May 28, 2002 |
2002-R-0516 | |
COMPARISON OF . 08 AND . 10 BAC LIMITS AND FATALITY RATES | ||
By: Kristina D. Arsenault, Research Fellow |
You asked for an update of three previous OLR Reports (98-R-0465, 99-R-0305, and 2001-R-0370) comparing the motor vehicle fatality rates in states whose driving under the influence (DUI) laws use a blood alcohol count (BAC) of . 08 with states with a . 10 limit. More specifically, you asked how Connecticut compares to both groups. BAC refers to the specific alcohol concentration in a driver's blood.
SUMMARY
The latest motor vehicle fatality rates are available only through calendar year 2000. Based on those statistics, regression analysis shows that there is no statistical difference between the average fatality rate in states that use a BAC of . 08 and states with a . 10 limit. The average 2000 fatality rate in the 20 states with a . 08 BAC limit was somewhat lower than the average in states with a . 10 limit (1. 45 vs. 1. 68 fatalities per 100 million vehicle miles traveled), but the difference cannot be measured as statistically significant. This finding was similar in past reports.
As of 2000, 30 states, including Connecticut have a . 10 BAC limit and 20 states have a . 08 BAC limit. Most of the states enforce a "per se" DUI law maintain that is illegal to operate a vehicle if you exceed the requisite BAC legal limit. Evidence of a person's BAC at or above the prescribed limit is illegal. In other words, there need be no finding of impairment.
In states without an "illegal per se" law, your BAC is just one of the factors that determines whether or not you're a drunk driver. Some examples of those other factors would include slurred speech and unsteady gait. Every state except Massachusetts and South Carolina has an "illegal per se" law. Massachusetts's law states that a BAC of 0. 8% is evidence of alcohol impairment but is not illegal "per se". South Carolina law states that a BAC of . 10 is evidence of alcohol impairment but is not illegal "per se".
Three states, Kentucky, Rhode Island and Texas have changed their BAC limit from . 10 to . 08 since the last report. Their 2000 fatality rates are included on the chart with those states having a BAC limit of . 08.
Massachusetts has the lowest motor vehicle fatality rate in the nation for 2000 with a . 8 fatality rate per 100 million vehicle miles traveled, followed by New Hampshire and Rhode Island with 1. 0 fatality rate respectively. Massachusetts had the lowest motor vehicle fatality rate in the nation for 1999 as well. Connecticut's fatality rate went up from the 1999 fatality rates by . 1. Connecticut had the second lowest motor vehicle fatality rate in 1999, and is currently tied with New Jersey and New York for the third lowest motor vehicle fatality rates in the nation in 2000. These results are based on data provided by the National Highway Traffic Safety Administration.
Tables 1 and 2 list the . 08 and . 10 states and figures 1 through 20 illustrate fatality rate comparisons, which are located in Appendix 1. Appendix 2 explains the statistical information and the methodology used.
EXPLANATION OF FIGURES
Figure 1 compares the motor vehicle fatality rates between states with . 08 and . 10 BAC limits. A statistical analysis reveals that there is no difference in their average fatality rates. Figure 2 compares the motor vehicle fatality rates for Connecticut, states with a 0. 8 BAC limit, and states with a . 10 BAC limit. Connecticut's rate is lower than both groups.
Figures 3 through 19 compare trends in Connecticut to each of the . 08 states. Connecticut's motor vehicle fatality rate has been lower than 12 of these . 08 states for all 25 years of this review and lower than the other eight in most years.
Figure 3 compares Connecticut's and Alabama's motor vehicle fatality rates. Connecticut's fatality rate has been lower than Alabama's for 25 years. Alabama established a . 08 BAC limit in 1995. Overall, the rate has been declining since 1986 where it had a dramatic increase.
Figure 4 compares Connecticut and California. Connecticut's rate has been lower than California's for 25 years. California established a . 08 BAC limit in 1990. Before that, its motor vehicle fatality rate had been declining since 1984. The decline continued until recently where the rate has stabilized since 1998.
Figure 5 compares Connecticut and Florida. Connecticut's rate has been significantly lower than Florida's for 25 years. Florida established a . 08 BAC limit in 1994. The rate has been declining since 1983 and experienced a slight increase in 1993. Florida's fatality rate has been fairly consistent from 1993 to 1999.
Figure 6 compares Connecticut and Hawaii. Since 1987, Connecticut's rates have been consistently lower than Hawaii's. Hawaii established a . 08 BAC in 1995. Overall, the rates have been declining since 1979. Since 1991, there have been slight fluctuations of increase and decrease. There was an increase in 1996 and again in 2000.
Figure 7 compares Connecticut and Idaho. Connecticut has had a lower rate than Idaho's for 25 years. Idaho implemented a . 08 BAC in 1997. Overall, the fatality had been declining since 1980.
Figure 8 compares Connecticut and Illinois. Connecticut has had a lower rate than Illinois' since 1985. Illinois has implemented a . 08 BAC in 1997. From 1975 to 1984, the Illinois' rate declined. From 1985 to 1989, it remained constant with one slight increase. From 1990 to 1993 the fatality rate declined again only to see an increase in 1994. After a slight decrease in 1996 and 1997, the rate has remained stable.
Figure 9 compares Connecticut and Kansas. Connecticut's rate has been lower than Kansas' for 25 years. Kansas implemented a . 08 BAC in 1993. From 1984 to 1992, the Kansas' fatality rate declined. From 1992 to 1993, it rose by 12. 5%. After implementation, the fatality rate has remained constant with some slight increases.
Figure 9A compares Connecticut and Kentucky. Connecticut's rate has been lower than Kentucky's for 25 years. Kentucky implemented a . 08 BAC in 2000.
Figure 10 compares Connecticut and Maine. Since 1983, Connecticut's rate has been lower than Maine's. Maine implemented a . 08 BAC in 1988. From 1983 to 1986 the rate declined. It remained constant from 1987 through 1988, and in 1989, decreased by 22. 7%. The rate has been relatively constant since 1989 and has experienced only small decreases.
Figure 11 compares Connecticut and Massachusetts. Since 1978, Connecticut's rate has been higher than Massachusetts'; except in 1996 through 1998, when it was lower. In 1994, Massachusetts changed its presumption law from . 10 to . 08. Massachusetts' rate has been declining since 1985 with slight increases in 1988 and 1996 through 1998. It has remained steady since 1999.
Figure 12 compares Connecticut and New Hampshire. New Hampshire implemented a . 08 BAC in 1994. Overall, New Hampshire's fatality rate has been declining since 1983 with minor increases in 1989 and 1996. Since 1994, Connecticut and New Hampshire have had similar motor vehicle fatality rates. New Hampshire was one of the two states in 2000 with a lower fatality rate than Connecticut.
Figure 13 compares Connecticut and New Mexico. Connecticut has had a lower rate than New Mexico's for 25 years. New Mexico implemented a . 08 BAC in 1994. New Mexico's fatality rate has been declining since 1981, with minor increases in 1988, 1993, 1997 and 2000.
Figure 14 compares Connecticut and North Carolina. Connecticut has had a lower rate than North Carolina's for 25 years. North Carolina implemented a . 08 BAC limit in 1993. Its rate steadily decreased from 1980 to 1983 and again from 1986 to 1992. The rate has been stabilized since 1993.
Figure 15 compares Connecticut and Oregon. Connecticut's rate has been lower than Oregon's for 25 years. Oregon implemented a . 08 BAC in 1983. Oregon's rate has been declining since 1976 except for a slight increase experienced in 1983.
Figure 15A compares Connecticut and Rhode Island. Rhode Island's rate has been lower than Connecticut's for most of the 25 years except for 1986-1988 and 1999. Rhode Island implemented a . 08 BAC limit in 2000.
Figure 15B compares Connecticut and Texas. Connecticut's rate has been lower than Texas' for 25 years. Texas implemented a . 08 BAC limit in 1999. Texas' fatality rates have generally decreased between 1981 and 1998 with an increase in 1996. The rate has remained the same since 1998.
Figure 16 compares Connecticut and Utah. Connecticut's rate has been lower than Utah's for 25 years. Utah implemented a . 08 BAC in 1983. Utah's rate has decreased each year since 1977 except for a slight increase in 1981. Utah's rate has been fairly constant since 1992.
Figure 17 compares Connecticut and Vermont. Connecticut's rate has been lower than Vermont's for 25 years. In 1991, Vermont implemented a . 08 BAC. Vermont's fatality rates have been decreasing since 1980 with slight increases in 1984, 1991, and 1995. The rate has been fairly constant since 1996.
Figure 18 compares Connecticut and Virginia. Connecticut's rate has been lower than Virginia's for almost all of the 25 years except for 1979-1982. Virginia's fatality rates have been steadily declining since 1984 with a few slight increases.
Figure 19 compares Connecticut and Washington. Connecticut's rate has been lower than Washington's for almost all of the 25 years with the exception of 1982-1983. Washington implemented a . 08 BAC in 1999. Washington's rates have been steadily declining since 1985.
APPENDIX 1
Table 1 lists the states with a . 10 BAC.
Table 1: States with a . 10 BAC as of 2000
Alaska** |
Maryland** |
North Dakota |
Arizona** |
Michigan |
Ohio |
Arkansas** |
Minnesota |
Oklahoma** |
Colorado |
Mississippi** |
Pennsylvania |
Connecticut |
Missouri** |
South Carolina* |
Delaware |
Montana |
South Dakota** |
Georgia** |
Nebraska** |
Tennessee |
Indiana** |
Nevada |
West Virginia |
Iowa |
New Jersey |
Wisconsin |
Louisiana** |
New York |
Wyoming** |
**South Carolina does not have an illegal per se BAC limit under its criminal law. However, it has a presumption law that someone with . 10 BAC is under the influence. For this reason, South Carolina was included in the . 10 states.
Since 2000, 13 states passed laws to establish a . 08 BAC limit. Nine of the states passed its law in 2001. Mississippi, South Dakota, and Wyoming's . 08 BAC limit goes into effect 7/1/02. For Louisiana, the . 08 BAC limit effective date is 9/30/03. Because the 2000 fatality rates compared in this report do not reflect a full year under the new standard, these states were included in the . 10 states.
Table 2 lists the states with a . 08 BAC and the year when the . 08 BAC was established.
Table 2: States with . 08 BAC and the year it was established as of 2000
States with . 08 BAC |
Year . 08 BAC was established |
Alabama |
1995 |
California |
1990 |
Florida |
1994 |
Hawaii |
1995 |
Idaho |
1997 |
Illinois |
1997 |
Kansas |
1993 |
Kentucky |
2000 |
Maine |
1988 |
Massachusetts* |
1994 |
New Hampshire |
1994 |
New Mexico |
1994 |
North Carolina |
1993 |
Oregon |
1983 |
Rhode Island |
2000 |
Texas |
1999 |
Utah |
1983 |
Vermont |
1991 |
Virginia |
1994 |
Washington |
1999 |
*Massachusetts does not have an illegal per se BAC limit under its criminal law. However, it has a presumption law that a person with a . 08 BAC is under the influence and a court-based license suspension system with a . 08 BAC that is similar to Connecticut's administrative per se system. For this reason, Massachusetts was included in the . 08 states category.
APPENDIX 2
We used a regression analysis to see whether there is statistically significant difference in fatality rates in states with . 08 BAC and states with . 10 BAC. We used this test to determine if the observed differences are due to (1) chance or (2) characteristics that are really different. As stated in the memo, the analysis demonstrated no statistical difference between the average rate of motor vehicle fatalities in states that use a BAC of . 08 and states with a . 10 limit.
The following is a discussion of the methodology we used.
REGRESSION ANALYSIS EXPLAINED
Regression analysis attempts to show a relationship between a dependent and independent variable(s). This analysis can be interpreted by looking at a number of measures. This first is called "R squared. " The R square indicates what percent of the variation across observations is due to the independent variable(s). Using the BAC data the R square is extremely low.
Additionally you may look at the t-score of any independent variable. The t-score is measured by dividing the X coefficient by the standard error of the coefficient. T-scores in excess of 2. 0 indicate that a variable is statistically significant even if it does not explain the variations as measured by the R square. In all cases analyzed the t-score is less than 2. 0 therefore not only are the BAC differences slight, but it is statistically insignificant.
KA: ts