Schwartz_Automobile

8/7/2019 Schwartz_Automobile

1/30

Social Benefits and Costs ofe Automobile

Joel Schwartz

Have 'we been forced into autom obile d epen den ce by a n unholy alliancebetw een carmakers, roadbuilders, an d governm ent pla nn ers? From t h eperspective of th e autom obile's critics, Americans have a n irrational loveaffair with th e a utom obile th a t resu lts in a range of social ills from pollu-tion t o congestion, makes u s dep ende nt on unstab le oil-producing coun-

ies, and drains our private and public coffersf scarce resources that would be better put to

n the other hand, for its propo-obile is the most convenient

l and economic opportunity and autono-human history

The automobile's critics-urban planners

automobile activists-claim that urban"sprawl" forces people to drive by pushingthem far from work, shopping, and otheractivities. Americans' reliance on the automo-bile, they say, is not a choice, but a necessityforced on people by suburban developmentpatterns that require an automobile for mosttravel. These development patterns have inturn been encouraged, so the argument goes,by implicit and explicit subsidies that make

Joel Schwartz is a visiting fellow at the American Enterprise Institute, where he studies air pot-lution, transportation, climate change, and chemical risks. He is the author of No Way Back:Why Air Pollution Will Continue to Decline and the forthcoming AEI book Air Quality in America.He holds a bachelor's degree in chemistry from Cornell University and a master's degree inplanetary science from the California Institute of Technology. In 1993,he studied Europeanapproaches to transportation and air quality policy as a German Marshall Fund Fellow.


2/30

traveling by car seem much cheaper than itreally is.

The result of automobile favoritism, in theview of critics, has been a wide range of auto-mobile-imposed ills, such as sprawling sub-urbs that thwart community involvement,long commutes that steal time away fromsocial and family interaction, air pollution,injuries, obesity, and a drain on families'budgets from excessive transportation costs.'If people had to pay explicitly for the fullsocial costs of driving, they argue, peoplewould live in denser, more urbanized housingand walk or use public transit for most travel.Furthermore, people would be happier andhealthier as a result of these lifestyle changes.

This view is mistaken. If Americans wereforced into suburbs and driving against theirwill, we would expect transportation and landuse to look quite different in other countrieswith policies less favorable to automobile hav-el. Yet people all over the world choose subur-ban lifestyles and automobile-based trans-portation as soon as they become wealthyenough to afford them. This is true in Europe,where, despite $yper-gallon gasoline andother high automobile-related taxes, as well aswidely available public transportation, auto-

mobiles account for about 78 percent of allmotorized travel and transit accounts for about16 percent.' Transit's share of all Europeanmotorized travel dropped 35 percent from 1970to zoo0 and continues to decline. Europe's oldcentral cities are now surrounded by suburbsthat look very much like their American coun-terparts. People also choose driving in thedeveloping world, where demand for automo-biles is rising faster than income, despite poorroads and high levels of traffic congestion.

Even without reference to other countries,the claim that Americans had the automobileforced on them does not stand up to scrutiny.By 1930, Americans already owned an averageof three automobiles for every four households,showing that Americans embraced automobiletravel long before there were interstate high-ways and long before the postwar suburban-ization of American metropolitan areas.

The critics have it exactly backwards. Theautomobile is a powerful enabling technolo-gy, allowing people the world over to satisfywhat seems to be a deep-seated human desirefor space, privacy, mobility, and autonomy.Automobile travel has vastly increasedhumankind's wealth and prosperity. Com-pared to other modes of travel, the speed and

'See, for example, Reid Ewing, Rolf Pendall, and Don Chen, Measuring Sprawl and Its Impact, Smart GrowthAmerica, 2003, at www.smartgroruthameric~1.com/spraw2index/MeasuringSprawl.PDF(May 16,2005); Jane H. Kay,Asphalt Natio n: How the Autornobile Took Ov er America and How W e Ca n Take I t Back (New York: Crown, 1997);James J. MacKenzie, Roger C. Dower, and Donald D. T. Chen, The Going Rate: Wh at It Really Costs to Drive(Washington, D.C.: World Resources Institute, 1992); Public Interest Research Group, More Highways, MorePollution: Road-Building and Air Pollution i n America's Cities, March 2.004, at wiuw.uspirg.oy/reyo+ts/~WoreHighwaysMorePollution3_oq.pdf(May 16, zooj); Michelle Emst, Mean Streets 2004: How Far Hav e W e Come?Pedestrian Safety, 1994-2003, Surface Transportation Policy Project, November 2004, at www.transact.org/libray/r~.ports~html/mszooqlpdl/Final~Mean-Streets-z004~~.pdf(May 16,2005); and Surface Transportation Pol-icy Project, "Transportation Costs and the American Dream: Why a Lack of Transportation Choices Strainsthe Family Budget and Hinders Home Ownership," July 2003, at www.trnnsact.org/library/decoder/american-dream.pdf (May 16 , 2005).

'Transit's share includes inter-city bus and rail travel in addition to travel within metropolitan areas. SeeEuropean Communities, Panorama of Transport: Statistical Oz~ovl'mof Transport in the European U nion , Part z (Lux-embourg: zoq),at http://~p.eurostat.ce~~eu.int/cffche~~OFFPUBKS-DA-o~-oo~-z~N~S-DA-o~-(May 16, ?.cog).


3/30

of automobile travel gives peopleto many times more choices of hous-

r pool of potential

f travel. Even after including the mostf air pollution and other nega-s of automobile travel, automo-

cost only about half as much as tran-tunities, allowing people to visit friends

relatives who would be too far away bymodes. Not only do wealthier

While the automobile confers enormousts, automobile travel also has undesirable

ts. As with most things in life, auto-des a set of unavoidable

o not want. Most people preferxible, rapid, and safe transportation to work,

ing, and leisure. Most people also dislikeThe history of the automobile is mainly a

f how Americans got more of whated out of the automobile and less of

they did not want. Technology has dras-

pared with the average car on theAutomobile safety has also substantiallyproved. Per mile of driving, the risk of

a car accident has declined more7; percent since 1960. Suburbanization,f millions of Americans to move to larg-

er homes and to areas with less traffic conges-tion, less crime, and quieter streets.

Based on evidence to be presented below,this chapter will show the following: Thedominance of automobile transportation overother modes is mainly the result of consumerchoice, rather than subsidies or coercion, andoverall, the automobile and its associatedroad and street infrastructure delivers hugenet benefits to Americans that could not beobtained by any other means currently avail-able or likely to be available for the foresee-able future. In addition, as will be shown hereand in other chapters of this book, the bene-fits of automobile travel can be retained andaugmented while, at the same time, undesir-able side effects can continue to be reduced.A Wo~ldwideLove Affair w ith theAutomobile

Americans have a "love affair" with theautomobile, or so goes the conventional wis-dom. Driving and suburbs are so popular inAmerica because p~iblicpolicy makes drivingartificially cheap, land-use planning favorssuburbs, and transit is starved for funding.On the other hand, in this view, other wealthycountries tax drivers to make them pay theirfair share of the costs they impose on others,and planners ensure that people live in densecities with convenient public transit.

Data from around the world, however,show that the conventional wisdom is mis-taken. After accounting for income-as meas-ured by gross domestic product (GDP) percapita-automobile ownership and use fol-lows a similar pattern all around the world.For example, FIGUREI shows the relationshipbetween cars per capita and GDP per capitain 1992 in a sample of 26 countries, rangingfrom very poor to the richest in the world.Note the strong correlation between incomeand car ownership.


4/30

40 2isl Centtiry Highways

Automobiles Per Capita vs. GDP Per Capita, 1992

r 'rer Grc+ Prt r lsr

r H e xr SKo

r lu r4 6 8 10 12 14 16 18 20

GDP Per Capita (thousands)Notes: Tpe countr:es cl~s tere dat the b w e r left are Pakstan. China, and lnoia.Aus = Austria: As? = A i i i t r~ l ia .GDP per ca pta isa: parc hsing ;iower partty (PPP), meaning tha t GDP levels across cointnes were est~matedby asking ho w m ucn it wou ld cost ina c o u n t r ~ slocal currency to buy a given basket o f goods and services.Source:j. Dargay and D.Ga teiy inccrnek Efec: on Cor ond Veh!c!e Owrerihtp, Worldwide: 1960-2015 (N ew York: C. V S tarrCenter New 'York University 1997).

FIGURE2 shows the trend in car ownershipversus income. The head of each arrow givesthe value for 1992, while the tail gives the valuefor 1970. Both axes are on a logarithmic scale,meaning that each factoruf-ro increase in GDPor car ownership covers the same physical dis-tance along an axis of the graph. For example,the distance traversed is the same in going froma GDP per capita of loo to 1,ooo as it is for goingfrom 1,000 to 10,ooo. This makes it easier to seethe path taken by countries at very low GDPsper capita. Note that increases in cars per capitatrack closely with increases in GDP per capita.Furthermore, note the steeper slope for poorercountries, which means that, at lower incomes,--- .- -..

automobile ownership rises more rapidly withincreasing income than it does for wealthiercountries.Infact, car ownership seems to be lev-eling off in the US., suggesting that ownershipis reaching "saturationr'-that is, an ownershiprate of about one car for every licensed driver.

These data suggest that high automobileownership in the US. is due largely to highincomes rather than to specific policies towardautomobiles. Per capita income in the UnitedStates has been and continues to be about 15percent to 40 percent higher than inother West-em countries.? This would be expected to beassociated with greater per capita automobileownership. Americans adopted the automo-

-.

?United Nations, Hum an D melopmenf Reports, at www.tindp.orX/hdrzooj/indi~afor/e.~~eI/11d1o~~tabie.xls


5/30

bile as their main form oftransportation long beforethe creation of the inter-state highway system orthe postwar suburbaniza-tion of metropolitan areasthat supposedly facilitatedthe automobile's domi-nance. By 1930, American5already owned more thanthree automobiles forevery four households.4

Once differences inincome are accounted for,these data show nothingparticularly special aboutthe United States in termsof automobile ownership.All around the world,people purchase automo-biles as soon as theybecome wealthy enoughto afford them. Given thatautomobiles represent amajor expenditure, thissuggests that peoplearound the world perceivelarge net benefits fromautomobile travel. --People purchase auto-

.................... -. ..................Trend in Car Ownership vs. Per Capita GDP,1970-1 992

I Carlpopulation ratio: 1970 & growth to 1992

....... ........I E-005 L .... .-,---L.,0.5 I 2 3 4 5 6 7 8 9 1 0Per-Capita GDP: 1970 & growth to 1992

. .mobiles even in countries with few roads, developing countries, while demand for roadwidespread public transit, and heavy traffic travel is relatively insensitive to the marginalcongestion. The demand for automobiles is costs of additional road travel. Thus, "incomeincreasing more quickly than income in matters more than gasoline prices."j Howev-

............ .. .-American households were substantially larger back then-more than four people per household, com-pared with about 2.6 today. The number of automobiles per person in 1930 was about 0.22. Automobile own-ership rates were calculated from US. Bureau of the Census, Historical S tatistics of the U nited Stntes; ColonialTimes to 197 0, Part 2 , September 1975, at wruw2.cens~~s.goz~/prodz/stntcomp/documents/CTg~opz-01.~df(May 16,2oo5), and Federal Highway Administration, Highway Statistics Sumnrnry to 1995,1997, at wruw.fhwa.dot.golr/ohirrr/st~rnmayg5/inde~.html(May 16, rooj).

5Peter Gordon and Harry Richardson, "The Geography of Transportation and Land Use," in R. Holcombeand S. Staley, eds., Smarter G m w th : Market-Bnsed Plnnn ing Strategies for Land-Use Plarrning in the 21st Ce nt ur y(Westport, Conn.: Greenwood Press, 2001).


6/30

42 21ct Century H i g h a y s

Total Passenger Miles of Travel by Travel Modein the European Union, 1970-2000

-.-Total -Automobile A Buses and Coaches -+Trams and Metros +-Rail .--AirNotes:"Raii" includes both com mu ter raii and inter-city rail. "Buses and Coaches" inciudes both urban and inter-city buses."Trams and Me tros" includes all m etrop olitan ra :i trans'it, except fo r c om mute r rail. Data are for the IS mem ber cduntries in theEuropean tinion .Source:European Comm unities, Panorama of i rons port Stalistaal Overv~ewof Transcon in the European Unon, P2r t 2,2003

er, because of poorer roads, these increases in cent of all person-miles of motorized travel indemand for cars and driving have occurred America and about 78 percent in Europe.7 Justunder "congestion levels and traffic condi- as in the U.S., public transit's share of all trav-tions that would appall ~mericans ."~ el has been declining for decades. In U.S. met-

When it comes to automobile use, people in ropolitan areas, transit accounted for about 3.6other wealthy countries also resemble Ameri- percent of motorized travel in 1970 but onlycans. Automobiles account for about 88 per- about 1.9 percent in 2000.~Europeans used

6 ~ b i d .7U.S. Department of Transportation, Bureau of Transportation Statistics, National Transportation Statistics

2004, January 2005, at www.bts.gov/publications/nafional~transportationstatistics/zoo/indextm1(May 16,2005);European Communities, Panorama of Transport, Part z.

'US.data are for travel within metropolitan areas only. Transit accounted for 1.1 percent of all person-miles of motorized travel in 2002 in the US. as a whole. National data for 1970 were not available. In 1980,transit accounted for 1.4 percent of all passenger-miles of motorized travel. Bureau of Transportation Statistics,National Transportation Statistics 2004; Wendell Cox, "US. Urban Personal Vehicle& Public Transport MarketShare from 1900,'' The Public Purpose, 2004, at zu?~izv.publicpurpose.com/ut-usptshare45ht(May 16,2oo5).


7/30

European Market Share of Each Transportation Mode, 1970-2000

Notes: "Raii",nri~;es oot-inco mm ut er '-a1 and inter--c:y :?A"Buses and Coaches" ~nc i~de?both w b a rard mTer-Cl~iyhilies. "T - ac s ard Verros" .cciudes ail rcer ropo lkan ra:i 'ran?::, ex:o2i kr co- i rmrer r a ~ l .aaa a-e for the 15 r-ember :oL?rr.es .n;Re European Lmon.

transit for 25 percent of motorized travel in1970, bu t only 16 percent by 2000.9 Between1980 and 1995, transit's share of all person-miles of travel declined 14percent in London,24 percent in Paris, 60 percent in Frankfurt,and 19 percent in Stockholm.'"

Many Americans are under the impressionthat Europeans travel a great deal on trains,but even in 1970, urban and inter-city railaccounted for only 1 2 percent of all passen-ger-miles of motorized travel, declining to 7.4percent in 2000.' ' Most of this travel is oninter-city and commuter rail. Urban rail-thetrams and metros that come to mind whenthinking about how Europeans get aroundtheir cities-accounted for only 1.1percent ofpassenger-miles in 2000. The decline in railtransport's share is due in part to increases in

. ~~.... ...,..

automobile travel, but probably even more soto increases in air travel, drivenby lower costsfrom airline deregulation. Air's share of allmotorized passenger transport nearlyquadrupled in Europe between 1970 andZ OOO . ' ~ Rail's share of passenger traveldecreased despite large expansions of high-speed rail in several European countries dur-ing the past few decades.

Air travel appears to be the main substitutefor rail travel, ~ ~ h i l ebuses appear to be themain substitute for automobile travel.'; In theU.S. and Europe, automobile and bus traveltogether account for about the same percent-age of total passenger-miles, and the same istrue for rail and air travel together.'i The dif-ference in the U.S. is that rail and buses makeup less than r percent of total passenger-miles

9The European data include both intra- and inter-city transit. Thus, for example, "buses and coaches"includes both transit buses and inter-city coaches, while "rail" includes both commuter and inter-city trains.European Commission, l'flnor~mzizof Transport, Part 2.

"'Wendell Cox, "Public Transport Market Share Trends: International Urban Areas from 1980," The PublicI'urpose, 2003, a t iuic~w.ptihlicpur~~0~e~co1n/ut-ittt1~~ktj.htm(May 16, 2005).

i l European Communities, Panorizrtm i,JTm~~spori,Part 2.'21birf.'34i r travel has probably also replaced some long-distance automobile travel.'4U.S. market share was calculated from Bureau of Transportation Statistics, Ni~tionnlli .anspovtdun Statis-

firs 2004.


8/30

44 iisr Centdry Highways

in these two groups, while in Europe, they slightly, from 59 percent to 62 percent,account for several percent of passenger-miles. between 1970 and 1994. Population densities

FIGURE3 displays the European trend in in European metropolitan areas declinedtotal passenger-miles of motorized travel, more than 60 percent between 1960 andbroken down by travel mode. Note the rapid 1990.'~Not only suburbanization, but also arise in total travel by automobile and air- decline in the number of people per house-- - -plane, while total travel by other modes has hold explains the density declines observedchanged little during the past few decades. in Western countries.'qTABLEI summarizes the trend in market share About the only major difference betweenof transit for the various modes. The "All the United States and Europe is in miles ofTransit" column gives the total market share automobile travel per capita each year. Whenfor the three public transit modes. compared at similar per capita incomes, Amer-

Europeans have also been moving to sub- icans drive their cars about 2,000 more milesurbs that look very much like suburbs in the per year than Europeans, as shown in FIGUREUnited States." The fraction of U.S. metropol- 4. The graph compares vehicle kilometers trav-- -itan-area residents living in suburban coun- eled each year versus per capita income, show-ties grew from 48 percent to 58 percent ing the trend in both variables from 1970 tobetween 1960 and 1990 and reached 62 per- 1997.2' Note that at any given income, Ameri-cent in 20oo.'~European cities are following cans drive more. On the other hand, per capitasimilar trends. Between 1968 and 1990, the driving is increasing more rapidly in Europefraction of Parisians living in suburbs grew than in the U.S. Between 1980 and 1998, pas-from 68 percent to 77 percent. Amsterdam's senger-miles per capita traveled by automo-suburban share grew from 20 percent to 33 biles increased 55 percent in the Europeanpercent between 1970 and 1994, while Union but only 25 percent in the U.S.2'Zurich's grew from 62 percent to 71 percent The data show that people in other parts ofbetween 1970 and 1995.'~London is a rare the world are not so different from Americansexception. Its suburban fraction grew only after all. As people become wealthier, they

IsHarry Richardson and Peter Gordon, Is Sprawl inevitable? Lessonsfrum Abroad (Los Angeles: University ofSouthern California, November 1999), at ruww-rcf.usc.edu/-pgordon/pdf/LESSON_A.pdf(May 16,2005).

%J.s. Bureau of the Census, "American Factfinder," at http://factfinder.cmsus.got~/serulet/DTGeoSearchByList-S e r v l e t ? d s - n a r n e = D E C - ~ o o o ~ S F ~ ~ U 6 ~ l a n ~ = 6 0 ~ 8 ;Alan E. Pisarski, Com mu ting i n America 11,EN0 Transportation Foundation, 1996; and G. Giuliano, Land Use Policy and Transportation: W h y W e Won 't GetThere from Here (Los Angeles: University of Southern California, July r999), at zoww.usc.ed t~/schools /spp~sk/research/pdfl?up_rooo_~ooz.pdf(May 16,2005).

'7lbid.' S ~ e n d e l lCox, "International Urbanized Areas: Change in Urbanized Population Density rg60-1gg0,"

The Public Purpose, 2001, at wiclro.demographia.com/db-intlua-areal(May 16,2005).'gGiuliano, Land Use Policy and Transportation, and Randal O'Toole, Transportation Costs and the Americ an

Dream, Reason Foundation, September 2003, at www.rppi.org/pbzj.pdf (May 16,zoo5).20Amile is 1.6 kilometers."Bureau of Transportation Statistics, h'ritional Transportation Statistics 2004; "Passenger Transport Indicators,"

European Environment Agency, Brussels, September 20,2001, a t htfp://themes.eea.eu.int/Sectors~and~activifies/t r a n s p o r t / i n d i c a t o r s / d e m a n d l r E R M x z , ~ ( M a y16,2005); U.S. Bureau of theCensus, Statistical Abstract afthe United States, 2004, at roruw.census.gov/prodhm~~/abs/statab.html(May 16,2005).


9/30

i ?e Social Herwf!:s anu Ccsts ot the Auiomooiie 45

Vehicle Distance Traveled Per Year Per Capita and GDP Per Capitain Selected Countries, 1970-1 99714000

\ i c n . c l e - l ( kne~ ret r a e t edb S,-'12000 i997,.v

,/'10000

,,& /'8000 I. ,. US --- United StatesF - Francei J K ~ - -United Kingdom

6000 Sf - SwedenDK - DenmarkNL -- Netherlands4000

2000 1970

$5 000 $10 000 $15 000 $20 000 $25 000 $30 000Per Capna GDP (1990US$ converted ustng Purchastng Power Pamy)

buy cars and choose suburban lifestyles. Thewealthier they become, the more they drive.

These similar transportation and housingtrends across countries have occurred despitewildly different public policies toward auto-mobiles and land use. Where the U S . has rel-atively low automobile sales and use taxesand decentralized land-use planning, manyother countries have made automobile own-ership and use far more expensive, and havecentrally controlled land-use policies that dis-courage suburban development. For example,while gasoline is selling for about $2.50 pergallon in the US. as of this writing and has

typically sold for much less than this amountduring the past decade, European gasolineprices are far higher-about 34.00 to $5.00 pergallon prior to the price run-up of the last twoyears-because Europe has much highergasoline taxes 22 High gasoline and other taxeson automobile ownership and use make auto-mobiles far more expensive in European andother countries, such as Singapore, but thesehigh prices have not prevented these countriesfrom following similar trajectories in terms ofautomobiles and suburbanization.

The case of Singapore-unlike the U.S., adensely populated country with excellent

"European Communities, Panorm a of Trmsporf,Part 2.


10/30

mass transit-shows how even extremely to bear these large costs even where viablerestrictive automobile policies have only a alternatives are available.modest effect on automobile ownership. In Despite the costs of owning and operatingresponse to a rapid rise in automobile own- an automobile, people choose automobilesership during the 1980s, Singapore in 1990 the world over because no other form ofintroduced a quota on the number of new transportation comes anywhere close to pro-cars allowed in the country each year 23 The viding comparable speed, flexibility, privacy,

Despite the costs of owning and operating an automobile, peoplechoose automobiles the world over because no other form oftransportation comes anywhere close to providing comparablespeed, flexibility, privacy, and convenience.

system was implemented by auctioning offpermits, which sold for about U S. $18,oooin 1997. Researchers estimate that this quotasystem, which increased the cost of purchas-ing a new car by more than 60 percent,reduced car ownership by only 7 percent to11percent. Based on 1997 data, even with-out the permit system, taxes and fees totaledabout zoo percent of the retail purchaseprice of a new car, so the ownership quotaadded to already high costs for purchasinga car. Singapore also imposes congestioncharges to discourage driving during peakperiods.

Thus, even extremely high taxes have a rel-atively small effect on vehicle ownership.People must perceive enormous and uniquebenefits from automobile travel to be willing

-~ ..

and convenience. Even in congested centralcities, automobile travel is much faster thanpublic transport. In the U.S., when compar-ing automobile and transit commutes withinthe same metropolitan areas, the averagetransit commute takes about 75 percentlonger than the average automobile com-mute, even though transit and automobilecommutes cover roughly the same distanceon a~e rage .~"

Automobile travel is much faster in Europeas well. For example, the average travel speedin French cities is 80 percent faster by privateautomobile when compared with tran~it.~5Transit is slower largely because of time spentgetting to and from stations, waiting for vehi-cles to arrive, stopping at stations, and trans-ferring between routes, which together add

'3Data on Singapore are for 1997and come from Giuiiano, Land Use Policy and Transportation.%ommute times by metropolitan area and mode were calculated by Wendell Cox from 2000 US. Census

data. See Urban Transport Fact Book, "US. Metropolitan Area Journey to Work Time: Transit and Other, 2000,"at wzu.publ icpurpose . cm/u t - j twzooomsut im (May 16,2005).National average commute time is from US.Department of Transportation, Summary of Travel Trends, 2001 National Household Travel Survey, 2004, athttp://nhts.ornl.gov/zoor/pub/S~.pdf(May16,2005).

'jR6my Prud'homme and Chang-Woon Lee, Size, Sprawl, Speed and the Eflciency of Cities, Observatoire del'~conomieet des Institutions Locales, Paris, November 1998, at http://mapagee.,~oosjr/'yrudhom/SSS,doc(May16,2005).


11/30

an average fixed time cost of about 16 to 20minutes to a typical transit trip.26

Given the speed differential between carsand transit, commuting by car providesaccess to more than three times the land areain a given amount of time.'j The speed differ-ential is even greater during off-peak travelperiods because driving becomes faster whiletransit service is either the same or slower.

Public transit also provides access onlyalong transit corridors, f~~rtherreducing thearea accessible by transit relative to cars. Forexample, if only one-half of households andemployers are practically accessible by transit,the automobile's speed and accessibilityadvantage would combine to put 12 times asmany employers w i t h i reach, when comparedwith transit, for any given commuting time.28The real transit-accessible fraction is probablylower than one-half, so this is likely an under-estimate of the automobile's advantage in pro-viding greater job access. Automobiles put farmore housing, work, shopping, entertainment,and social choices and opportunities withinreach than do other modes of transportation.

Research on employment. supports the con-clusion that access to an automobile increasesjob opportunities.A recent shtdy of people in awelfare-to-work program in Ltrs Angeles con-cluded that owning an automobile increasedby 12 percent the likelihood that participantswould land a job.'9 A study of people withouta high school degree in Portland, Oregon, con-cluded that car ownership was associated withan 80 percent increase in the likelihood of beingemployed and a $275 increase in average week-ly wages.3' Researchers from the University ofCalifornia estimate that boosting minority carownership rates to the same level as whiteswould eliminate 45 percent of the black-whiteemployment gap and 17 percent of the Lati-no-white employment gap.jl

Policymakers have focused on transit as theway to increase job access for low-income peo-ple, but as a recent study concluded, transitdoes not serve the needs of poor households.Rather, automobile-based solutions will "allow[welfare] participants to be both independentand to optimize their work force, educational,personal, familial, and social opport~nities."3~

26~dw ardL. Glaeser and Matthew E. Kahn, Sprawl and Urban Growth, working paper, National Bureau ofEconomic Research, May 2003, at ruww.tricc.urg/docs/delivrry.pdf(May 16, 2 ~ ~ 5 ) .

'TThe land area accessible within a given amount of time scales as the square of the ratio of vehicle speeds.If cars are 75 percent faster, then a car provides access to three times the land area (1.75 x 1.75= 3.06).

28 Calculate this as follows: Transit can cover one-third the land area based on speed. Furthermore, assumethat only one-half of all homes have access to transit and that only one-half of all employers are accessible bytransit. 113 * 112 1 / 2 = 1/12; that is, one-twelfth as many potential employers are within reach by transitwhen compared with the automobile.

'91Jaul M . Ong, Car Ozunership and We/jiare fo Wo rk, University of California Transportation Center, Berkeley,February 26, 2001, at www.nctc.net/papers/iqu.pdf(May16,2005).

i0Kerri Sullivan, Transportntion C. Wo rk: Exploring Car Usage and En ~p loy nie nfOu t c on~c sin the LSAL D ata,Portland State University, Portland, Oregon, 2003, at http://web.pdx.edu/-jdill/r.SALLSulliz~a~i.pdf(May 16,2005).

3'Steven Raphael and Michael Stoll, Can Boosting M inority Cnr-Own ership Rates Narrow Inter-Racial E~ npl oy-merit Cups? University of California, Berkeley, June 2000, at 7 U 7 U U J . b r ~ ~ k i ~ ~ S . ~ d ~ / ~ ~ / ~ ~ e n t s l b ~ohraph-stoll.pdf(May 16,2005).

i2Alissa D. Gardenhire and M. William Sermons, "Understanding Automobile Ownership Behavior ofLow-Income Households: How Behavioral Differences May Influence Transportation Policy," in PrrsonalTravel: The Long and Short of It, Transportation Research Board Transportation Research Circular E-Co26, June28-July I, 1999, at http://gulliz~er.trb.org/p~iblicatioizs/circii!nrs/eco~6/o~~pnr~nhire.pdf(May 16,2005).


12/30

Walking and transit are insufficient to sat-isfy the full range of people's needs anddesires in work, shopping, socializing, andentertainment. As a result, even large increas-es in residential density have only a smalleffect on driving per capita. For example, datafor the US.indicate that, all else equal, dou-bling residential density reduces per capitamiles of driving by only about l o percent.33

Experience around the world thus suggeststhat suburban, automobile-based lifestyles arenot just an American cultural or policy phe-nomenon, but are also the result of deep-seat-ed human aspirations that transcend cultureand public policy. If so, rather than bemoan-ing the way people want to live, public policyshould be geared toward helping people reapthe benefits of their preferred lifestyles whileminimizing the undesirable side effects. Auto-mobile travel offers something that people theworld over find compelling and far superiorto any other option available.Net Costs or N et Benefits?

Critics of the automobile and suburbanlifestyles claim that the popularity of automo-biles and suburbs results not from people'sgenuine preferences in a free market, but fromgovernment policies that subsidize automo-biles and roads, and therefore suburbs, at theexpense of transit and central cities.'"n reali-ty, as shown above, people choose automo-bile travel worldwide, regardless of govern-ment policies for or against cars. Policies thatdiscourage automobile travel work againstpeople's preferences and have only a margin--. ---- -.-

al impact on automobile ownership and use.People choose automobiles even though theyare costly to own and operate-suggestingthat people the world over perceive enor-mous net benefits from automobile travel.

Could the automobile's critics be correctthat reliance on the automobile for most travelis actually making people worse off overall?Given the global preferences for the automo-bile, this charge seems absurd on its face. Partof the problem may be that we tend to take thebenefits of automobile travel for granted whilethe costs, such as air pollution, congestion, andaccidents, are more salient. Thus, a more hard-headed look at all the costs and benefits ofautomobiles and roads is worthwhile to seehow our best quantitative estimates comparewith people's actual behavior.

For several reasons, estimating the costsand benefits of automobile travel turns out tobe a difficult enterprise.

First, some of the costs are non-monetary-for example, the health damage caused by airpollution or the time motorists spend driving-and therefore must be estimated indirectly.

Second, many costs related to automobileuse are "bundled" into the cost of other goodsand services. "Free" parking at a shoppingmall is an example. Parking facilities costmoney to build and maintain, and there isalso a cost of devoting land to parking thatcould otherwise have been put to other pro-ductive uses. Motorists do not pay any ofthese costs directly; instead, they end up pay-ing them implicitly through the cost of prod-ucts purchased at the mall.-

j3Don Pickrell and Paul Schmiek, Trends in Personal Motor Vehicle Ow ner shi p and Use: Evidence from theNaiaiionruide Personal Transportation S ur vey, Volpe Center, U.S. Department of Transportation, Cambridge, Mass-achusetts, April 23, 1998,at http://npts.ornl.gov/npts/sgg~/Doc/Envecon.pdf(May16,2005).

%ee, for example, Richard J. Jackson, M.D., M.P.H., and Chris Kochtitzky, M.S.P., Cr at in g a Healthy Envi-ronment: Th e Impact oj the Built Envir onment on Public Health, Sprawlwatch Clearinghouse/Centers for DiseaseControl, Washington, D.C., 2002, a t w?ow.sprawlwatch.org/health.pdf(May 16, zooj), and MacKenzie, Dower,and Chen, The Going Rate: What It Really Costs to Drive.


13/30

Tizird, it is not clear whether some costsrelated to automobiles should be included inthe cost tally. For example, many employersprovide free or subsidized parking to theiremployees. Should this driving-related costbe included in a tally of the costs and benefitsof automobiles, or is it-like, for example,employer-paid health care, or employer-sub-sidized office cafeterias or transit passes-justanother employee benefit or business cost?

Fourth, for reasons that will be discussedbelow, the benefits of automobile travel aredifficult to quantify precisely.

With these issues and concerns in mind, therest of this section evaluates the costs andbenefits of automobile travel. As we shall see,estimates in the research literature supportwhat worldwide experience seems to tell us:The benefits of automobile travel far out-weigh the costs.

Benefits of Automobile Travel. At a mini-mum, we can assume that the benefits ofautomobile travel are at least as large as thetotal amount that consumers willingly spendeach year to travel by automobile. People sim-ply would not spend the thousands of dollarsthey spend each year on automobile owner-ship and use if they did not believe that thebenefits of automobile travel at least equaledthe direct costs of obtaining those benefits.But the full social benefits of automobile trav-el are much Larger than this.

The previous section showed that owningan automobile increases poor people's likeli-hood of landing a job. The speed and flexibil-ity of driving also provides access to betterhousing and a greater choice of both housesand neighborhoods.35 Large supermarketsand warehouse stores could not exist without

automobiles, and these stores provide accessto a greater variety of consumer goods andlower prices for those goods. While manyconsumers may lament the demise of the cor-ner grocery or the local department store, it isconsumers who have voted with their feet-and their cars-for greater variety, lowerprices, and the ability to transport severalshopping bags' worth of purchases home in asingle trip.

Automobiles also increase social andrecreational opportunities, putting withinreach friends, relatives, and parks that wereformerly too far away. And the automobilecreates greater quality-of-life competitionamong cities by allowing people to move toplaces that most closely fit their particularinterests.

The automobile has greatly increasedAmerica's economic productivity by givingworkers a greater choice of accessible jobs andemployers a greater choice of workers. Theresult is better matches between people'sskills and employers' needs. The road net-work and motor vehicles have also increasedAmericans' safety by providing for rapidresponses to medical emergencies.

Not only are Americans far more mobilethan they were before the advent of the auto-mobile, but that mobility is much more even-ly spread throughout society. Only thewealthy rode inter-city trains in 1900. Mosturbanites did not regularly ride streetcars andrarely traveled more than several miles fromhome. Today, more than go percent of house-holds own at least one car, and whether theircar is a new Mercedes or an old Geo Metro,both provide fast, flexible travel, and driversof both cars "deal on equal terms for space on

.-

35This qualitative discussion of automobile benefits is summarized from Randal O'Toole, The VanishingAuto~nobilea t x i Other Urban Myths (Bandon, Ore.: Thoreau Institute, 2001), pp. 85-88, and from e-mail discus-sions with Randal O'Toole.


14/30

the highway, parking at the shopping mall,and fuel in their tanks."'6

Transit advocates imagine there was a pre-automobile Golden Age where anyone couldgo anywhere at a time of their choosing. ThisGolden Age never existed when transit wasthe only travel option. However, automobiletravel, while not an unmitigated blessing (seebelow), has provided for the vast majority ofAmericans a level of mobility and opportuni-ty that is unparalleled in human history.

The concrete benefits of automobile travelcan be quantified by looking at how muchpeople appear to value automobile travel in

example, in a world with no vehicle travel, theability to travel a few hundred miles per yearby car would be enormously valuable to mostpeople.On the other hand, those already driv-ing 10,000 miles per year might not find a fewmore miles of driving particularly valuable.After all, they would have used the first ~o,ooomiles of travel to make trips that they consid-ered more important than the trip they wouldmake with the next few miles of travel.

The diierence between the price consumerspay for a good and the value to them of thatgood is known as the consumers' surplus.~7The total benefits of automobile travel are the

Transit advocates imagine there was a pre-automobile GoldenAge where anyone could go anywhere at a time of their choosing.This Golden Age never existed.

terms of dollars. FIGURE5provides a schemat-ic of the benefits of automobile travel. Theshaded area represents what people spenddirectly on automobile travel, including pur-chase and maintenance of cars, gasoline, tolls,parking, and taxes. This total expenditureequals the cost per mile of travel, multipliedby the total miles traveled.

The curved line is called the demand curve.It represents how much automobile travelpeople would be willing to buy at any givenprice per mile of travel. In other words, thedemand curve represents how much peoplevalue each additional increment of automo-bile travel. The demand curve slopes down-ward toward the right because each addition-al increment of travel would be expected to beless valuable than previous increments. For

sum of out-of-pocket costs and consumers'surylus (the total area under the demand curvein FIGURE5). AS shown in FIGURE5, con-sumers' surplus could represent the largestportion of the total benefits. Uncertainty abouttotal benefits arises because we are unsure howmuch people would value those first few hun-dred miles per year of automobile travel.

The consumers' surptus may seem abstract,but it represents real value that people derivefrom the goods and services they purchase. Tosee this, consider the following thought exper-iment: Eggs cost about lo cents each, butimagine we could somehow arrange for theprice of eggs to be 20 cents each. At that price,people would, on average, buy fewer eggs.Even so, many people might not change theiregg consumption at all, and many eggs would

360'~oole,The Vanishing Automobile, p. 87.37See, for example, Steven E. Landsburg, Price Theoy and Applicatzons (New York:West,2001)


15/30

..... .-,-1-5 .-.... .. . . -Benefits of Automobile Travel = Out-of-Pocket Costs + Consumers' Surplus

Current arnouciof travel ai-Gcost per rn !e

Total Vehicle Miles ofTravelSource: Adapted f r m T bogarty The Benefrrs of Road i-ovel and Jrsnspon (Washingtor.D.C.:Amer!car Petm:eum Institilte.jaruary 1988).

oubt continue to be sold. Clearly, peopled be willing to pay 20 cents for eggs

10 cents are gainingo cents of consumers' surplus valueegg they consume.Now imagine we could arrange for eggs to

cents each. This would no doubt causecontinue to buy at least a

each month, and people who reallyof considerable financial means, might

their egg consumption at all. Peo-

ple who are willing to pay 50 cents for eggsthey can get for l o cents are gaining at least40 cents of consumers' surplus value for eachegg they consume.

In short, the consumers' surplus is real andcan represent the lion's share of the value peo-ple derive from their purchases.

While dozens of studies have assessed thecosts of automobile travel, there appears to beonly one study that has attempted to quanti-fy the full benefits of automobile travel 38 In a1998study, Thomas E Hogarty estimates theaverage cost of automobile travel to be 48

3 ' ~ . Hogarty, The Be~zeJitsofRoad Truvel and Transport (Washington,D.C.: American Petroleum Institute, Jan-


16/30

cents per vehicle-mile.3Tombining this withtotal personal vehicle travel estimated at 2.23trillion miles in 1995, he estimates that con-sumers spent $1.1 trillion out of pocket in1995 for personal vehicle travel.soThis is theportion of benefits represented by the shadedarea in FIGURE5. Once again, these costs arealso counted as benefits, because peoplewould not spend the money they spend forautomobile travel unless they believed theexpenditures provided benefits at least equalto the out-of-pocket costs.+1

In order to estimate the consumers' surplusbenefits, Hogarty reasoned that, because manypeople are willing to pay taxi fares rather thanwalk or take public transit, the per-mile cost ofa taxi ride represents a lower limit on howmuch some people are willing to pay for auto-mobile travel in a case where traveling by pri-vate car is not an option. In the mid-~ggos,anaverage taxi ride cost about $4.30 per mile. Carrentals represent another way to assess how

much people value automobile travel. Basedon car rental costs in the mid-rggos, luxury carrentals in New York City cost about $7.60 permile, given known rental costs and reasonableassumptions about total driving distance.Assuming that people value the first few milesof automobile travel at between $4.30 and$7.60 per mile, Hogarty calculates the con-sumers' surplus benefits of automobile travelto be $4.3 trillion to $7.9 trillion per year.

Adding together out-of-pocket costs andconsumers' surplus gives a total annual bene-fit of $5.4 trillion to $9.0 trillion for personalautomobile travel in 1995. As a further refine-ment, Hogarty includes the inferred value oftime spent traveling to arrive at a total benefitof $5.8 trillion to $9.5 trillion. This may be anunderestimate of the true benefits because theuse of taxi and rental car costs likely under-states how much people would be willing topay rather than give up automobile travelaltogethers' Hogarty also estimates the bene-

39This is likely an overestimate. Based on estimates of actual consumer expenditures on all things related toowning and operating automobiles, the direct cost of driving was about 29 cents per vehicle-mile in 1995. Thisresult was calculated by Wendell Cox, based on data in US. Bureau of Economic Analysis National Accountsof consumer expenditures. Cox's calculation gives the cost per passenger-mile in 2002 dollars. This was con-verted to a cost per vehicle-mile by multiplying by an average vehicle occupancy of 1.6 people, as given bythe 1995 National Personal Transportation Survey. The value was then converted to 1995 dollars using theConsumer Price Index. Hogarty's figure is high because he uses automobile ownership cost estimates fromthe American Automobile Association. The AAA assumes that the cost of a new automobile is amortized overa much shorter period than the actual average life of an automobile, resulting in spuriously high averageautomobile ownership costs. The National Accounts data are more representative because they track actualexpenditures. See ~uww.publicpurpose.com/ut-drz7gz96o.htm;www.pubJicpurpose.com/~lt-drvgl96o.htm;andzuwzu.bea.doc.gov~ea/dn/nipa~ueb/~electtableasp?selected=,V#Szfor Cox's calculations and the National Accountsdata, respectively.

4'Based on the National Accounts data, the actual out-of-pocket cost of automobile travel in 1995 wasabout $650 billion. Accounting for this difference would increase the fraction of benefits attributed to con-sumers' surplus, but it has little effect on the overall benefit estimate. Thus, for simplicity, Hogarty's esti-mates are used in the main text.

4'To see this, consider another example: If Jane buys an apple for 25 cents, she must expect that the applewill provide her with at least 25 cents worth of benefits. If Jane felt the benefits of eating the apple would beless than 25 cents, she would not buy the apple.

4'As Hogarty also notes, another reason his estimate is biased low is that he used the average per-mile costrather than the marginal per-mile cost when calculating the consumers' surplus. The marginal cost is lower,which would make the consumers' surplus larger.


17/30

of road freight transport, which he tallies.1 trillion per year, for a total benefit range

all on-road motor vehicle transport of $6.90.6 trillion per year in 1995.

Costs of Automobile Travel. Automobileof very concrete pri-

public costs such as the cost of auto-

s a cost in terms of the time it takes tother. Some of this

unt of time it takes to get from one placel time costs due to congestion. Driving

o includes the harm from air pollution andspeculative costs such as

potential costs of future human-inducedA portion of US. military

The most detailed and comprehensive esti-f these costs to date was performed byi of the University of Califor-

of reports published inting for all of the costs listed above

d many more, DeLucchi estimates that thel social cost of all on-road motor vehicle

trillion in 1991. This estimate is in 1991lars. Assuming that all of these costs areional to the total amount of vehicle

nd adjusting to 1995 dollars gives anated cost range of $ 2 trillion to $4 tril-

n per year in 1995. Even the high end of thiser-bound estimate of the benefits of motor

s discussed above. If these estimates

are reasonable, we can conclude that the totalbenefits of motor vehicle travel are far greaterthan the total costs.

Subsidies and Externalities.Critics of auto-mobile travel are most concerned aboutexplicit and implicit subsidies that theybelieve are encouraging people to drive morethan they otherwise might." For example,drivers pollute the air without having to takeaccount of the costs this imposes on others.Getting on the freeway during rush hour isalso free, but this causes delay for other peo-ple, who are then forced to spend extra timegetting to their destination. Some portion ofUS. defense expenditures might be attributa-ble to maintaining access to oil in politicallyunstable areas of the world, but these costsare paid out of general tax revenues ratherthan through, say, user fees on petroleumfrom the Middle East.

These subsidies are often referred to as"externalities" because they are not reflectedin, and are therefore external to, the markettransactions involved in driving. Additionalsubsidies include costs of road building andmaintenance that are paid out of general rev-enues rather than gasoline taxes, registrationfees, or other revenues directly related toautomobile use. Some analysts also includeemployer-paid parking at work as a subsidyto the automobile because this benefit is a fed-erally tax-deductible business expense.A detailed discussion of the reality andmagnitude of all claimed subsidies and exter-nalities is beyond the scope of this chapter.Note, however, that some of these costs are,at least arguably, not really subsidies. As dis-cussed earlier, employer-provided parking atwork can be viewed as just another employee

~..... ~ ~~ ~

'3For the text of these reports, see w?c~~~~.its.ucd~z~is.edu/people~act~lty/deluccl/(May17,zoog)."See, for example, MacKenzie, Dower, and Chen, Tke Going Rote: Wlznt It Re d l y Costs tu Drioe,and Sierra

ca's Autos on Welfare," undated, at z~~i~~zu.sii.rraclub.urg/sprazul/a~~ticles/subsidies.asp(May 17,2005).


18/30

benefit, similar to other tax-deductible busi-ness expenses such as employer-providedhealth care or subsidized office cafeterias.Indeed, many employers similarly offeremployees free or reduced-price transit pass-es, which transit advocates normally do notinclude on the ledger of transit costs.

Some studies also consider the portion ofroad maintenance and services not coveredby gas taxes as subsidies to drivers, yet some

and the negative economic effects of thenation's trade deficit in its tally of purportedsubsidies for automobile travel.45

Costs of Alternatives to the Automobile.Debates over the societal effects of the auto-mobile often seem to proceed on the implicitassumption that transit is a cheaper alterna-tive. In fact, transit costs about four times asmuch per passenger-mile when comparedwith the direct costs of driving. In 2002, driv-

Transit costs about four times as much per passenger-mile whencompared with the direct costs of driving.

portion of these expenditures would be nec-essary regardless of whether people usedautomobiles to get around. Even if we couldimagine a world in which people did notdrive, they would nevertheless insist on hav-ing the rapid response to fires, crimes, andmedical emergencies that is uniquely afford-ed by road vehicles. Furthermore, some roadswould be necessary for goods deliveries tosupermarkets and other retail stores. A streetand road network would be necessary to pro-vide all of these services and might reason-ably be funded by property and sales taxes.

While the reality and magnitude of theabove subsidies are open to reasoned debate,some of the automobile subsidy claims madeby anti-automobile activists are not credible.For example, the Sierra Club includes "prop-erty taxes lost from land cleared for freeways"

ing cost about 20 cents per passenger-mile,while transit cost 82 ~ e n t s . 4 ~The estimatedcosts of driving include all direct costs of own-ing and operating automobiles as well as roadbuilding and maintenance, including govem-ment expenditures not covered by gas taxesand other user fees. The estimated costs oftransit include all capital and operating costs.

The costs of automobile use do not includeexternalities, such as air pollution and con-gestion; but based on even the highest costsclaimed by anti-automobile activists, theseexternal costs would add about 23 cents perpassenger-mile, for a total cost of driving of43 cents per passenger-mile.47 Thus, even ifwe add an implausibly huge value for theimplicit subsidies to motorists, the real costof driving would still be well below the realcost of transit.'@

45Sierra Club, "America's Autos on Welfare."46~endel lCox, "US. Cost of Automobiies/SUVs and Public Transport per Passenger Mile from 1960," The

Public Purpose, 2~~14,at wruw.publirpurpose.co~/ut-dr~g~~6o.htrn(May 17,2005).47Calculated by assuming external costs of the automobile equal to $7.50 per gallon of gasoline, an average

fuel economy of 20 miles per gallon, and an average vehicle occupancy of 1.6 people.ransi sit also imposes external costs not included here, such as air pollution from buses, diesel trains, andpower plants that generate electricity for electric trains. Assuming the high-end damage estimates from airpollution used by activists, these costs would add several cents per passenger-mile to the cost of transit.


19/30

These cost comparisons also ignore the factby transit is of lower quality than

of the destinations accessible by auto-le, is available only during certain hours

the day, and does not provide any privacybut also provides less value per dollar.

In short, not only does automobile travel

actually be more expensive.g the Undesirable Side Effects

f Automobile TravelAre the negative aspects of automobileNo--just the opposite. With

w exceptions, the history of automobilery of increasing the bene-

of automobile travel while reducing theAir Pollution: Going Down. Air pollution has

the federal Clean Air Act was adopted.

s and the 1960s" Ozone smog waszed as a problem in Los Angeles

e 194os, driven by rapid growth inbut ozone was already in decline

the mid-1950s as a result of local efforts toso

In reports such as the Public InterestResearch Group's More Highways, Afore Poilu-tion, environmental and land-use activistsclaim that increasing air pollution is aninevitable result of increasing automobileuse.jl The director of Smart Growth America,an anti-suburb, anti-automobile group, con-tends that "[slprawl and higher-emittingSUVs are proliferating faster than technologi-cal fixes can keep up."i2 These claims are mis-taken and are contradicted by real-world dataon ambient air pollution levels and motorvehicle emission trends.

Due to population growth and increasingaffluence, Americans have more than dou-bled their total annual miles of driving dur-ing the past 30 years, yet air pollution has fall-en dramatically. FIGURE6 displays trends indriving and ambient air pollution levels from1975 to 2003. All variables are set to an indexvalue of 1.0 in 1975, and the graph showsannual changes from the 1975 base level. Ifmore driving causes more air pollution, weshould have seen a huge increase in air pollu-tion during the past few decades. But as FIG-URE 6 shows, just the opposite occurred: Asdriving increased, air pollution declined.

While 60 percent of the nation's ozone mon-itors violated EPA's one-hour ozone standardin the late 197os, only about l o percent do sotoday, and average annual days exceeding thestandard have dropped more than 95 per-cent.53 The nation has likewise made greatprogress on airborne particulate matter, or PM(soot, dust, and so forth). Only a few percentof the nation's monitoring locations stilt vio-

4YIndur M. Goklany, Clearing th e Air: The Real Story u fthe Wa r on Air Pollution (T,Vashington,D.C.: Cato Insti-1999).

iO1bid.5' Public Merest Research group, More Highwnys, More Pollution.i2David Coldberg, "If Smog Isn't Routed, It Returns," Atlanta Journal-Constitution, September 1, zoo5 p. Ai j .j3All ambient pollution trend statistics are based on analysis of national air pollution monitoring data down-

EP.4 at wr~~u~.epa.gov/ttn/nirs/airs~~qs/detail~iafa/dow~loadqsdata.fm.For detailed charts showing


20/30

56 21s: Century Highways

1Trend in Total Vehicle Miles vs. Trends in Air Pollution, 1975-2004 iIiIIi

-T Total Vehicle Miles Traveled -x-Annual-Average Nitrogen Dioxide-- Peak 8-hour Carbon Monox~de --I -hour Ozone Exceedances

Note: Pollution data are national averages of all m on itoring stations in the U.S.Nitrogen dioxide is theannual-average level, carbon monoxide is the peak daily 8-hour level, and ozone is the average numbero f days per year exceeding the federal I -hour ozone standard.Sources:Vehicle miles, Bureau ofTrarsportation Statistics,Not~onolTransponation Statistics 2005: Population,Census 6ureau;Air ?oilution. Environmental Protection Agencfs AirData pollution m onitoring database, at~ w . e p o . g o v / t t n l a ~ ~ / a i ~ a q s / d e t o i l d ~ t ~ / d ~ ~ n i o a d a q s d n t o ~ h ~ n(March 15,2004).----- -.------,-------"

late the Environmental Protection Agency'sstandard for particulate matter under l omicrons in diameter (PM"). About 15 percentviolate the EPA's new and much more strin-gent standard for particulate matter up to 2.5microns in diameter ( PM ~ ' ~ ) ,but P M ~ . ~lev-els have steadily declined, dropping nearly 50percent during the past 25 years and 14 per-

cent in just the past five years. During the-tg70s, the EPA initially considered carbonmonoxide to be the worst air pollution prob-lem, but all monitoring locations in the coun-try now attain the federal health standard, andmonitored levels in air continue to decline.

Air pollution continues to decline becausecam--including SUVs and pickup t r u c k h a v e

-.ambient ai r p ol lu t ion t rends, seeJoel Schwartz an d Steven Hay ward, Air Quality in America: A Dose of Reality onAir Pollution Levels, Trends and Health Risks, Am erica n Enterpr ise Inst i tute, Ap r i l 28, 2004, a t www.aei.org/docLib/zoo4o4jo1_schwartzhayward.pdf(May 17,2005).


21/30

en getting cleaner much more quickly thans been increasing. For example, on-

ad trend measurements show that the aver-o percent per year as the fleet turns over

more recent models that start out and stay54 Yet driving is

2 percent per year, for a netns decline of 8 percent per year 55 These

because old high polluters

FIGURE;shows recent trends in automobileemissions (including SUVs and pickup trucks)as measured in a tunnel in the San FranciscoBay Area from 1994 to 2001. Emissions of allthree of the major automobile air pollutants-carbon monoxide, volatile organic com-pounds (VOCs), and nitrogen oxides-aredeclining rapidly. The rates of decline areabout 9 percent, 13 percent, and 15 percent peryear, respectively, for nitrogen oxide, carbon

Old high-polluting cars continue to be scrapped and new cars'emissions and durability continue to improve. The same goes forSUVs and pickup trucks. Data from vehicle emissions inspectionprograms show that SUVs and pickup trucks have had about thesame emissions as cars since at least the 2001 model year.

be scrapped and new cars' emissionsdurability continue to improve. The samefor S W s and pickup trucks as for regular

show that SUVs and pickup trucksd about the same emissions as cars since

2001 model year EPA standardshave the same low emissions as

57 Theobile fleet that will be on the road in 15 to

years will emit at leastgopercent less air pol-on per mile of travel even if we do nothingto reduce automobile emissions.

monoxide, and volatile organic compounds.Emissions are reported here in terms of

grams of pollutant emitted per gallon of fuelburned for the average automobile. This rateof change in emissions per vehicle can be con-verted into a rate of change in total vehicleemissions by taking account of increases ingasoline consumption over time. According tothe California Department of Transportation,gasoline use increased about I .3percent peryear from 1994 to 2001. Thus, for example,while VOC emissions per vehicle declined 15percent per year, total VOC emissions

"%ajal J. Pokharel et al., "Emissions Reductions as a Result of Automo bile Improvem ent,'' Environmental Sci-an d Technolo~y,Vol. 37 (zoo3), p p . 5097-5'01, an d Joel Sch wa rtz, 1Vo bt'ny Back: W h y Air Poiintion Wi ll Con-

n w to Llrcline, American Enterprise Institute, July 2003, at www.oei.orK/ilocLib/~oo@o4_4.pdf(May 17, 2~0 5).5j"Congestion Data for Your City," Texas Transportation hs titu te, a t http:/l,nobility.tninzi.e~iu/tinis/congestion~dat~/

17,2005).S6schwartz, Ko 6 % ~Bnck.57Environmental Protection Agency, "Control of Air Pollution from New Motor Vehicles: Tier 2 Motor

cle Emissions Stand ards a nd Gasoline Sulfur Co ntrol Requirements; Final Rule," F r~femlRegister, Febru-pp. 669&6870.


22/30

[rend in Automobile Emissions Measured in a San Francisco Bay Area Tunnel

1992 1994 1996 1998 2000 2002 2004Calendar Year

Notes:N O x : nitroger ox~de s:\iOC= volatlle organic compou rds: C O = carbon rroro xide. ir ris sio ni were measured ;sgrams o f pciiutanr emirted per gallon o f gasoline burned. Lines through tbe data pon rs are based an iiRing he data r a dnexponenriai furci cn .Source:A. Kean et aI,,"Trends in Exbauit Ernissiors fimm In-Use Califo rnia L,ght-D utyVe hicIes , l99+2OO I:' Society o fAutomotive Eqyeers,Warrenda ie, Pennsylvania. 2CO2, at w+w.uc!cic.net!pcpe:~!i84pdf (May l7.2CO5).

- -""-"-.-.-,---..----.----"-a--- ~ ~ . . --=--

declined a little over 13 percent per year. Thisshows that, contrary to activists' claims,growth in driving and growth in the popular-ity of SUVs-both of which are accounted forby adjusting for increases in gasoline con-sumption-are doing little to offset the bene-fits of inherently cleaner automobiles.

Soot emissions from diesel trucks have

also declined a great deal. On-road measure-ments show that the average truck's sootemissions declined more than 80 percentbetween 1975 and ~ 9 9 9and 50 percentbetween 1997 and 2o04.j8 EPA regulationsrequire a further go percent reduction in sootand other pollutants from diesel trucks start-ing with the zoo7 model year.59 Based on cur-

i s ~ l a nW.Gertleret a/.,Real-World Particulate M atter and Gaseous Enzissionsfrorn Motor Vehicles in a Highway Tun-nel, Research Report, Health Effwts Institute,2~02,at ~~m.heal'Rlt~e~ts.org/pubs/CertGros.pdf(May17,200j);ThomasKirchstetter et al., "Characterization of Particle and Gas Phase Pollutant Emissions from Heavy- and Light-DutyVehicles in a California Roadway Tunnel," American Geophysical Union Fall Meeting, San Francisco, 2004.

j9Environmental Protection Agency, "Regulatory Impact Analysis: Heavy-Duty Engine and Vehicle Stan-


23/30

Although air pollution has greatly improveds still contribute a substantial fraction of

However, a small frac-f all automobiles account for most of thefor example, that the worst ro percent of

c compound emissions, whichand particulate matter air pol-

The worst jpercent of cars contribute0 percent of VOC emissions. These high pol-

in poor repair. In other words, the typ-al car is so clean that driving per se is not a

rce of air pollution. Rather, a fewof' the problem.

The distribution of air pollution emissionsThe most successful motor vehi-

nges in vehicles-new designs and newologies-that have resulted in progres-

ly more effective and more durable auto-ution control systems. Yet much

behavior-getting them to driveby encouraging carpooling, making driv-

ing less convenient by restricting freewayexpansion, or providing alternatives such astransit.

Policymakers and environmental activistshave mistakenly defined the motor vehicle airpollution problem as, "all cars pollute, soreducing driving is an effective way to reduceair pollution." In fact, most cars pollute hard-ly at all, while a few cars pollute a lot. Evenas far back as the early rgSos, researchers real-ized that a small percentage of automobilesaccounted for most em i ~ s i o n s . ~ ~Thus, even ifthey could be effective in principle, policiesaimed at reducing how much people driveare guaranteed to put most of their effort intoreducing vehicle-miles driven by cars thatemit hardly any air pollution. Technology hasgreatly weakened the link between drivingand air pollution and will virtually eliminateit over the next two decades.

Policies that try to affect how m uchpeopkdrive in the name of reducing air pollutionhave been notably unsuccessful in influenc-ing how much people drive, but they havecaused a great deal of collateral damage byincreasing road congestion and divertinghundreds of billions in transportation fundsover the past few decades to transportationmodes that hardly anyone chooses to use.

Safer Cars, Safer Roads. More than 42,000Americans were killed in motor vehicle acci-dents in 2003, and millions were injuredb2Yetthe toll would be far higher without the con-tinuous improvements in road and vehiclesafety implemented over the past fewdecades. Per vehicle-mile of travel, the death

rds and Highway Diesel Fuel Sulfur Control Requirements," December 2000, at w7u;u.epn.gov/otniI/diesei.him"Schwartz, No Way Bock. See also reports on various on-road emissions measurement studies at Fuel Effi-

Test, at ~ ~ ~ . i u u ~ ~ e n t. b i o i ~ ~ e ~ i ~ . if ~ i . ~ ~ ~ i ~ / i i ~p l ~ t ~ d ~ ~ t y ~ ~ ~~ h i c l e s . h t l(May 17, 2005).61Lowell G. Wayne and Yuji Horie, Evaluatio~zo f ARB s In-Use Vehicle Surveillance Program, F i r m 1 Report: California Air Resources Board, October 1983).62Bureau of Transportation Statistics, National Trtmsportntion Stiltistics 2004.


24/30

60 21s: Cectury Highwaysrate for automobile occupants declined 75 vehicle, the driver and/or the pedestrian havepercent between 1960 and 2003, while the rate alcohol in their blood 6 jOne-third of pedestri-for pedestrians declined 84 percent. In other ans killed in car accidents have blood alcoholwords, 40 years ago, the risk of dying in a car levels greater than 0.08 percent, as do 6 percentaccident was four times higher if you were of drivers involved in those accidents. Thus,driving and seven times higher if you were while suburbanization appears at worst towalking. In addition, injuries per vehicle-mile have no effect on automobile risks to pedes-declined 32 percent from 1990 to zoo2 63 For trians, alcohol-impaired judgment on thethe past decade, the risk of injury has been part of pedestrians and/or drivers may be adropping faster than the risk of death. major factor.

Critics of the automobile claim that subur- I t has become fashionable among publicbanization decreases road safety, particularly health experts and advocates to claim that sub-for pedestrians. For example,Mean Streets, an urbs discourage walking and physical activi-annual report by the Surface Transportation ty.66 Could it be that suburbs are safer forPolicy Project (STPP), an anti-automobile pedestrians because suburbanites simplygroup, decries the supposedly great and spend less time walking? Probably not. Theincreasing danger that automobiles pose for Centers for Disease Control and universitypedestrians 64 Yet per vehicle-mile, the fatali- researchers report that suburbanites are actu-ty rate for pedestrians has dropped faster ally more physically active than people inthan the rate for vehicle occupants. Further- cities or rural areas and less likely to be obese f ~ 7more, not only has the rate declined, but the Congestion. Because getting on the road istotal annual number of pedestrians killed has free, congestion is the default system foralso declined 27 percent since 1990 even as rationing road space. However, it should betotal miles of driving increased 31 percent noted that motorists have adjusted to conges-over the same period. tion in their housing choices. Families choose

The STPP blames pedestrian deaths on where to live based on a range of factors. Forurban "sprawl." Yet the extraordinary increas- example, families with children rate access toes in pedestrian safety over the past few good schools near the top."' However, commut-decades occurred at the same time that Ameri- ing time to work must surely be a factor as well.cans chose suburbs by the tens of millions. The Total vehicle-miles driven in urban areas grewSTPP also fails to note that in 45percent of acci- at 2.5 times the rate of growth in urban road-dents in which a pedestrian is killed by a motor capacity during the qgos, yet average cornrnut-... -.....

631990 is the earliest year for which national injury data are available.64~rnst ,Menn Streets 2004: How Far Have WeCome? Pedestriiin Safety, 1994-2003.65U.S.Department of Transportation, National Highway Traffic Safety Administration, I+desiria,r Roadzciay

Fatnlities, April 2 ~ 3 ,at w?~~ru-nud.nhtsa.dot.goz~/p~~/~1rd-~o~CSA/Rpts/~oo/o-j6.pdf(May 17, 2005).66~ee ,for example, Ewing, Pendall, and Chen, Measuring Sprawl and l f sZnrpnct, and Jackson and Kochtitzky,

Creating a Healthy Environment.6 7 ~ .S. House rt a!., "Excess Mortality Among Urban Residents: How Much, for Whom, and Why?" Arrieri-

cnn journnl of Public Ilr~zl th,Vol. 90 (zooo), pp. 1898-1904; Centers for Disease Control, National Center forE-iealth Statistics, Health United States, 2001. at ~ ~ 7 ~ i u . ~ ~ i ~ . g o ~ / ~ l c h ~ / d ~ t a A 1 ~ i ~ / 1 ~ ~ 0 l . p d f(May 17,2005).

68~e t e rGordon and Harry Richardson, "Critiquing Sprawl's Critics," Cato Institute Policy Anuly sis No.365, January 24,2000, at zuwzo-rcf.~rsc.edu/-p~01do11/piff/pa~6j.pdf( M a r7,2m5).


25/30

:: time rose only a few minutes between 1990d zoo~ .~YPeople adjusted their housing and

g times relative-constant even as congestion increased. Jobs

people to the suburbs, facilitating thef relatively short commute timesf increasing road congestion.7'

Emen surveyed, a majority of people say theybuilding more transit to reduce conges-

ty of new transit facili-

I t is not clear that such a large increase intransit use could even be accomplished. From1990 to rooo, inflation-adjusted spending ontransit increased 29 percent and transit rider-ship increased 14 percent, but driving increasedfaster, resulting in an 11percent decline in tran-sit's share of work-commute trips 7' Transit'sshare of all transportation is at an all-time low.Only 1.6percent of all person-trips and 1.2 per-cent of all person-miles are traveled via public

Per vehicle-mile, the fatality rate for pedestrians has dropped fasterthan the rate for vehicle occupants. Not only has the rate declined,but, since 1990,the total annual number of pedestrians killed hasalso declined even as total miles of driving have increased.

s to reduce congestion is limited because tran-nts for only about 4.6 percent of all

congestion is at its worst. Even apercent increase in transit's share of commuted still leave transit accounting for onlypercent of ail commute hips. A recent study

a jo perrent increase in transit's22 seconds, or about 1.5 percentT1

transit 73 Transit's decline occurred despite $360billion in taxpayer subsidies for transit betweenthe mid-1960s and the late 199os7%ile auto-mobiles also receive taxpayer subsidies, transitreceives about 57times as much in subsidies perpassenger-mile as automobiles receive75 Publictransit received more than 15 percent of all taxdollars spent on transportation between 1977and 1995 despite accounting for less than 2 per-cent of actual travel during this period 76

('91bid.; L.S.Department of Transportation, Summary of Travel Trends, 2001 National Household Travel Survey7"Randall Crane and Daniel G. Chatman, "Traffic and Sprawl: Evidence from US . Commuting, 1985 to

Planning nnri Markets, Vol. 6 zoo^), pp. 14-22; Gordon and Richardson, "Critiquing Sprawl's Critics.""Wendell Cox and Randal O'Toole, "The Contribution of Highways and Transit to Congestion Relief: ABackgrounder 1721, January 27, 2004, at wzuw.heritnge.urg/Researcl.~/

7' lbid.73U.S. Department of Transportation, Strtnniary of Trnzrel Trends, 2001 National Household Travel Survey

74Gordon and Richardson, "Critiquing Sprawl's Critics."75Wendell Cox, "Highway and Mass Transit Subsidies in the United States: 2002,'' The Public Purpose,

zuzc;w.yublicpurposr.com/trt-mfhiuy$zooz.ht(May 17,2oo5).761'ietro S. Nivola, Laws $the Landscape:How Policies Sl~npeCities in Ew ope and America (Washington, D.C.:


26/30

Although most people like the idea of build-ing transit, it is clear that the vast majority willnever actually use these transit services them-selves. The Onion, a satirical weekly newspa-per, accurately captured people's real views ofpublic transit with the tongue-in-cheek head-line, "98 Percent of U.S. Commuters FavorPublic Transportation for Others."77

Road congestion has worsened largelybecause policymakers, supported by anti-automobile activists, have prevented roadinfrastructure from expanding to meet newdemand for automobile travel. Between 1980and 2003, total miles driven in the UnitedStates increased 89 percent, but total urbanroadway lane-miles increased only 51 per-cent7* Indeed, a key goal of many urban plan-ners and activists is to make driving slowerand less convenient in order to encourage peo-

ple to switch to other modes of travel79 Forexample, the regional planning authorities inPortland, Oregon, and Minneapolis, ~Minneso-ta, seek to increase congestion in order tomake driving less attractive and to increasethe use of alternative transportation modesOther than deciding to live with road con-gestion, there are probably only two approach-es that have a decent chance of making a dentin the problem: Provide more road capacityand/or charge for road space during peak trav-el hours. Metropolitan areas with more high-way and arterial lane-miles per capita also havelower traffic densities, so more road space perperson is associated with less congestion8' Butexisting road space also is not used efficiently.Congestion charges can encourage people tomove discretionary travel away from peaktravel

nUReport:98 Percent of US. Commuters Favor Public Transportation for Others," The Onion, November29,2000, at www.theonion.com/auto/news-3643.php(May r7,2005).

78"Urban" means all non-rural roadways. The change in urban roadway lane-miles includes all freeways,arterials, collectors, and local streets. Freeway and arterial lane-miles alone increased 48 percent. Bureau ofTransportation Statistics,National Transportation Statistics 2004, Table 1-13,

79For general examples, see Steve Inskeep, "Commuting IV," A11 Things Considered,National Public Radio,May 30, 1997, at ~~~w.npr.org/templateslStory/stoy.php?stoyld=10~943~,and O'Toole, The Vanishing Automo-bile, pp . 260-261. See also Dom Nozzi, "Traffic Congestion: Friend or Foe?" Walkable Streets, undated, atwww.walkablestreets.rom/congest.htm("It is a serious strategic blunder for sprawl-busters and other communi-ty and environmental advocates to oppose traffic congestion"); "Fewer Marylanders Using Mass Transit,"Montgomery Journal,June 4,2002 (IIMy hope is that it will get so congested that people will want to invest inpublic transit," Marie Howland, Professor of Urban Studies, University of Maryland); "The Automobile andthe Environment: MIT Conference Addresses the Costs of Congestion and Looks for Solutions," MIT Centerfor Environmental Initiatives Newsletter, June 1999, p. 7, at ht@://lfee.mit.edu/public/CEl-Newslette~-VzN~,pdf("Mr. Michael Replogle of the Environmental Defense Fund argued that from an environmental point of view,reducing traffic congestion can be counterproductive"); and MichaelA. Replogle, minority Statement ofMichael A. Replogle" in Transportation Research Board, Expanding Metropolitan Highways: lmplicationsfor AirQuality and Energy Use (Washington, D.C.: National Academies Press, 1995), p. 358, at www.nap.edu/books/0309061o7$1tml("Limiting further highway capacity expansion, reducing highway capacity, and calmingtraffic (especially in central areas) can be effectivestrategies for reducing energy use, air pollution, and otherenvironmental problems").

%ktro, Regional Transportation Plan Update (Portland, Ore.: March 1996), and Metropolitan Council,Transportation Policy Plan (St. Paul, Minn.: 1996).8'M1endell Cox, "Traffic Volumes & Highway Capacity," The Public Purpose, 2000, at www.publicpurpose.comhwy-chartsggJlesfiame.htm(May 17,2005)."For a detailed discussion of how urban areas can deal effectively with road congestion, see Chapter 4.


27/30

Lt'hile increasing congestion might be instrument for dealing with various subsidiesand externalities related to driving depends

riving experience. on the nature of the cost. For example, roadhave added progressively more congestion would be addressed most by

to cars over time: for example, more charging for road space during periods ofortable seating, more spacious passenger peak usage; air pollution, by levying pollu-ns, better stereos, and better soundproof- tion charges on motorists based on their vehi-

Drivers can listen to books on tape or CD, while cell phones andheadsets allow them to conduct business or talk to family on theroad. Many restaurants cater to morning commuters by providingdrive-through windows. These innovations give the auto an advan-tage over mass transit, which does not cater to individual tastes.

who want intellectual stimulationcollege courses on tape or

l phones and headsets allow com-rs to conduct business or talk to friends ora n d cafes cater to morning commuters by

s on the same side of the street as the morn-sting time making a left turn.'3 All of these

ol over their commutingce, giving the personal automobile an

mass transit, whichoffer privacy and does not have the

xibility to cater to individual tastes.Using Pricing to Mitigate the Negative Side

The appropriate policy

cles' actual emissions; carbon dioxide emis-sions, by taxing gasoline; and accidents, bycharging more to the riskiest drivers. Howev-er, realizing that gasoline taxes are adminis-tratively simple and comparatively well-accepted, two researchers from Resources forthe Future (RFF), an environmental and ener-gy think tank, recently set out to estimate theoptimal level of gasoline taxation that wouldaccount for all of the estimated external costsimposed by drivers on society at large $4 Theyconcluded that the U.S. gasoline tax shouldbe about $1per gallon to account for thesecosts, or about 60 cents per gallon more thancurrent US. gasoline taxes.

The largest components of this tax are dueto congestion (29 cents per gallon) and acci-dents (24 cents per gallon), followed by region-al air pollution (14 cents per gallon), with theassumed climate change impacts of carbon

8 3 ~ .Shaver, "Pursuit of a C r a n d e L a t t e Ma y Be St i r r ing up Gridlock," The W nshington Post , Apri l 18,'4lan W. H. Parry an d Ken A. Small, "Does Britain o r the U nited States Hav e the R ight Gaso line Tax?"

Rrvicw, forthcoming 2005.


28/30

dioxide emissions adding 5 cents per gallon. Atcurrent rates of gasoline consumption, adding60 cents per gallon to the gasoline tax wouldadd about $79billion to the total annual cost ofdriving. This works out to about 3 cents pervehicle-mile, 2 cents per passenger-mile, and$330 per year for the average auto m~bil e.~ j

Although the reality and magnitude of thesecosts are uncertain and matters of both scientif-ic and political controversy, charging driversfor real costs they may be imposing makessense. But even assuming that the RFF analysiscorrectly estimates these costs, charging driv-ers for these costs would have only a smalleffect on how much people actually drive. Thedirect costs of driving to U S . motorists areabout $850 billion per year. An additional $80billion, white obviously not a trivial increase, istoo small to have a noticeable impact on theoverall amount of driving, especially given thatAmericans likelv derive trillions of dollars peryear in net benefits from driving.

In any case, as the RFF analysis itself notes,much of the external cost imposed by auto-mobiles is due not to total miles driven per se,but to driving at certain times of the day (con-gestion), to a few high-polluting cars (air pol-lution), or to particularly risky drivers (acci-dents). Only greenhouse gas emissions close-ly track total miles driven. Thus, more target-ed approaches would address many of thecosts imposed by driving without substantial-ly affecting total miles driven.

TheRFF analysis is based on pollution emis-sions estimated for 2000,but whatever the costsof automobile air pollution were in 2000, theyare substantially lower now because automo-bile per-mile emissions have been dropping byabout lo percent per year. The air pollutioncosts of driving will continue to declineH6

It is also worth noting that estimates ofsubsidies to drivers ignore the fact that someof the costs drivers pay might actually be toohigh. For example, researchers from RANDestimated that changes in public policiesregarding automobile liability insurancecould cause substantial reductions in the costof auto insurance by reducing the costs of per-sonal injury litigation and that the savingscould be achieved without reducing compen-sation for economic losses to injured driv-er~.~7Accordingto one estimate, the total sav-ings could be as high as $50 billion per year.88

While we have not eliminated all of theundesirable side effects of driving, we havemade great strides in making driving safer,less polluting, and more comfortable. A rangeof untapped policy approaches could furtherincrease the net benefits people derive fromautomobile travel.

Critics of the automobile and suburbanlifestyles believe that charging motorists thefull costs that driving imposes on societywould decrease driving and increase transituse. What they fail to appreciate is that apply-ing this principle across all transportation

s5~alculatedbased on national data on gasoline consumption and total vehicle-miles traveled by automo-biles, and assuming that the average automobile gets 20 miles to the gallon and is driven about i1,ooo milesper year. The actual increase in costs would be somewhat smaller than this because the extca tax would causea small decrease in gasoline consumption through a combination of greater fuel economy and fewer milesdriven. Bureau of Transportation Statistics, iVRtiona1 Transportation Sta tistic s 2004.

86~okharetef al., "Emissions Reductionsas a Result of Automobile Improvement,"and Schwartz, No W ay Back.'?stephen J. Carroll and Alan F. Abrahamse, The Effects of a Choice Automobile Insurance Plan on Insurance

Costs and C ompensation: A n An alysis Based o n 1997 Data (Santa Monica, Cal.: RAND, ~ y y y ) ,a t roww.rand.org/publications/MR/MRr134/ (May 17,2005).

88~ivola ,Laws of the Landscape.


29/30

iiicrcnse driving and prefer to do away with air pollution and cartransit use because public transit is accidents, and although we all like the free-

and more heavily subsi- dom to drive where we want when we want,n driving." Transit costs about 82 most of us would prefer that other people not

52 cents, or 64 use that freedom to drive down our streets or

Most people's needs, desires, goals, and plans are too diverse andever-changing to satisfy with travel mainly by foot and transit.

f this cost is paid through taxpayerthese subsidies would raise fares to

by nearly a factor of three. On ther hand, adding both direct and implicit

o motorists would add about 13 per-enger-mile cost of driving,

a total driving cost of about 23cents per pas-le."' Even if one could justify impos-

stantially larger costs on motorists, driv-still remain far cheaper than transit.

Clearly, increasing the costs of automobileress explicit and implicit subsi-little effect on the dominance of

the automobile. This is confirmed by experi-imposing costs on

vers well in excess of any harms caused byayed Europeans' adop-

f American-style transportation choices.

What Is the Alternative?Americans have chosen automobile-basedrban lifestyles, but driving, as with most

things in life, involves trade-offs.We all would- ~~~,......, , ..... ~.~.......~~.

congest our roads. Over the years, we havefigured out ways to progressively mitigatemany of the negative effects of driving whilemaintaining or augmenting the benefits.

To be sure, there is still plenty of room forimprovement; but to listen to the automobile'scritics, you would think not only that drivingis making people worse off overall, but alsothat we are irrationally ignoring an alternativetransportation and land-use Nirvana that issitting there, ripe for adoption-a worldwhere work, shopping, school, restaurants,piano lessons, soccer practice, cafes, and moreare all just a short walk or transit ride away.

In reality, most people's needs, desires,goals, and plans are too diverse and ever-changing to satisfy with travel mainly by footand transit, while most people's lifestyle pref-erences are incompatible with the residentialdensities necessary to make walking and tran-sit practical. As one researcher recently put it:

[Olne can imagine how nice it would beto be able to walk to local shops instead

'92.lark DeL ucchi, "Shcrnld We Try to G et the Prices Iiight? " Access, Vol. 16 (m oo ), p p . 10-14, at ;oi~w. i t s .u~-~faois.cdu/pirhiicnti~~iishooo.:I


30/30

of driving to Wal-Mart, but the reasonpeople drive to Wal-Mart is that theselection is better and the prices arelower. One can imagine how nice itwould be to walk to the corner groceryrather than drive to the supermarket,but again, supermarkets put the cornergroceries out of business because theyhave tower prices and a better selec-

national average. Furthermore, most peopleprefer to have more living space and moreprivacy than high-density living can provide.

There is no realistic alternative to automo-bile-based transportation that would notrequire large reductions in people's autonomy,prosperity, and quality of life. Around theworld, and in a wide range of cultural, econom-ic, and policy environments, people adopt auto-

There is no realistic alternative to automobile-based transporta-tion that would not require large reductions in people's autono-my, prosperity, and quality of life. People adopt automobile-basedtravel and suburban living when they become wealthy enough toafford them.

tion.. ..Bicycles seem like a good alter-native to driving, unless the weather istoo hot or too cold, or it is raining.9'What is the alternative to the automobile

and the suburban home? Nobody likes roadcongestion, but travel by transit is far slowerand offers a far more restricted range of desti-nations. Higher residential densities mightincrease opportunities to walk and wouldmake transit more viable, but trip times byboth transit and automobile are slowest incities with the highest densities and the great-est transit use. The New York metropolitanarea has the longest average commute time inthe U.S.gZMeanwhile, sprawling metropoli-tan areas such as Phoenix, Las Vegas, Denver,and Austin have commute times around the

mobile-based travel and suburban living whenthey become wealthy enough to afford them.The dominance of automobile-based travel isnot just an American phenomenon, and is notthe result of public policies favoring the auto-mobile, but is instead the manifestation of deep-seated human desires for opportunity, privacy,space, convenience, and autonomy.

Nevertheless, for the past few decades, pol-icymakers and activ

Schwartz_Automobile

Documents

Transcript of Schwartz_Automobile