages encouraging safety belt use (Figures 4). Pre-survey (May 2003) results indicated that 73 percent of respondents had seen or heard messages encouraging safety belt use, suggesting that belt messages are generally commonplace. The post-survey (June 2003) measured an increase in the affirmative responses, to where 82 percent of respondents reported exposure to messages encouraging safety belt use.

Figure 4. Past 30 Days, Seen or Heard Messages Encouraging Safety Belt Use by State

d

* = Significant

A more specific survey question asked respondents whether or not they had seen or heard about the special enforcement efforts towards belts in the past 30 days. Pre-survey results indicated that few had just before the May Mobilization (16 percent). Figure 5 shows that changed over the course of the Mobilization. Just after the Mobilization, 40 percent of survey respondents indicated exposure to an enforcement message, an increase of 24 percentage points.

Figure 5. Past 30 Days, Seen or Heard of Special Police Efforts towards Belts by State

d

Respondents who indicated hearing or seeing a safety belt enforcement message in the past 30 days were also asked where they had seen or heard that message. Respondents indicated television as the most common source of information both before and after the mobilization (Figure 6). Post-survey results indicated that the proportion of respondents seeing a belt enforcement message on television nearly doubled over the course of the May Mobilization (28 to 50 percent). Post-survey results also indicated that respondent’s recall of radio messages nearly doubled (11 to 20 percent).

Figure 6. Where Saw or Heard of Special Enforcement Effort (subset of total respondents)

d

Figure 7 shows that large increases were measured in the percentage of respondents recalling the “Click It or Ticket” and the Click It or Ticket [State Name] slogans (26 and 24 percentage point increases). A small increase was measured in recall of the already well known Buckle Up [State Name] slogan (6 point increase). The alcohol awareness slogan, Friends Don’t Let Friends Drive Drunk remained the most recalled program slogan. Other slogans showed less recall and little or no change in recognition.

Figure 7. Recognized Principal Program Slogan by State

d

Perception of Law Enforcement
The telephone survey asked respondents a number of questions concerning perceived safety belt enforcement. Respondents were asked if they agree with the statement “that police in the community are writing more tickets now than a few months ago” (Figure 8). Pre-survey results indicated that about a third (34 percent) of respondents “agreed.” The proportion “agreeing” after the mobilization was larger (47 percent). The sharpest increase was measured among respondents saying that they “strongly agree” (9 point difference).

Figure 8. “Strongly Agree” Police in Community are Writing More Tickets Now by State

d

A related question asked respondents if they agree with the statement “police in my community generally won’t bother to write tickets for safety belt violations.” A sizeable proportion of pre-survey respondents (38 percent) expressed agreement before the mobilization. After the mobilization, that proportion measured somewhat lower (33 percent), a decrease of 5 points.

Figure 9. Police Generally Won’t Bother to Write Tickets for Safety Belt Violations

d

Respondents were asked how likely they think the chances are to receive a ticket for not wearing a safety belt. Post survey results found that the proportion of respondents who perceived a ticket “very likely” increased 6 percentage points; the proportion perceiving a ticket as “somewhat likely” increased 3 points.

Figure 10. Reports That Over Past Six Months Ticket for Non-Use was “Very Likely” by State

d

Self Reported Usage
Respondents were asked to report on their frequency of belt use. First, a series of questions asked what kind of belt system is in the respondent’s primary vehicle, then how often the belt system was used, and when was the last time they did not use it. Ninety-two percent of respondents indicated that their vehicle had the combination of lap and shoulder belt, seven percent indicated their vehicle had a shoulder belt only, and the remaining one percent just a lap belt. Among the respondents with a combination lap/should belt, 87 percent said they used the restraint system “all the time” and that did not change over time (Figure 11). That proportion was an obvious exaggeration that did not match with direct on-the-street measurements of belt use. A follow-up question asked respondents to indicate when was “the most recent time driving without a safety belt” (Figure 12). One-in-four respondents indicated within the previous year they had made at least one trip in which they did not use their safety belt.

Figure 11. Frequency of Lap/Shoulder Belt Use

d

Figure 12. Most Recent Time Driving Without a Safety Belt

d

Figure 13 shows the reported frequency of belt use adjusting for the most recent time driving without a safety belt. “All of the time” respondents slightly decreased over time (from 78 to 74 percent) and “all the time minus a month” respondents increased by as much (9 to 13 percent). Overall, reported frequency of belt use did not change during the National Mobilization.

Figure 13. Frequency of Safety Belt Use (Adjusted)

d

Opinion of Safety Belt Law
Research indicates that wearing a safety belt can reduce injury by nearly 50 percent. However, among telephone survey respondents, over one-third (35 percent) indicated agreement with the statement “safety belts are as likely to harm as help you” and nine percent strongly agreed that “putting on a safety belt makes them worry.” On the other hand, a vast majority (86 percent) strongly agreed that they would want a safety belt on if in a crash.

The proportion of respondents that indicated “stricter enforcement of adult safety belt laws” is very important measured higher after the May Mobilization (from 56 to 62 percent) adding proof that a majority of respondents believe safety belt enforcement is a necessary tool for improving safety belt use.
Additionally, a majority of respondents (69 percent) indicated that primary enforcement of safety belt laws should be allowed, and that remained relatively unchanged over the course of the National Mobilization. Roughly one-quarter (27 percent) of the respondents indicated the opinion that primary enforcement should not be allowed.

Comparison with Previous National Surveys
The May 2003 telephone surveys were preceded by two national surveys conducted in the same pre/post fashion; the first during May 2002 and the second during November 2002. The figures that follow show results for selected questions from all three pre/post telephone surveys. Results are explained categorically as they pertain to: 1) exposure to program message; 2) perception of law enforcement, and; 3) opinion of safety belt law.

Exposure to Program Message
Awareness of special enforcement efforts increased, from pre-to-post, each survey wave (Fig. 14). The largest increase was measured after the May 2003 Mobilization (24 percentage points), where 40 percent of those surveyed recalled hearing or seeing about a special enforcement efforts in the past 30 days. Comparatively, after the May 2002 Mobilization, 33 percent of respondents recalled hearing or seeing about special enforcement and only 30 percent after the November 2002 Mobilization. These results are not surprising, given that prior to May 2003, a nationwide advertisement purchase did not occur.

Another interesting finding is that pre-surveys measured awareness at nearly the same level each wave (14 to 16 percent), indicating that relatively low levels of publicity and information were focused on safety belt enforcement in the months between mobilizations.

Figure 14. Percent Who had Heard or Seen Special Enforcement Efforts

d

Click It or Ticket has become the most used safety belt enforcement campaign slogan. Currently, 32 states use it as their principle slogan. Motorists’ recall of the CIOT slogan has increased each National Mobilization, even though prior to 2003, there was no national CIOT campaign. As previously explained, the entire national media buy for May 2003 broadcast CIOT repeatedly to the nation’s motorists, and recognition increased most during the May 2003 Mobilization, 26 points and post-surveys measured the highest recall to date, 61 percent.

Figure 15. Percent Who had Heard or Seen CIOT Message in Past 30 Days

d

Perception of Law Enforcement
The May 2003, National Mobilization experienced the largest increase in proportion of respondents who agree “police are writing more tickets now than before.” The May 2003 increase equaled 13 points, compared to an 8 point increase in May 2002, and a 6 point increase in November 2002. At the end of the May 2003 National Mobilization, nearly half of respondents (47 percent) believed police were doing more safety belt enforcement.

Figure 16. Percent Who Agree Police are Writing More Tickets Now

d

Previous National Mobilizations experienced no appreciative change in the perceived likelihood of being ticketed for not wearing a safety belt. The six point increase in 2003 was the first measured statistically significant improvement.

Figure 17. Perceived Likelihood of Being Ticketed for Not Wearing a Safety Belt

d

Opinion of Safety Belt Law
Surveys measured only slight increases in the proportion of respondents that thought it very/fairly important for the state to enforce the safety belt law more strictly. On average, across all survey waves, 77 percent of respondents agreed that strict enforcement is very/fairly important. Respondents were most likely to say so after the May 2003 Mobilization (80 percent). All three pre/post surveys indicated that a high level of public support persists, even after the public was exposed to widespread, enforcement-centered, publicity and enforcement.

Figure 18. Percent Who think it is Important for State to Enforce Safety Belt Law More Strictly

d

Surveys have measured relatively little change in the percentage of respondents indicating that a primary enforcement safety belt law should be allowed. Nearly seven of every ten respondents have consistently indicated support for a primary enforcement law. That corroborates with other data that indicate majority support for strong safety belt laws exists and that support does not diminish even after the occurrence of widespread enforcement and publicity.

Figure 19. Percent Who Think Primary Enforcement Should be Allowed

d

Driver License Office Survey
Surveys (N=12,924) of licensed drivers were reported by 10 states so far. These states were Alabama, Arkansas, Connecticut, Illinois, Kentucky, Maryland, Rhode Island, South Carolina, Utah, and West Virginia. While such surveys were conducted in many other states, these particular states are a small sample based on those who had submitted their results to the evaluator. Driver survey data were collected in two waves. First wave surveys were collected during the week preceding any program publicity. Results of the first wave surveys are considered baseline. Second wave surveys were collected during the week just after publicity and enforcement ended. Results from the second wave surveys are considered post.

Driver surveys included questions about self reported safety belt use, exposure to messages concerning safety belt enforcement, perceived risk of a ticket for not using a safety belt, and program slogan recognition (see example of questionnaires in Appendix A). Survey questions were ordered and worded similarly among the 10 states, with only minor modifications to names of states, type of law, and names of law enforcement agencies as they appear on the questionnaire. Order and wording was not changed so that pre-to-post differences could be observed.

General Survey results are presented in Table 6.

The driver survey included questions about exposure to messages concerning safety belts and safety belt enforcement. Questions also asked respondents what sources of information they encountered. Survey results indicated that the public became aware of new messages focusing on safety belts. Respondents were asked if they had recently seen or heard a safety belt message. Over the course of the May Mobilization, awareness of new messages increased 23 percentage points (from 61 to 84 percent). Baseline surveys indicated that nearly half (49 percent) of respondents recognized the Click It or Ticket slogan. That proportion increased by 20 percentage points, to 69 percent, over the course of the mobilization.

Respondents were asked the more specific question have you seen or heard about police enforcement towards safety belt use. The baseline survey found about a third of the survey respondents had. That proportion nearly doubled in the post survey (58 percent), indicating that enforcement centered messages were widely received.

Respondents were also asked to identify sources of safety belt information they had encountered. Respondents identified television as the prime conduit before and after CIOT. Respondents also identified radio as a prevalent source, but not to the extent of television. Respondents identified the newspaper, but to an even lesser extent than radio. Between the time of the baseline and post surveys, television exposure increased most. That is not surprising given the magnitude of the effort to place paid advertisements on television broadcasts.

The driver survey also included questions regarding perceived enforcement and if a personal experience occurred involving safety belt enforcement. Respondents were asked to indicate how strictly they think their State Police and local police enforce the safety belt law. Respondents perceived State Police as more strict when it comes to safety belt law enforcement. A larger proportion of respondents indicated State Police as “very strict,” compared to local police, and that was true in every state and evident in both survey waves. Measures of perceived police strictness did not change over the course of the May 2003 Mobilization, at least not in the Driver Licensing Office Survey. Although, as presented above, telephone survey results did show increases among various measures of perceived enforcement, including a belief that police are writing more tickets and increased risk of a ticket.

The driver survey also asked respondents questions regarding personal encounters with police enforcement of safety belt laws. The first question asked respondents if they had experienced enforcement focused on safety belts in the past month. The second question asked respondents if they have ever received a ticket for not using a safety belt. A small but noticeable increase occurred in the proportion of respondents who reported experience with safety belt enforcement (13 to 17 percent). In regard to having received a ticket for not using a safety belt, a relatively low number of respondents indicated they had (13 percent) and that did not change over time.

Several significant correlations resulted from comparing pre to post changes in the DMV results with other program results. These analyses compared the changes in the responses from those respondents who filled out questionnaires prior to the mobilization to those who filled out questionnaires afterwards with other self reported responses and with other program results (e.g. media dollars spent per person, number of belt tickets written per person). These correlations establish a relationship between both degree of enforcement and media and self-reported and measured behaviors.

Change in pre to post self reported belt use was significantly correlated (r = .633, p = .049) with whether or not the respondent had heard a belt message. That is, the greater the change in self reported belt use the greater the change from pre to post campaign in the percentage of people who reported hearing a belt message. This result can be interpreted to suggest that media penetration does affect self reported belt use. Further investigation suggests that the effect is more influenced by TV than radio, although affected by both. The change in percentage of people who reported hearing the message on TV was significantly correlated (r = .891, p = .001) with change in self reported belt change as was the change in having heard a message on the radio (r = .716, p = .02). The change in the percentage of people having heard the CIOT message specifically was related to change in self reported belt use (r = .652, p = .041).

Result also show that pre to post change in judgments of local police strictness is correlated with the number of belt tickets written per person in the state (r = .612, p = .045). That is, states that reported writing more belt tickets per person tended to have a larger pre to post change in the percentage of respondents saying that the local police were “very strict.” One interpretation of this finding is that people seem to be aware of the higher ticket writing by police (assumedly by local police). Theoretically then, increased belt ticket writing may influence attitudes (e.g. police strictness) that are related to belt use. The stricter one perceives the police to be, the more likely that person may be to properly engage their safety belt.

Another finding supports the relationship between heightened enforcement and actual safety belt use. In those states where there was a greater change in the number of people who reported having been through a police check point, there was also a higher conversion rate (r = .662, p = .037). That is, relative increases in the self reported experience with police enforcement efforts directed at safety belt use are related to the percentage of non-users who subsequently became safety belt users. Again, this suggests a link between perceptions of enforcement and actual belt use.

Lastly, a near significant (p = .057) correlation exists between respondents’ change in perceived chances of being ticketed and the actual observed change in 2002 to 2003 belt use. That is, there is some evidence that the extent to which people are made to feel a change in their likelihood of being ticketed for non-compliance with a safety belt law, affects their actual observed change in safety belt use. States that had bigger changes in belt use from 2002 to 2003 tended to be those states where there was also a change from pre to post campaign with regards to how likely respondents felt they were to get a ticket. This finding has some merit despite the weak correlation. First, it should be noted that the small number of states’ data entered into this analysis makes it more difficult to show a significant relationship. Additionally, prior research has also shown such a relationship between beliefs regarding likelihood of being ticketed and self-reported belt use (Chaudhary, Solomon & Cosgrove, 2003). Still, one would expect that this relationship should strengthen with the addition of additional sates’ data.

The correlations presented above add evidence that perceptions of heightened enforcement, and exposure to media indicating such enforcement, are related to self reported and observed changes in belt use behavior.

Table 6. Pre/Post Driver License Office Survey Results (10 States*)

 

Back

Top

Next