**CHAPTER 4 Modelling road safety in twenty-one countries**

**4.7 Denmark**

For Denmark, a model of the fatality rate was constructed using the all-occupant seat belt wearing rate, alcohol deaths per vkt and a dummy variable for the introduction of speed cameras in 1991 and increases in 1996. As well, dummy variables were included in the period prior to 1973 for higher speed limits and the fatalities peak in 1970-71. The results are shown in Table 4.47.

Table 4.47 Regression results for predicting the Danish fatality rate

*Regression Statistics*

Multiple R 0.9955961

R Square 0.9912116

Adjusted R Square 0.9901398 Standard Error 0.8860933

Observations 47

ANOVA

*df* *SS* *MS* *F* *Significance F*

Regression 5 3630.773992 726.15479 924.84786 5.33522E-41

Residual 41 32.19161529 0.7851613

Total 46 3662.965607

*Coefficients Standard Error* *t Stat* *P-value* *Lower 95%* *Upper 95%*

Intercept 17.009100 1.840962 9.239247 0.000000 13.291202 20.726997

Seat belts -0.125322 0.017714 -7.074786 0.000000 -0.161095 -0.089548

SpeedLimits 5.387422 0.752069 7.163467 0.000000 3.868587 6.906256

Alcohol 0.921971 0.249446 3.696080 0.000642 0.418205 1.425736

Dum7071 5.277569 0.743477 7.098497 0.000000 3.776087 6.779051

Speedcams -0.839246 0.386344 -2.172276 0.035671 -1.619484 -0.059008

Figure 4.145 shows the pattern of the fatality rate is fairly accurately predicted by the model.

Figure 4.145 Actual/predicted Danish fatality rate

Fatalities per billion safety-wgt vkt

2029 2027 2025 20212023 2019 2017 2015 2013 2011 2009 2007 2005 2003 2001 1999 1997 1995 1993 1991 1989 1987 1985 1983 1981 1979 1977 1975 1973 1971 1969 1967 1965

Actual Predicted

0 5 10 15 20 25 30 35 40

Figure 4.146 shows the components of the Danish fatality prediction/forecast. A major influence is the increase in seat belt wearing. From the mid-80s, alcohol control starts to have an effect, as do speed cameras from the early 90s.

Figure 4.146 Components of the Danish fatality rate prediction

2029 2027 2025 2023 20192021 2017 2015 20112013 2009 2007 2005 20012003 1999

Fatalities per billion safey-wgt vkt

Fatality rate

Impact of seatbelts + alcohol + speed cameras + dummies Impact of seatbelts + alcohol + speed cameras

Impact of seatbelts + alcohol Impact of seatbelts

0

Figure 4.147 shows that the modeling also produces a fairly accurate prediction of the level of fatalities over time.

Figure 4.147 Actual/predicted Danish road deaths

2029

Actual Predicted

0

Chapter 4 • Modelling road safety in twenty-one countries The safety weights applied in Denmark were cars, light commercial vehicles and other vehicles 1.0, motorcycles 12.0, buses 1.5, trucks 2.0 and mopeds 6.0. The injury weighted vkt calculation has motorcycles at 9.0 and mopeds at 4.5.

In Denmark, the injury rate (road injuries per billion injury-weighted vkt) moved in sync with the fatality rate. This is shown in Figure 4.148.

Figure 4.148 Danish injury and fatality rates

-30 70 170 270 370 470 570 670 770 870

2011 2009 2007 2005 2003 2001 1999 1997 1995 1993 1991 1989 1987 1985 1983 1981 1979 1977 1975 1973 1971 1969 1967 1965

Fatalities per billion vkt

Injury rate Fatality rate

Injuries per billion vkt

0 5 10 15 20 25 30 35 40

A model of the injury rate was constructed using the fatality rate and a dummy variable. The regression results are shown in Table 4.48.

Table 4.48 Regression results for predicting the Danish injury rate

*Regression Statistics*

Multiple R 0.9909971

R Square 0.9820752

Adjusted R Square 0.9812604 Standard Error 30.7202402

Observations 47

ANOVA

*df* *SS* *MS* *F* *Significance F*

Regression 2 2275052.419 1137526.2 1205.34729 3.76717E-39

Residual 44 41524.25908 943.73316

Total 46 2316576.678

*Coefficients Standard Error* *t Stat* *P-value* *Lower 95%* *Upper 95%*

Intercept -29.091063 9.536737 -3.050421 0.003863 -48.311094 -9.871032

FatalRate 24.580929 0.508629 48.327828 0.000000 23.555855 25.606003

Dum8595 -57.187969 10.297735 -5.553451 0.000002 -77.941690 -36.434248

Figure 4.149 shows the pattern of the injury rate is fairly accurately predicted by the model.

Figure 4.150 shows that the modeling also produces a fairly accurate prediction of the level of injuries over time.

Figure 4.149 Actual and predicted Danish injury rate

2029

Actual Predicted

Injuries per billion injury-wgt vkt

0

Figure 4.150 Actual and predicted Danish road injuries

2029

Actual Predicted

0

Chapter 4 • Modelling road safety in twenty-one countries

### 4.8 Finland

For Finland, a model of the fatality rate was constructed using the all-occupant seat belt wearing rate and measures of alcohol deaths per vkt and speed fines per vkt, as well as a dummy variable and a time trend prior to 1973. The results are shown in Table 4.49.

Table 4.49 Regression results for predicting the Finnish fatality rate

*Regression Statistics*

Multiple R 0.9940780

R Square 0.9881911

Adjusted R Square 0.9867853 Standard Error 1.4203631

Observations 48

ANOVA

*df* *SS* *MS* *F* *Significance F*

Regression 5 7090.548715 1418.109 702.9283 2.54858E-39

Residual 42 84.73211847 2.017431

Total 47 7175.280833

*Coefficients Standard Error* *t Stat* *P-value* *Lower 95%* *Upper 95%*

Intercept 17.408412 2.276334 7.647565 0.000000 12.814584 22.002240

Seat belts -0.220491 0.017597 -12.52978 0.000000 -0.256004 -0.184979

Alcdths/vkt 4.080406 0.484054 8.429656 0.000000 3.103546 5.057266

Timeles73 1.229333 0.184329 6.669223 0.000000 0.857341 1.601324

Dumles74 2.373011 1.483486 1.599618 0.117180 -0.620786 5.366807

SpFines/vkt -0.000238 0.000194 -1.228629 0.226051 -0.000630 0.000153

Figure 4.151 shows the pattern of the fatality rate is fairly accurately predicted by the model.

Figure 4.151 Actual/predicted Finnish fatality rate

Fatalities per billion safety-wgt vkt

2029 2027 2025 2023 2021 2019 2017 2015 2013 2011 2009 2007 2005 2003 2001 1999 1997 1995 1993 1991 1989 1987 1985 1983 1981 1979 1977 1975 1973 1971 1969 1967 1965

Actual Predicted

0 5 10 15 20 25 30 35 40 45 50

Figure 4.152 shows the components of the Finnish fatality prediction/forecast. The major influence is the increase in seat belt wearing. From the early 90s, alcohol control starts to have an effect, as does speed control from the early 2000s.

Figure 4.152 Components of the Finnish fatality rate prediction

2029

Fatalities per billion safey-wgt vkt

Fatality rate

Impact of seatbelts + alcohol + speed fines + dummies Impact of seatbelts + alcohol + speed fines

Impact of seatbelts + alcohol Impact of seatbelts

0

Figure 4.153 shows that the modeling also produces a fairly accurate prediction of the level of fatalities over time.

Figure 4.153 Actual/predicted Finnish road deaths

2029

Actual Predicted

0

Chapter 4 • Modelling road safety in twenty-one countries The safety weights applied in Finland were cars, light commercial vehicles and other vehicles 1.0, motorcycles 6.0, buses 1.5, trucks 2.0 and mopeds 3.0. The injury weighted vkt calculation has motorcycles at 10.0 and mopeds at 5.0.

In Finland, the injury rate (road injuries per billion injury-weighted vkt) moved roughly in sync with the fatality rate. This is shown in Figure 4.154.

Figure 4.154 Finnish injury and fatality rates

2011 2009 2007 2005 2003 2001 1999 1997 1995 1993 1991 1989 1987 1985 1983 1981 1979 1977 1975 1973 1971 1969 1967 1965 60 160 260 360 460 560 660

Fatalities per billion safety-wgt vkt

Injury rate Fatality rate

Injuries per billion injury-wgt vkt

0 5 10 15 20 25 30 35 40 45 50

A model of the injury rate was constructed using the fatality rate and a dummy variable from 2007 on. The regression results are shown in Table 4.50.

Table 4.50 Regression results for predicting the Finnish injury rate

*Regression Statistics*

Multiple R 0.992021

R Square 0.984106

Adjusted R Square 0.983384 Standard Error 18.136794

Observations 47

ANOVA

*df* *SS* *MS* *F* *Significance F*

Regression 2 896172.9221 448086.46 1362.1997 2.67255E-40

Residual 44 14473.50475 328.94328

Total 46 910646.4269

*Coefficients Standard Error* *t Stat* *P-value* *Lower 95%* *Upper 95%*

Intercept 68.325649 5.058516 13.507055 0.000000 58.130881 78.520418

FatalRate 11.227514 0.230919 48.620975 0.000000 10.762127 11.692901

Dum07on -9.348847 9.142921 -1.022523 0.312123 -27.775193 9.077499

Figure 4.155 shows the pattern of the injury rate is fairly accurately predicted by the model.

Figure 4.156 shows that the modeling also produces a fairly accurate prediction of the level of injuries over time.

Figure 4.155 Actual and predicted Finnish injury rate

2029 19791981 1977

Actual Predicted

Injuries per billion injury-wgt vkt

0

Figure 4.156 Actual and predicted Finnish road injuries

2029

Actual Predicted

0

Chapter 4 • Modelling road safety in twenty-one countries