Evaluation of Transport Accident Commission Road Safety Television Advertising

Monash University Accident Research Centre - Report #52 - 1993

Authors: M. Cameron, N. Haworth, J. Oxley, S. Newstead & T. Le

Abstract:

The Transport Accident Commission (TAC) has made a major investment in road safety advertising on television in Victoria since 1989. This high profile and intense advertising has captured the Victorian public's attention as representing a substantial commitment by the TAC to improving road safety in Victoria.

Monash University Accident Research Centre (MUARC) has conducted an evaluation of the road safety benefits of TAC's television advertising campaigns. Part 1 of this report describes the TAC campaigns and summarises an attempt to estimate the role of the publicity campaigns in Victoria's road safety performance. Part 2 builds on recent MUARC evaluations of two major road safety programs which combined TAC advertising campaigns with increased enforcement efforts by the Victoria Police. An economic analysis of the effects on crashes of the TAC advertising supporting enforcement is described. Part 3 describes an attempt to evaluate the Concentrate or Kill campaign, which differs from the speed and drink-driving advertising campaigns in that it was not designed to support enforcement. The overall findings from this multi-faceted and relatively complex evaluation study are synthesised in Part 4.

The research indicates clear links between levels of TAC publicity supporting the speed and alcohol enforcement programs and reductions in casualty crashes when other major factors are held constant. For levels of advertising at the point of diminishing returns, the estimated benefits in terms of reduced TAC payments were respectively 3.9 and 7.9 times the costs of advertising supporting the speed and alcohol enforcement programs.

The road safety effects of TAC publicity with themes not related to enforcement (ie. concentration) is less clear. The Concentrate or Kill advertisements appear to raise awareness of the issue, but there is no conclusive evidence at this stage that they have reduced the crash involvement of the specific target group of the advertisements, namely young drivers on country roads.

Executive Summary

GENERAL CONCLUSIONS (Part 4)

OVERVIEW AND GENERAL EFFECTS (Part 1)

The monthly levels of TAC road safety advertising on television varied considerably between December 1989 and December 1992. This variation represented an opportunity to examine the link between the TAC advertising and the monthly road safety performance in Victoria.

The Victorian serious casualty rate per 100 million kilometres travelled displayed a substantial downward trend from mid-1989, and at a greater rate of decrease than in NSW. The ratio of the Victorian rate to the NSW rate was considered to represent the net effect of the unique Victorian programs (including publicity campaigns) relative to those in NSW.

An attempt was made to develop a method to estimate the contributions of the unique major road safety programs in each State to the reduction in the corresponding serious casualty rate. The aim was to remove the contributions of the unique road safety programs from the ratio of the serious casualty rates calculated each month during 1989-92, so that any link between the residual changes in the Victorian serious casualty rate and the monthly levels of TAC television advertising could be seen. It turned out that it was not possible to develop a satisfactory method because of the absence of some necessary data.

ECONOMIC ANALYSIS OF THE EFFECTS ON CRASHES OF THE TAC ADVERTISING SUPPORTING ENFORCEMENT (Part 2)

A relationship between the monthly levels of TAC television advertising (all themes) and the reductions in "low alcohol hour" (mainly daytime) casualty crashes in Melbourne during 1990-91 had been found as part of a previous evaluation of the Speed Camera Program. This provided the basis of an economic assessment of the investment in publicity at different levels. The present study found a stronger relationship between the same type of crash reductions and the monthly levels of publicity with "speeding" or "concentration" themes. The level of publicity was measured using TARPs (Target Audience Rating Points - a measure of reach of the designated target audience).

Another previous study had evaluated the effects during 1990-91 of the Random Breath Testing Program using "booze buses". This study was extended to fit relationships between monthly levels of TAC television advertising and casualty crashes during the high alcohol hours of the week in Melbourne and country Victoria to the end of 1992. The relationships took into account monthly variations in other explanatory variables found to have significant effects (ie. unemployment rates, number of random breath tests, alcohol sales, and seasonal variation and trend). The most reliable relationships were those for TAC "drink-driving" publicity measured by "Adstock" (a function of TARPs) which represents cumulative awareness in the month due to current and previous advertising levels.

An economic analysis compared the value of the estimated crash savings per month at each level of monthly TARPs with the total costs of investing in those TARPs. These costs were estimated including both the fixed costs of developing each television advertisement (plus the costs of supporting media), and the costs of television media placement, measured by the average cost per TARP. The estimated reductions in casualty crashes were valued from two viewpoints: (a) reductions in TAC payments to injury claimants, and (b) reductions in the social costs of the crashes (the social costs include the value of future productivity forgone, pain and suffering, and damage to vehicles and property).

(a) Return on investment to reduce TAC payments to injury claimants

An investment of 540 TARPs per month on average in a combination of "speeding" and "concentration" publicity was estimated to be economically justified before diminishing returns occurred. This level of average investment could lie in the range from 310 to 750 TARPs per month (with 68% confidence). An investment of 540 TARPs per month (plus necessary fixed costs to develop the television advertisements used, as well as for supporting publicity in other media) is estimated to return benefits of reduced TAC payments at a level 3.9 times the investment costs. It was also estimated to result in a 9% reduction in monthly low alcohol hour casualty crashes in Melbourne.

An investment of 800 TARPs per month on average in "drink-driving" publicity was conservatively estimated to be economically justified before diminishing returns occurred. This level of average investment could lie in the range from 660 to 930 TARPs per month (with 68% confidence). An investment at an average level of 800 TARPs per month (plus necessary fixed costs) is estimated to return benefits of reduced TAC payments at a level 7.9 times the investment costs. It was also estimated to result in an 18.5% reduction in monthly high alcohol hour casualty crashes in Victoria.

In both the above cases, even higher benefit/cost ratios would be realised if lower levels of TARPs per month were used, but the reduction in monthly casualty crashes would be lower.

(b) Reduced social costs of casualty crashes

An investment of 1080 TARPs per month on average in a combination of "speeding" and "concentration" publicity was estimated to be economically justified before diminishing returns occurred. This level of average investment could lie in the range from 610 to 1490 TARPs per month (with 68% confidence; the latter figure also being a less reliable estimate). An investment of 1080 TARPs per month (plus necessary fixed costs) is estimated to return benefits of reduced social costs of crashes at a level 5.4 times the investment costs.

It was not possible to reliably estimate the point of diminishing returns for the investment in "drink-driving" publicity if the casualty crash savings were valued by their average social costs, except that it would be considerably higher than the corresponding level based on TAC payments.

Assumptions and qualifications

These results were based on a number of assumptions and warrant a number of qualifications which are described in the report. In particular it should be noted that a high level of television advertising at the beginning of a new campaign (perhaps at a higher level per week than indicated by the estimated point of diminishing returns) may be necessary. The TAC launches of each new advertisement had this characteristic, which may be an essential part of establishing in real life the relationships between publicity levels and casualty crash reductions observed in this study.

EFFECTS OF CONCENTRATE OR KILL - AN ADVERTISING CAMPAIGN NOT DIRECTLY LINKED TO ENFORCEMENT (Part 3)

Concentrate or Kill differs from the speed and drink-driving advertising campaigns in that it did not have an associated enforcement campaign. Thus an evaluation of this campaign has the potential to give a clearer assessment of the benefits of a road safety advertising campaign not accompanied by an enforcement effort. The campaign included the advertisements Country Kids and Morgue which promoted the need to concentrate while driving.

It was initially envisaged that statistical analysis would be used to link numbers of crashes with the advertising exposure levels to measure the extent to which the Concentrate or Kill advertising led to a reduction in the number (or severity) of the group of crashes that were targeted. However, for both Country Kids and Morgue, it did not appear possible to identify valid target groups of drivers or crashes which were large enough to make this link satisfactorily. The analysis was restricted to a comparison of crash rates of target and control groups before and after the introduction of each of the advertisements.

For both Country Kids and Morgue, there was no reliable evidence of reductions in the risk of serious casualty crashes involving the target groups of the advertisements after the commencement of the advertising campaigns. These findings could have resulted from the crash numbers being too small to show statistically significant reductions or from the effect of the advertisements being relatively small. A number of characteristics of the Concentrate or Kill campaign which may have affected its benefit/cost ratio are discussed in Part 3.

Sponsor: Transport Accident Commission