Evaluation of Iowa s 70 mph Speed Limit Four Year s Later

Transcription

1 Evaluation of Iowa s 7 mph Spee Limit Four Year s Later Final Report December 21 Sponsore by the Iowa Department of Transportation (CTRE Project 6-247) Sponsore by The Iowa Department of Transportation (InTrans Project 6-247)

2 About the Institute for Transportation The mission of the Institute for Transportation (InTrans) at Iowa State University is to evelop an implement innovative methos, materials, an technologies for improving transportation efficiency, safety, reliability, an sustainability while improving the learning environment of stuents, faculty, an staff in transportation-relate fiels. Iowa State University Disclaimer Notice The contents of this report reflect the views of the authors, who are responsible for the facts an the accuracy of the information presente herein. The opinions, finings an conclusions expresse in this publication are those of the authors an not necessarily those of the sponsors. The sponsors assume no liability for the contents or use of the information containe in this ocument. This report oes not constitute a stanar, specification, or regulation. The sponsors o not enorse proucts or manufacturers. Traemarks or manufacturers names appear in this report only because they are consiere essential to the objective of the ocument. Iowa State University Non-iscrimination Statement Iowa State University oes not iscriminate on the basis of race, color, age, religion, national origin, sexual orientation, gener ientity, sex, marital status, isability, or status as a U.S. veteran. Inquiries can be irecte to the Director of Equal Opportunity an Diversity, (515) Iowa Department of Transportation Statements Feeral an state laws prohibit employment an/or public accommoation iscrimination on the basis of age, color, cree, isability, gener ientity, national origin, pregnancy, race, religion, sex, sexual orientation or veteran s status. If you believe you have been iscriminate against, please contact the Iowa Civil Rights Commission at or Iowa Department of Transportation s affirmative action officer. If you nee accommoations because of a isability to access the Iowa Department of Transportation s services, contact the agency s affirmative action officer at The preparation of this (report, ocument, etc.) was finance in part through funs provie by the Iowa Department of Transportation through its Agreement for the Management of Research Conucte by Iowa State University for the Iowa Department of Transportation, an its amenments. The opinions, finings, an conclusions expresse in this publication are those of the authors an not necessarily those of the Iowa Department of Transportation.

3 Technical Report Documentation Page 1. Report No. 2. Government Accession No. 3. Recipient s Catalog No. InTrans Project Title an Subtitle 5. Report Date Evaluation of Iowa s 7 mph Spee Limit 4 Years Later December Performing Organization Coe 7. Author(s) 8. Performing Organization Report No. Reginal R. Souleyrette an Dan Cook InTrans Project Performing Organization Name an Aress 1. Work Unit No. (TRAIS) Institute for Transportation Iowa State University 11. Contract or Grant No South Loop Drive, Suite 47 Ames, IA Sponsoring Organization Name an Aress 13. Type of Report an Perio Covere Office of Traffic an Safety Final Iowa Department of Transportation 14. Sponsoring Agency Coe 8 Lincoln Way Ames, IA Supplementary Notes Visit for color PDF files of this an other research reports. 16. Abstract On July 1, 25, the State of Iowa implemente a 7 mph spee limit on most rural Interstates. This ocument reports on a 4 year upate of a stuy of the safety effect of this change. Daytime an nighttime serious crashes were stuie for a perio of 14 years prior to the change an 4 years afterwars. Cross meian crashes were stuie for 4 years before an 4 years after. Simple escriptive statistics reveal increases in most crash severity categories for the 4 year perio following the spee limit increase when compare to the 4year perio prior to the increase. When compare to longer term trens, the increases were less pronounce in some severity levels an types, an for a few severity levels, the average crash frequencies were observe to ecrease. Few of the changes in crash frequency were larger than the normal year to year variation is these statistics. Three types of crash were foun to have increase by amounts larger than might be expecte by normal variation: nighttime fatal crashes (52%), serious cross meian crashes (25%) an all (total) crashes (25%). Only the increase in all (total) crashes was foun to be statistically significant at the 9% confience level. While not statistically significant at a high level of confience, the results suggest that further stuy shoul be unertaken to unerstan an consier actions to mitigate nighttime an cross-meian crashes. 17. Key Wors 18. Distribution Statement 7 mph spee limit safety spee limit increase No restrictions. 19. Security Classification (of this report) 2. Security Classification (of this page) 21. No. of Pages 22. Price Unclassifie. Unclassifie. 23 NA Form DOT F 17.7 (8-72) Reprouction of complete page authorize

4 i

5 EVALUATION OF IOWA S 7 MP SPEED LIMIT 4 YEARS LATER Final Report December 21 Principal Investigator Reginal R. Souleyrette Geral an Aurey Olson Professor of Civil Engineering Center for Transportation Research an Eucation, Iowa State University Grauate Assistant Dan Cook Preparation of this report was finance in part from funs provie by the Iowa Department of Transportation through its research management agreement with the Institute of Transportation, Center for Transportation Research an Eucation, CTRE Project A report from Institute for Transportation Iowa State University 2711 South Loop Drive, Suite 47 Ames, IA Phone: Fax:

6 Table of Contents ACKNOWLEDGMENTS... IV INTRODUCTION...1 METODOLOGY...1 Crash Data Assembly an Processing...2 Crash Analysis...4 RESULTS...4 Descriptive Statistics...4 Paire T-test...1 SUMMARY AND CONCLUSION...11 APPENDIX: PAIRED T-TEST COMPUTATIONS... A-1 iii

7 ACKNOWLEDGMENTS The contents of this report reflect the views of the authors, who are responsible for the facts an the accuracy of the information presente herein. The opinions, finings, an conclusions expresse in this publication are those of the authors an not necessarily those of the sponsors. The sponsors assume no liability for the contents or use of the information containe in this ocument. This report oes not constitute a stanar, specification, or regulation. This project was supporte by the Iowa Department of Transportation, Office of Traffic an Safety, using Traffic Safety Improvement Program funs. iv

8 INTRODUCTION On July 1, 25, the State of Iowa implemente a 7- mph spee limit on most rural Interstates. Concern has been expresse that the change may have resulte in a ecrease in safety, reflecte in an increase in crashes. An early stuy (Lun et al, 27; Lun 27) reporte the preliminary safety effect of the spee limit change using eighteen months of after perio ata (June 25 December 26). That stuy escribe how the spee limit change may have affecte spees an safety on other categories of highways in Iowa. Those roaways inclue parallel (to the rural Interstates) routes, expressways, primary highways, urban Interstates, an rural other highways. The stuy also examine specific portions of the rural Interstates, namely, those within 2 miles of a metropolitan area. The safety performance of I-8 east an west as well as I-35 north an south was examine. The initial stuy was later upate using 2.5 years of ata (Souleyrette et al, 29). This report presents the results of a stuy of the impacts of that spee limit increase, comparing 4 years of after-perio crash ata to 14 years of ata from before the change. Descriptive Crash analysis is presente for the following crash types: o Total crashes o Crashes by severity o Cross-meian crashes o Day-night crashes METODOLOGY To examine the effect of spee limit change, it is important to isolate the effect of the change from other variables that systematically might affect crash frequency. Typically, traffic level (ADT, VMT) is the most significant preictor of crash rate. The analysis presente in this report was simplifie as VMT change very little on the stuy roas (rural Iowa Interstate) over the 4 year before an after stuy perio. Other effects can be policy changes (DWI, seatbelt laws), enforcement, vehicle technology changes, etc. There were no significant policy changes that woul have affecte crash likelihoo on Iowa Interstates uring the perio, an enforcement levels an vehicle technologies were assume as constant for this stuy. Figure 1 shows the annual vehicle miles travele on rural interstates in Iowa in millions of vehicle miles travele. The four year average VMT preceing the change to 7 mph (21-24) was 52 million vehicle miles travele compare to 51 million vehicle miles travele for the average four year after perio (26-29). The change in traffic volume from the before to after perio is only an increase of 1.6%. 1

9 6 Iowa Rural Interstate VMT In Millions (1^6) Figure 1: Annual VMT on Iowa Rural Interstates Crash Data Assembly an Processing Crash ata were obtaine from the Iowa DOT Office of Traffic an Safety. The ata inclue the perio 1991 to 29, covering 14 years before an 4 years after the year of the spee limit change (25). ighway ata an attributes were extracte from the Iowa DOT s statewie Geographic Information Management System (GIMS) ata base. The proceures to ientify interstate crashes for the subset (pre-gims) involve linking crash information recors to point crash locations, an using spatial an attribute recors to select crashes on the interstates. Sections with 7-mph spee limits were ientifie using 26 GIMS ata. These sections were also ientifie in oler versions of GIMS that more closely match the recore locations of crashes in the before perio. Using GIS selection methos, crashes in these areas were selecte. Cartographic representation of the highway system in Iowa change significantly over the time frame of this stuy. As such, there are many crash locations that o not match perfectly with any one GIS representation of the roaway system. The istance from each crash location to both the nearest rural interstate segment an to the closest other type roa was establishe. Crashes were then classifie into four groups: 1. Crashes greater than 1 meters from the Interstate (7 mph sections, or INT7) 2. Crashes within 1 meters of INT7 3. Crashes between 1 an 1 meters of INT7 where istance to the Interstate is less than the istance to other roas 4. Crashes between 1 an 1 meters of INT7 where istance to the Interstate is more than the istance to other roas 2

10 The secon an thir groupings were classifie as stuy area crashes (those that occurre on segments that now have a 7 mph spee limit). Figure 2 presents the urban an rural Interstates in Iowa. Re segments inicate the sections where spee limit was raise to 7 MP. A statewie list of locations of possible/probable cross meian crashes was provie by the Iowa DOT Office of Traffic Safety (TAS). This list was use in the present stuy an merge to the list of crashes occurring on segments where the spee limit was change to 7 MP. While the overall magnitue of cross meian crashes coul be significantly ifferent (most likely higher) from the reporte number, the methoology use by TAS was applie equally in the before an after perios. Therefore, more confience shoul be place on the magnitue of the change expresse as a percent, rather than the absolute numbers of crashes. To ientify nighttime crashes, a table of sunrise an sunset times was constructe covering the 1991 to 29 perio, base on Ames observations (as an average for the state) 1. This table was joine to the crash ata base an use to select nighttime crashes (crashes before sunrise or after sunset on any given ay). The time of ay was available on all but 237 of the 35,733 total crashes in the atabase. Detaile steps of the ata collection, preparation an processing may be foun in Appenix A of the 2.5 year upate (Souleyrette et al, 29). Figure 2: Iowa Interstate Stuy Sections 1 U.S. Naval Observatory ( 3

11 Crash Analysis Simple measures of crash performance incluing averages an stanar eviations were prepare from the assemble crash ata. Daytime an nighttime serious crashes were stuie for a perio of 14 years prior to the change an 4 years afterwars. Due to limitations of ata, cross meian crashes were stuie for 4 only years prior to the change. Mean changes by crash severity categories for the 14 an 4 year before an 4 year after perios were compute an compare to 14 or 4 year stanar eviations to inicate the most important changes. 25 ata were exclue to eliminate the impact of temporary changes an to provie equal number of seasons in the before an after perios. Paire T-tests were performe on the crash types experiencing larger increases using ata from the 3 counties which contain rural expressway sections. The results of this test inicate which increases are statistically significant. Aitional information about the t-tests may be foun in the appenix to this report. RESULTS Descriptive Statistics Fatal crashes increase on average from 19.5 to 25.3 (5.8) per year resulting in a 29.5 percent increase when compare to the 4 year before perio. Compare to the average crash frequency over the entire 14 year before perio, fatal crashes increase from 2.8 to 25.3 (4.5 or a 21.2 percent increase). owever, these increase are on the orer of the annual variability (stanar eviation) in crash frequencies over the longer perio (21.3 percent of the mean value or about 4.4 crashes per year). Therefore, while the increase in fatalities is tragic in terms of the aitional 5 lives lost per year, the changes are similar to what might be expecte from ranom variation in the ata (see Figure 3). 4 Fatal Crashes sections increase to 7 mph Figure 3: Annual Fatal Crash istory 4

12 Serious (fatal an major injury) crashes increase on average from 8. to 87.3 (7.3) per year resulting in a 9.1 percent increase when compare to the 4 year before perio. owever, compare to the average crash frequency over the entire 14 year before perio, serious crashes ecrease from 14 to 87.3 (16.7 or a 16.1 percent ecrease). The annual variability (stanar eviation) in crash frequencies of this severity over the longer perio (2.7 percent of the mean value or 21.5 crashes per year) is also on the orer of these changes. See figure 4. Serious Crashes sections increase to 7 mph Figure 4: Annual Serious Crash istory Nighttime fatal crashes increase, although to a lesser extent than all fatal crashes. This type of crash increase on average from 6.75 to 1.25 (3.5) per year resulting in a 51.9 percent increase when compare to the 4 year before perio. owever, compare to the average crash frequency over the entire 14 year before perio, fatal crashes increase from 8.34 to 1.25 (1.91 or a 22.8 percent increase). The annual variability (stanar eviation) in nighttime fatal crash frequencies over the longer perio is 28.5 percent of the mean value or 2.38 crashes per year. Therefore, while compare to the 14 year before perio, the increase is on the orer of what might be expecte ue to ranom variation, compare to the 4 years before the spee limit change, the increase is more significant. See figure 5. 5

13 Nighttime Fatal Crashes sections increase to 7 mph Figure 5: Annual Nighttime Fatal Crash istory Serious nighttime crashes increase slightly from 3.3 to 31. (.7) per year resulting in a 2.5 percent increase when compare to the 4 year before perio. But, compare to the average crash frequency over the entire 14½ year before perio, serious nighttime crashes ecrease from 43. to 31. (8. or a 27.8 percent reuction). Again, however, the annual variability (stanar eviation) in serious nighttime crash frequencies over the longer perio (23.3 percent of the mean value or 1. crashes per year) is on the orer of magnitue of this change. See figure Nighttime Serious Crashes sections increase to 7 mph Figure 6: Annual Nighttime Serious Crash istory Fatal cross-meian crashes increase on average from 5.5 to 8. (2.5) per year resulting in a 45.5 percent increase when compare to the 4 year before perio. owever, the annual variability (stanar eviation) in fatal cross meian crash frequencies over the 4 year before perio is 43.3 percent of the mean value or 2.38 crashes per year. See figure 7. 6

14 12 Cross Meian Fatal Crashes Figure 7: Annual Cross-Meian Fatal Crash istory Serious cross-meian crashes increase on average from 1. to 12.5 (2.5) per year resulting in a 25. percent increase when compare to the 4 year before perio. The annual variability (stanar eviation) in serious cross meian crash frequencies over the 4 year perio (14.1 percent of the mean value or 1.41 crashes per year) is much less than the observe increases, inicating a more significant fining. See Figure Cross Meian Serious Crashes Figure 8: Annual Cross-Meian Serious Crash istory All cross-meian crashes (total of all severity levels) increase on average from 46.3 to 55.5 (9.2) per year resulting in a 2. percent increase when compare to the 4 year before perio. The annual variability (stanar eviation) in total cross meian crash frequencies over the 4 year before perio (15.3 percent of the mean value or 7.9 crashes per year) is on the orer of the observe increases. See Figure 9. 7

15 8 Cross Meian Crashes All Severities Figure 9: Annual All Severities Cross Meian Crash istory All nighttime cross-meian crashes (total of all severity levels) increase on average from 15.5 to 16.8 (1.3) per year resulting in an 8.1 percent increase when compare to the 4 year before perio. The annual variability (stanar eviation) in total cross meian crash frequencies over the 4 year perio (16.3 percent of the mean value or 2.52 crashes per year) is greater than the observe increases, inicating the change in crashes is within the expecte variability. See Figure 1. 4 Cross Meian Crashes Nighttime All Severities Figure 1: Annual All Severities Nighttime Cross Meian Crash istory All (total) crashes increase from 1747 to 2187 (44) per year resulting in a 25.2 percent increase when compare to the 4 year before perio. Compare to the average crash frequency over the entire 14 year before perio, all crashes increase from 1785 to 2187 (42 or a 22.5 percent increase). The annual variability (stanar eviation) in total crash frequencies over the longer perio (6.8 percent of the mean value or crashes per year) is much smaller than magnitue of these changes inicating significant increase. See figure 11. 8

16 3 All Crashes sections increase to 7 mph Figure 11: Annual All Severities Crash istory Analysis of off-system (non- rural interstate an other) roas was conucte earlier uring the stuy, using 18 months of before an after ata. Results from that analysis are presente in Appenix C of the 2.5 year upate report (Souleyrette et al, 29). Table 1 reports increases in the crash severity categories for the 4 year perio following the spee limit increase (26-29) as compare to the 4 year perio prior to the increase (21-24). Mean changes that are large compare to stanar eviation are highlighte in yellow. 9

17 Rural Interstate crashes before an after 7MP spee limit Annual Average for 14 years before Annual Average for 4 years before*** Annual Average for 4 years after? change ** (compare to 4 year avg) change ** (compare to 14 year avg.) annual stanar eviation (14 yrs. before) % annual stanar eviation (14 yrs. before) Fatal crashes % 21.2% % Fatal crashes Fatalities % na % an Cross-meian fatalities * % * 2.38* 43.3% * crashes Night-time % 22.8% % crashes Serious (fatal plus major injury crashes) All serious crashes Serious Crossmeian crashes Serious Nighttime crashes % -16.1% % * % * 1.41* 14.1% * % -27.8% % All crashes All Crashes % 22.5% % All Crossmeian crashes * % * 7.9 * 15.3% * Night-time cross-meian * % * 2.52 * 16.3% * crashes * cross meian crash ata only available for 4 years of the "before" perio ** change shoul be large compare to stanar eviation for high statistical confience *** previous 4 year perio Jan 21 - Dec 24? after 4 year perio Jan 25 - Dec 29 Table 1: Descriptive Statistics an Results Paire T-test Paire T-tests were performe on the crash types experiencing larger increases using ata from the 3 counties which contain rural expressway sections (night-time fatal crashes, serious crossmeian crashes, an all crashes. The results of this test inicate that only total (all) crashes 1

18 increase significantly. Aitional information about the t-tests may be foun in the appenix to this report. SUMMARY AND CONCLUSION This stuy investigate the safety effect of the July 1, 25 7mph spee limit in Iowa. Crash statistics for sections of rural Interstate were compare using 14 years of ata before an 4 years after ata. Three types of crash were foun to have increase by amounts larger than might be expecte by normal variation: nighttime fatal crashes (52%), serious cross meian crashes (25%) an all (total) crashes (25%). Only the increase in all (total) crashes was foun to be statistically significant at the 9 percent confience level. owever, while not statistically significant at a high level of confience, the results suggest that further stuy shoul be unertaken to unerstan an consier actions to mitigate nighttime an cross-meian crashes. 11

19 REFERENCES V. Lun (27), The 7-MP Spee Limit: Spee Aaptation, Spillover an Surrogate Measures of Safety. Masters Thesis. Iowa State University. Department of Civil, Construction an Environmental Engineering. August. Lun, V., R. Souleyrette an T. Stout (27), 7-MP Spee Limit: Spee Aaptation, Spillover an Surrogate Measures of Safety, Mi-Continent Trans Research Symposium. Ames, Iowa. August. Souleyrette, R., T. Stout an A. Carriquiry (29). Evaluation of Iowa s 7 mph Spee Limit 2.5 Year Upate, Iowa DOT, CTRE Project 6-247, January. 45 pp. 12

20 APPENDIX: PAIRED T-TEST COMPUTATIONS Comparing only serious crashes (fatal an major injury crashes) Before perio: After perio: Note: n=3 counties with rural interstate µ n = 3 α =.5 =.9667 s t = after before a µ : µ = : µ = 6.94 = 2.45 D.9667 t statistic = = s n Reject.8688 < t if t - statistic t Cannot Reject =.8688 Result: p-value = between.1 an.25 (greater than alpha) There is no significant increase or ecrease of serious crashes from 65 mph segments to 7 mph segments. Comparing ALL crashes Same efinitions as before

21 µ n = 3 α =.5 = s t = µ after before a : µ : µ = = 2.45 Reject = D t statistic = = s n > t if t - statistic t Reject = Result: p-value <.5 There is a significant change of total crashes from 65 mph segments to 7 mph segments. µ : µ : µ n = 3 α =.5 t α = µ after before a = Reject > D t statistic = = s n > t if t - statistic t α Reject α = Result: p-value <.5 There is a significant increase of total crashes from 65 mph segments to 7 mph segments. Comparing SERIOUS Cross-Meian Crashes Same efinitions as before A-2

22 µ n = 3 α =.5 =.3333 s t = after before a µ : µ = : µ = = 2.45 D.3333 t statistic = = s n Reject < t if t - statistic t Cannot Reject = Result: p-value = between.1 an.25 (greater than alpha) There is no significant increase or ecrease of serious cross-meian crashes from 65 mph segments to 7 mph segments. Comparing FATAL Nighttime Crashes Same efinitions as before µ n = 3 α =.5 =.3333 s t = after before a µ : µ = : µ = = 2.45 D.3333 t statistic = = s n Reject < t if t - statistic t Cannot Reject = A-3

23 Result: p-value = between.1 an.25 (greater than alpha) There is no significant increase or ecrease of fatal nighttime crashes from 65 mph segments to 7 mph segments. A-4