Effects of Mandatory Seat Belt and other Traffic Safety Laws in the US
Supriya Sirsi
Source Dataset:
https://vincentarelbundock.github.io/Rdatasets/csv/AER/USSeatBelts.csv (I have also added an extra column/attribute that holds the year in which seatbelt laws first came into effect in various states. Source: https://en.wikipedia.org/wiki/Seat_belt_laws_in_the_United_States)
Source Description:
Balanced panel data for the years 1983–1997 from 50 US States, plus the District of Columbia, for assessing traffic fatalities and seat belt usage.
Question for Analysis:
This dataset contains information about various measures implemented by US States to decrease the average number of fatalities caused by road accidents, including enforcing traffic safety laws like: minimum drinking age of 21, maximum permissible blood alcohol level, wearing seatbelts, setting speed limit of 65mph, setting speed limit of 70 mph etc.
Has the enforcement of any of these laws, especially those about seatbelts, actually resulted in a decrease in fatalities? __________________________________________________________
Data Exploration & Wrangling:
Here is a look at a section of the first few rows of the data:
## X state year miles fatalities seatbelt speed65 speed70 drinkage alcohol
## 1 1 AK 1983 3358 0.04466945 NA no no yes no
## 2 2 AK 1984 3589 0.03733630 NA no no yes no
## 3 3 AK 1985 3840 0.03307291 NA no no yes no
## 4 4 AK 1986 4008 0.02519960 NA no no yes no
## 5 5 AK 1987 3900 0.01948718 NA no no yes no
## 6 6 AK 1988 3841 0.02525384 NA no no yes no
## income age enforce lawenforcedfrom
## 1 17973 28.23497 no 1991
## 2 18093 28.34354 no 1991
## 3 18925 28.37282 no 1991
## 4 18466 28.39665 no 1991
## 5 18021 28.45325 no 1991
## 6 18447 28.85142 no 1991Rename four data columns:
(“state” to “stateabbr”,“miles”to “mtmperyear”,“fatalities” to “fatalitiesperyear”, and “drinkage” to “drinkingagelaw”)
## X stateabbr year mtmperyear fatalitiesperyear seatbelt speed65 speed70
## 760 760 WY 1992 6217 0.01898022 0.66 yes no
## 761 761 WY 1993 6770 0.01772526 0.67 yes no
## 762 762 WY 1994 6689 0.02152788 0.70 yes no
## 763 763 WY 1995 7044 0.02413402 0.71 yes yes
## 764 764 WY 1996 7360 0.01942935 0.72 yes yes
## 765 765 WY 1997 7576 0.01808342 0.75 yes yes
## drinkingagelaw alcohol income age enforce lawenforcedfrom
## 760 yes no 18704 33.96467 secondary 1989
## 761 yes no 19535 34.21438 secondary 1989
## 762 yes no 19865 34.45578 secondary 1989
## 763 yes no 20685 34.76661 secondary 1989
## 764 yes no 21524 35.07435 secondary 1989
## 765 yes no 22596 35.38646 secondary 1989Add a column (absfreq) to the dataset, calculated as the product of two other columns:
(Absolute frequency of fatalities per year) = (million traffic miles per year) x (fatalities per million traffic miles per year)
mydata$absfreq <- mydata$mtmperyear * mydata$fatalitiesperyear
head(mydata)## X stateabbr year mtmperyear fatalitiesperyear seatbelt speed65 speed70
## 1 1 AK 1983 3358 0.04466945 NA no no
## 2 2 AK 1984 3589 0.03733630 NA no no
## 3 3 AK 1985 3840 0.03307291 NA no no
## 4 4 AK 1986 4008 0.02519960 NA no no
## 5 5 AK 1987 3900 0.01948718 NA no no
## 6 6 AK 1988 3841 0.02525384 NA no no
## drinkingagelaw alcohol income age enforce lawenforcedfrom absfreq
## 1 yes no 17973 28.23497 no 1991 150
## 2 yes no 18093 28.34354 no 1991 134
## 3 yes no 18925 28.37282 no 1991 127
## 4 yes no 18466 28.39665 no 1991 101
## 5 yes no 18021 28.45325 no 1991 76
## 6 yes no 18447 28.85142 no 1991 97Summarise the US Seat Belts data:
## X stateabbr year mtmperyear
## Min. : 1 Length:765 Min. :1983 Min. : 3099
## 1st Qu.:192 Class :character 1st Qu.:1986 1st Qu.: 11401
## Median :383 Mode :character Median :1990 Median : 30319
## Mean :383 Mean :1990 Mean : 41448
## 3rd Qu.:574 3rd Qu.:1994 3rd Qu.: 52312
## Max. :765 Max. :1997 Max. :285612
##
## fatalitiesperyear seatbelt speed65 speed70
## Min. :0.008327 Min. :0.0600 Length:765 Length:765
## 1st Qu.:0.017341 1st Qu.:0.4200 Class :character Class :character
## Median :0.021199 Median :0.5500 Mode :character Mode :character
## Mean :0.021490 Mean :0.5289
## 3rd Qu.:0.024774 3rd Qu.:0.6500
## Max. :0.045470 Max. :0.8700
## NA's :209
## drinkingagelaw alcohol income age
## Length:765 Length:765 Min. : 8372 Min. :28.23
## Class :character Class :character 1st Qu.:14266 1st Qu.:34.39
## Mode :character Mode :character Median :17624 Median :35.39
## Mean :17993 Mean :35.14
## 3rd Qu.:21080 3rd Qu.:36.13
## Max. :35863 Max. :39.17
##
## enforce lawenforcedfrom absfreq
## Length:765 Min. :1984 Min. : 44
## Class :character 1st Qu.:1986 1st Qu.: 255
## Mode :character Median :1987 Median : 640
## Mean :1988 Mean : 847
## 3rd Qu.:1991 3rd Qu.:1039
## Max. :1997 Max. :5504
## Display Aggregates: Average Seatbelt Usage Rates and Average Fatalities grouped by State:
(Absolute frequency of fatalities per year) = (million traffic miles per year) x (fatalities per million traffic miles per year)
Analysis:
This boxplot above indicates that US seatbelt usage rates kept increasing from 1983 to 1997, with very few outliers.
Analysis:
The boxplot above indicates that traffic miles covered in the US had been very slowly but steadily increasing from 1983 to 1997. The histogram shows most of the values distributed over a smaller range, thus reinforcing that the overall rate of change in annual traffic miles covered must be very small.
Scatter Plot & Line Plots:
Analysis:
The above Scatter Plot gives an indication that annual fatalities have been on a downward trend during the period from 1983 to 1997. Since most states started enforcing seatbelt laws after 1986, this can be interpreted as the increasing enforcement of seatbelt laws could have been a factor in the decrease of fatalities. However, since all states did not implement these laws at the same time, the correlation is not very clear.
Analysis:
On taking subsets of the data only for the states of California & Mississippi and analysing the above line plots, the correlation between enforcement of seatbelt laws and the downward trend in fatalities during the period from 1983 to 1997, is visible.
Analysis:
From the above line plot, the correlation between a maximum speed limit of 65 mph and the downward trend in fatalities during the period from 1983 to 1997, is visible.
Conclusion:
From the graphs seen above:
1. Annual fatalities per million traffic miles kept decreasing from 1983 to 1997
2. Seatbelt Usage Rate increased slowly from 1983 to 1997
3. Enforcement of seatbelt laws in California and Mississippi seems to have been a significant factor in the decrease in fatalities in both states
4. Enforcement of maximum speed limit of 65 mph in California seems to have been a significant factor in the decrease in fatalities in that state.
Thus we can conclude that enforcement of traffic safety laws like enforcement of seatbelt usage and setting specific maximum speed limits, was effective in decreasing the annual traffic fatality rate during the period from 1983 to 1997.