ML11356A489

Hudson River Sloop Clearwater (Cle) Pre-Filed Evidentiary Hearing Exhibit CLE000022, Socioeconomic Differences in Household Automobile Ownership Rates: Implications for Evacuation Policy

ML11356A489
Person / Time
Site:	Indian Point
Issue date:	12/22/2011
From:	Berube A, Deakin E, Raphael S The Brookings Institute, Univ of California - Berkeley
To:	NRC/SECY
SECY RAS
Shared Package
ML11356A487	List: ML11356A487 ML11356A488 ML11356A489 ML11356A490 ML11356A491 ML11356A492 ML11356A493
References
RAS 21633, 50-247-LR, 50-286-LR, ASLBP 07-858-03-LR-BD01
Download: ML11356A489 (47)

Text

Socioecon omic Differences in Ho usehold Automobile Ownership Rates: Im plic a tions for Evacuation Policy Alan Berube The Brookings Institution E-mail: ABERUBE@brookings.ed uElizabeth Deakin Depart m ent of City and Regional Planning University o f California

, Berkeley E-mail: e dea k i n@ber keley.e duSteven Rap hael Gol d m a n School o f Public Policy University o f California

, Berkeley E-mail: ste venraphael@berkeley.e d u June 2006 Abstrac t The devastation wrought by hurricane Kat rina laid b are many o f the d ispariti e s t h a t continue to separate Americans by race and cla ss. One disparit y that was i m m ediately apparent in Katrina's after math concerned the size and c omposition o f t h e area's populations that lacked access to a n auto mobile. The s e hou sehold s , largely d ependent on the li mited e mergency publ i c transp o rtati o n available to evacuate the city i n advance o f t h e st o r m , were the m ost likely to b e left behind. In New Orleans, this population see m ed quite large in size - and overwhel m ingl y black. In this paper, we doc u ment difference s in car-ownership rates betwee n racial and socioecono m i c groups. W e present patterns for the nation as a whole as well as for the pre-Katrina New Orleans metropolita n area usin g data fro m the 2000 5% Public Use Micro data S a mple (PUMS) o f the U.S. Cen s u s o f Populat ion and Ho using. W e also present esti mates of the nu mber of people for all U.S. metrop olitan a reas that reside in a household with o u t access t o an auto m obile. F i nally, we explore the relati onship between residentia l housing segregation and spatial proxi mity to other hou sehold s without acce ss to automobile s a m o n g African-A m erican s. Exhibit CLE000022 Submitted 12/22/11

21. Introduction The devastation wrought by hurricane Katrina has laid bare many of the disparities that continue to separate Americans by race and class. One disparity that was immediately apparent in Katrina's aftermath concerned the size and composition of the area's population that lacked access to an automobile. These households, largely dependent on the limited emergency public transportation available to evacuate the city in advance of the storm, were those most likely to be left behind. In New Orleans, this population seemed quite large in size - and overwhelmingly black and poor. In retrospect, the size and demography of the population stranded by Katrina is not particularly surprising. Nationwide household access to automobiles is quite low for African-Americans (especially poor African-Americans), with a disparity relative to white households that is somewhat smaller than, yet comparable in magnitude to, the racial disparity in home ownership rates. Moreover, New Orleans was a predominantly black city, with an overall poverty rate that exceeded by far that for the nation. The evacuation failure in New Orleans was compounded by the racial segregation of the New Orleans metropolitan area. Given that most black households resided in predominantly black areas, black households without cars were physically surrounded by neighbors without cars, rendering an evacuation strategy based on private transportation particularly ineffective. This is not to suggest, however, that the outcome we observed in Ne w Orleans is one that would not be reproduced elsewhere should such a catastrophe hit another major American metropolitan area. New Orleans is a mid-size American city with a population that happens to be disproportionately African-American. In many larger U.S. cities, the black population is 3comparable in magnitude, highly residentially segregated, disproportionately poor, and has low car-ownership rates (the precise traits that rende red New Orleans blacks particularly vulnerable). In this paper, we document differences in car-ownership rates between racial and socioeconomic groups. We present patterns for the nation as a whole as well as for the pre-Katrina New Orleans metropolitan area using data from the 2000 5% Public Use Microdata Sample (PUMS) of the U.S. Census of Population and Housing. We also present estimates of the number of people for all U.S. metropolitan areas that reside in a household without access to an automobile. Finally, we explore the relati onship between residentia l housing segregation and spatial proximity to other households without access to automobiles among African-Americans.

The results of the analysis suggest that there are sizable populations without access to household automobiles in most U.S. metropolitan areas. Similar to New Orleans, the members of these households are disproportionately poor and minority. Moreover, we show that the combination of low car-ownership rates among black households coupled with racial housing segregation renders evacuation plans based on private transportation most ineffective for predominantly minority communities. The results suggest that cities facing risks requiring evacuation cannot completely depend on decentralized private evacuation strategies.

2. Was New Orleans Unique? Basic Cross-City Differences in Households Access to Automobiles Nationwide, roughly 8 percent of the U.S. population resides in a household with no access to an automobile. As we will soon see, there are great disparities across racial and ethnic group, income class, and other demographic characteristics such as age. Before turning to these comparisons, however, here we present some basic estimates of the population with no access to 4a car for specific metropolitan areas and cities. We seek to assess whether access to private transportation in New Orleans prior to Katrina was particularly weak. In addition, we seek to provide estimates for policy makers and planners of the size of their region's population that does not have access to a car. Table 1 presents estimates of the percent of people in a household with no car as well as the overall population estimate for each metropolitan area in the United States. At the top of the table, the ten metropolitan areas with the largest percentage figures are isolated. Not surprisingly, the New York Consolidated Metro politan Statistical Area has the largest population of individuals in households without access to a single automobile (42 percent of the population accounting for almost 4 million people). Moreover, among the nation's top ten are some of the largest metropolitan areas in the country, such as Los Angeles-Long Beach, the San Francisco Bay Area, and Chicago. The New Orleans metropolitan area ranks at the top of this distribution, with Table 1 indicating that New Orleans has the fourth hi ghest rate of individua ls with no access to a household auto among the nearly 300 metropolitan areas in the country. As we will soon see, this fact is driven largely by the internal demographic composition of the metropolitan area. Figure 1 summarizes the relative position of the New Orleans metropolitan area. In the lion's share of metropolitan areas, the pe rcent of individuals in hous eholds with no cars lies below eight percent, with a large share of metropolitan areas below four percent.

Table 2 reproduces these tabulations for all U.S.

central cities identifie d in the 5 percent 2000 PUMS. Again, the table presents the percent of the population with no auto access as well as the corresponding estimate of the absolute population with access to a car. The top ten cities are isolated at the top of the table. The percentage s without access to a car ar e clearly higher for the 5nation's central cities relative to the comparable figures for entire metropolitan area. This fact reflects the relative poverty of central city populations , the higher costs asso ciated with owning and maintaining a car in more urban settings, and the wider availability of public transit in and around major American cities.

The percent of residents without access to a car in the city of New Orl eans is again high enough to place the city among the top ten. The remaining nine cities are all on the east coast, with Washington, D.C., Boston, and Baltimore being added to the list. Note, in nearly all of these top ten cities, the absolute size of the population with no access to a car is either comparable in magnitude or exceeds the estimate for New Orleans. Figure 2 summarizes the distribution of the percentage figures across central cities. Again, New Orleans is in the tail with a percentage with no car far above the average for ot her central cities As a final set of tabulations, Table 3 presents comparable city-specific estimates of the percent with no car access specifically for the non-elderly and elderly poor. In the aftermath of hurricane Katrina, it become evident that most of those stranded by the hurricane were poor, and many were elderly. Here we present figures for central cities only. The percentages in households with no cars are discretely higher for those residing in poverty. Averaging across all central cities, nearly 47 percent of the central city poor reside in house holds without a single automobile (with little difference between the elderly and non-elderly). This figure exceeds by far the percent without a car among the poor nationwide. Again, New Orleans is near the top of the pile along this dimension as are the same set of east coast central cities that appear in each of the two prior tables.

63. The demographic correlates of car-ownership That African-Americans have low car-ownership rates follows directly from sharp racial differences in household income and poverty. Low income households are often unable to generate the resources needed for down payments on expensive consumer durables that cannot be fully financed. Moreover, the cost of capital faced by poor households is likely to be higher on average due to such factors as poor credit histories and perhaps even discriminatory treatment in capital markets. In 2004, roughly 24.7 percent of African-Americans lived below the federal poverty line, compared with 8.6 percent of non-Hispanic whites. Median household income in

2004 for black households stood at $30,134 per year, only 62 percent of the median household income for whites.

The racial disparity in durable asset ownership that has perhaps received the most attention is housing. As of the fourth quarter of 2005, 48 percent of households headed by an African-American owned their residence, compared with 76 percent of non-Hispanic white residents and 50 percent of Hispanic residents. This racial disparity in home-ownership rates has received much attention, as home ownership is one of main avenues of wealth accumulation for U.S. residents. A house, in contrast to an automobile, generally appreciates in value over time. Moreover, borrowing against home equity is often considerably cheaper relative to the cost of capital for unsecured debt. Racial disparities in car ownership rates have received far less attention. Nonetheless, substantial disparities exist that may be linked to racial disparities in labor market outcomes (see for example, Raphael and Stoll (2001) and Raphael and Rice (2002)), time endowments, and, as has become evident post-Katrina, other important outcomes that affect the welfare of poor 7households. In this section, we document racial and ethnic differences in car ownership rates along with how car access differs by household income relative to the poverty lane. Table 4 presents tabulations from the 2000 five percent PUMS of the percent of people who reside in households without access to a single automobile. For all U.S. residents, residents of the New Orleans metropolitan statistical area, and residents of the city of New Orleans, the table presents estimates of the percent without cars by race/ethnicity (non-Hispanic white, non-Hispanic black, other non-Hispanic, and Hispanic) and income (those with incomes below 100 percent of the poverty line (the poor), those with incomes between 100 and 200 percent of the poverty line (the near poor), and everyone else). Beginning with the results for all U.S. residents, African-Americans have the lowest car-o wnership rates of all racial/ethnic groups. In particular, 19 percent of blacks reside in households without a single car, compared to 4.6 percent of whites, 13.7 percent of Hispanics, and 9.6 percent of those falling into the other category. These racial disparities are particularly large among the poor, suggesting an interaction effect betw een race and poverty on the likelihood of owning a car. For example, the black-white difference in car-ownership rates among the poor is over 22 percentage points (33 percent of poor blacks compared with 12 percent of poor white s), compared with an overall racial disparity of roughly 14 percentage points. This disparity narrows somewhat among the near poor but is still quite large (on the order of 12 percentage points). Within the New Orleans metropolitan area, and in particular within the city of New Orleans, car ownership rates were very low. For example, the overall percentage of residents with no access to a car was 14 percent in the metropolitan area and approximately 26 percent for the city, compared with 8 percent for the nation as a whole. Among the poor, 36 percent of the poor in the metropolitan area and nearly 47 percent of poor residents of the city did not have access to a 8single automobile, compared with 20 percent of th e poor nationwide. Interestingly, most of the differences between New Orleans and the nation are driven by particularly high figures for African-Americans and, more specifically, the African-American poor. R oughly half of African-Americans residing in poverty in New Orleans (using either the city or metropolitan area definition) had no access to a car, compared with 33 percent of poor blacks nationwide. In addition to these high percentages with no access to a car for specific racial and income groups, the racial composition of New Orleans coupled with an unusually high poverty rate further contributed to the overall size of the population at risk of being stranded. Table 5 presents tabulations from the 2000 PUMS of average socioeconomic and demographic characteristics for all residents of the U.S., residents of the New Orleans metropolitan area as well as residents of the city of New Orleans. The Table presents tabulations for all residents as well as the residents in a household with access to at least one car and residents in households with no cars. A simple comparison of the figures in the first column reveals much about why those stranded by Katrina were disproportionately poor and black. While African-Americans comprise 12.5 percent of the U.S. population in 2000, blacks accounted for 40 percent of the population of the New Orleans metropolitan area, and 67 percent of the city population. Moreover, while the national poverty rate stood at 14.6 percent in 2000, the poverty rate for the New Orleans metropolitan area was six percentage points greater (20.8 percent), while the poverty rate for the city of New Orleans (30.5 percent) was more than double the national rate.

To be sure, it is unlikely that those left behind in the New Orleans Superdome or Convention Center were a random sample of New Orleans residents. While such a random sample would surely select a population that was disproportiona tely black and poor, the figures in the second 9two columns of Table 5 suggest that differentia l access to autos within New Orleans between racial and income groups was an important contributor. For example, within the New Orleans metropolitan area, African-Americans constituted 34 percent of residents residing in a household with access to at least one car, while at the same time accounting for 77 percent of residents in households without access to a single automobile. Within the city of New Orleans, blacks

constituted 61 percent of residents in households with cars and 86 percent of residents of households without cars. Similar patterns are observed in the proportional representation of the poor. Those residing in poverty accounted for 16 perc ent of those in households with access to at least one automobile, but 53 percent of those in households w ith no cars. For the city of New Orleans, the comparable figures are 21 percent and 56 percent respectively. Thus, both nationwide and in the New Orleans metropolitan area, there are large racial and income disparities in car ownership rates, with unusually low car ownership rates within defined groups for New Orleans. The relative poverty and racial composition of those who did not evacuate certainly reflects in part the racial and income composition of the city of New Orleans, pre-Katrina. However, these sharp differences in auto-access rates likely diminished the probability of evacuating before the flooding for the city's poor and African-American residents.

4. Racial Housing Segregation and Neig hborhood Access to Private Automobiles Thus far we have demonstrated two general facts: (1) in nearly all metropolitan areas there are numerically large populations of individuals residing in households with no auto access, and (2) these populations are dispro portionately poor and minority. We have also shown that in 10post-Katrina New Orleans auto-access among poor and minority communities was particularly low. Clearly, one would expect a relationship between the ease of evacuation in response to a crisis and access to a private household auto, and the lower household auto ownership rates of African-Americans would clearly put them at a disadvantage. This disadvantage may be offset, however, if African-American house holds have neighbors with cars who could help in the event of a disaster. Interestingly, the stock of privately-owned automobiles in the New Orleans metropolitan area was more than sufficient to evacuate the entire population had neighbors with cars extended help to neighbors without. Tabulations from the 2000 census indicate that there were 2.3 persons for each automobile in New Orleans, only slightly lowe r than the 2.2 persons per car figure for the nation as a whole. Assuming an average carrying capacity of three persons

per auto, there should have been sufficient priv ate capacity to evacuate all residents of the metropolitan area. In practice, however, the racial housing segregation that is common in most U.S. metropolitan areas is likely to have compounded the African Americans' lower access to autos in New Orleans. Simply put, given that African-Americans have low car-access rates and African-Americans are quite likely to live in disproportionately African-American neighborhoods, black households with no access to a car are likely to be physically surround ed by other households without access to a car. Thus, racial segregation mechanically limits the potential for neighborly behavior to offset the consequences of racial disparities in car ownership. In this section, we explore the relations hip between racial segregation in housing and various measures of the degree of auto access among one's neighbors. To do so, we analyze census tract level data from the 2000 Census Summary Tape Files 3 for the entire nation and use 11 these d a ta to characterize the auto access rates of the neig hbor hoods o f the averag e white and black residents of all m e tropo litan areas in the U.S. W e use census tracts 1 as our m easure of neighborhood.

Characterizing the degree of modern-day se grega tion and its effect on neighborhood auto access Before discussing specific racial dif f erences in auto access, we first depict th e deg r ee of racial hous ing segregation characteristic of U.S.

m e tropolitan areas as o f the las t census. W h ile there are m a ny ways to nu m e rically characterize r acial housing segregation, a standard and easily inte rpre table segregation m easure is the diss im ila rity index. The dissim i larity index provides a single num b e r that depicts the degree of racial h ousing segregation in a given m e tropolitan area.

The index describes the degree of dissim ilarity between the spatial distribution of two population groups (for exam ple whites and blacks) across th e neighborhoods of a give n m e tropolitan area.

The index varies from 0 to 1, with a value corresponding to the propor tion of either of the m e tropolitan area's sub-populations that would h a ve to be relocated to achieve perfect balan ce.2 For exam ple, our tabulations from the 2000 STF3 f iles yield a black-white dissim ilarity score for New Orleans of 0.687, indicating that 68 percent of New Or leans blacks (or whites) would have had to m ove in 2000 to achi eve perfect in tegration.

Figure 3 su mmarizes th e degree of racial housing segregation betw een blacks and whites as of the 2000 census.

We constr uc t th e f i gur e as f o llows. First, we tab u lated th e dissim ilarity score b e tween black s and whites f o r each m e tropolitan area in the co untry. Nex t , we placed

1 Cen s u s tracts are sm all su b-co un ty g e og raph ic un its th at are m ean t to captu re relativ ely hom oge n ous areas withi n cities th at m a y b e thou gh t to ap pro x im ate n e ig hbo rho o d s. In th e 200 cen s u s , t h e av erag e po pu latio n of a cen su s tr act is app r ox imatel y 4 , 000 per s on s. 2 Th e d i ssim ilar ity in d e x is fairly si m p le to calcu late. D e fi n e th e v a riab le Black as th e t o tal b l ack pop u lati o n in a g i v e n m e tro p o litan area, Wh ite as th e to tal wh ite po pu lation in a g i v e n m e tropo litan area, Black i as the black p opu latio n of neig hb orh ood i, and w h ite i as th e w h ite p opu latio n o f n e i g hbo rho o d i. Th e d i ssimilarity in d e x fo r th e metro p o litan area is calcu lated b y th e expressio n I i i i White White Black Black D 1.2 1 12metropolitan areas into segregation groups based on 5 percentage point intervals. Finally, we allot African-Americans across these groups according to the size of the black population in each metropolitan area. Thus, the figure presents the distribution of segregation scores across metropolitan areas after weighting by the relative size of the black population in each region. The degree of black-white segregation varies considerably from a low of roughly 0.25 to a high of 0.83. However, most African-Americans live in fairly segregated metropolitan areas. The median black resident in U.S. metropolitan areas resides in an area with a segregation score of 0.645 (that is to say, 65 percent of blacks in the median African-American's metropolitan area would have to relocate to desegregate the city). While this represents a substantial improvement relative to the degree of segregation in 1990, the median level of segregation experienced by blacks remained high at the close of the 20 th century.3 The degree of segregation experienced by black residents of New Orleans (a dissimilarity score of 0.687) is slightly above the national median. To assess the extent to which racial housing segregation translates into lower neighborhood car access rates in the respective neighborhoods of blacks and whites, we performed a series of simple tabulations using the STF3 data. First, for each census tract in the country, we calculate the proporti on of households without access to a car as well as the number of persons per car in the census tract. Next, we tabulate the average values for these two neighborhood level auto access measures using various alternative populations as weights. First we tabulate averages using the total census tr act population as weights.

These averages are indicative of the car access measures in the neighborhood of the average American. Next, we tabulate averages using census tract white population as weights, giving means that can be 3 See Raphael and Stoll (2002) for a discussion of how racial segregation and racial disparities in geographic job access changed between 1990 and 2000.

13interpreted as the access rates for the average white American. Finally, we tabulate averages using the census tract's black population as weights, providing access rates for the average black American. To the extent that racial segregation concentrates black households in neighborhoods with low car ownership rates, the fraction without a car and the number of persons per car should be higher in the typical black neighborhood re lative to the typical white neighborhood. Figures 4 and 5 graphically depict these tabulations for the nation as a whole and for the New Orleans metropolitan area. Starting with Figure 4, the average U.S. resident resides in a census tract where roughly 10 percent of households have no access to a car. The comparable figure for New Orleans is substantially higher (15.4 percent), lik ely reflecting the higher poverty rates and higher fraction black in New Orleans. For the average wh ite U.S. resident, 8 percent of the households in one's neighborhood do not have single automobile, compared with approximately 20 percent for the neighborhood of the average African-American. In New Orleans, both figures are higher (9 percent for white and 26 percent for blacks), especially for African-Americans. Figure 5 presents the comparable tabulati ons for the ratio of neighborhood residents to cars. Nationwide as in New Orleans, there are slightly more than two persons per car in the neighborhood of the average U.S. resident. However, there are notable racial disparities. For the nation and New Orleans, the rati o of people to cars in the typical whit e neighborhood is under two. In black neighborhoods, the ration of people to cars ranges from 3.3 (New Orleans) to 3.5 (nationwide). Racial segregation and neighborhood auto access in New Orleans The tabulations in Figures 4 and 5 provide a nice summary of the typical neighborhood access to autos and how such access varies by race. A simpler way to reinforce this point is to 14analyze maps of specific regions in order to assess the visual correlation between racial housing segregation on the one hand and the spatial concentr ation of households with no autos on the other. Figures 6 and 7 present such an analysis for the city of New Orleans. Figure 6 presents the fraction of each census tract's residents that are African-American. As is readily apparent, the black community is visibly concentrated in a handful of neighborhoods, including New Orleans East, the lower ninth ward, and the seventh ward. Figure 7 gr aphically depicts two variables: the proportion of households without access to an automobile (reflected in the different shadings of the geogra phic sub-units of the city) and the number of households without an auto (each dot representing 20 households). A comparison of Figures 6 and 7 reveals that household without access to automobiles are not randomly distributed. There is a notable visual correlation between the proportion of households th at are black and the proportion of households without cars. In particular, predominantly African American neighborhoods such as the Lower Ninth Ward and Holy Cross, the Seventh Ward, and Tremé, all inundated with flood waters, show large concentrations of car-less households, in comparison to the mostly white, also-inundated Lakeview district.

Is the proportion carless in black neighbor hoods higher in more segregated cities? Thus far, we have shown that neighborhood access to autos is lower for African-Americans (nationwide as well as in New Orleans) and, within New Orleans, is negatively correlated with the proportion of neighborhood residents that are black. An interesting implication of these patterns concerns whether blacks in less segregated cities have better access to neighbors with cars than blacks in more segregated cities.

Figure 3 reveal s considerable variation in racial segregation across metropolitan areas, although most bl acks reside in fairly 15 segregated conditions. Here, we briefly e xplore how segregation and neighborhood auto access vary across metropolitan areas. Figure 8 presents a scatter plot of the proportion of households with no access to a car in the neighborhood of the average black resident against the black-white dissimilarity index. Each data point represents one of the approximately 280 metropolitan areas in the country. There is notable positive relationship between the proportion of neighbors without a car in black neighborhoods and the degree of segregation between blacks and whites. This is precisely what one would expect given the relatively low car ownership rates of black households. To ascertain whether the co rrelation depicted in Figure 8 is driven by some underlying characteristic that varies across metropolitan areas and that affects car-ownership rates for both blacks and whites, Figure 9 presents a comparable scatter plot where the variable on the vertical axis is now the difference in the proportion of households without a car be tween the typical black and white neighborhoods of the city. By cal culating the difference relative to white neighborhoods, the figure nets out any unobserved metropolitan area characteristic that may influence overall car access rates. Again, there is a strong positive correlation between the relative proportion of households without cars in black neighborhoods and the degree to which blacks are segregated from whites.

5. Conclusion The patterns documented in this paper are several. First, we show that in most metropolitan areas there are sizable populations of households without access to automobiles.

Individuals in these households are more likely to be poor and minority and are perhaps the most likely to be left behind in the event of an emergency evacuation. Second, there are quite large 16disparities in car-ownership rates between black and white households. These disparities certainly reflect average differences in income as well as differential costs of owning a car. Nonetheless, even among the poor, black househol ds are considerably less likely than white households to have access to an auto. Finally there is a compounding effect of racial segregation on access to automobiles via the fact that low black car ownership rates coupled with racial housing segregation concentrates poor minorities in neighborhoods where many households lack automobile access. These findings raise several questions regarding how federal, state, and local policy makers should plan for the evacuation of a region's most vulnerable residents. Given low car-ownership rates among certain sub-groups, one might ask whether a city's existing public transit infrastructure can be adequately mobilized to fill the gaps left by a private transport evacuation. Moreover, given many of the difficulties experienced in the evacuation of Houston soon after the Katrina disaster, one might furt her inquire whether higher car ownership rates would appreciably help. In the case of Katrina, the pre-existing public transit evacuation plan failed miserably. New Orleans had an emergency plan to evacuat e one to two hundred thous and people by bus in the event of an emergency, and as the hurricane approached the City established ten pickup points to take people to emergency shelters. Drivers were to take their own families to safety on the first loaded bus run and th en continue to evacuate others. However, the plan failed completely when only a few drivers reported to work during the evacuatio n period. Residents then were directed to the Superdome and the Convention Center despite the lack of adequate water, food, or security there.

17Even had the bus drivers done their jobs, the plan was not likely to succeed. A few simple calculations demonstrate the enormity of the task of evacuating those without private transportation and the likely limited ability of transit to rise to the occasion. Optimistically, 500 buses 4 could have evacuated perhap s forty persons per vehicle or 20,000 persons per set of outbound trips. A fifty-mile trip to shelters outside the likely high damage areas (Baton Rouge is 80 miles away, so 50 miles is conservative) would at minimum take one hour assuming orderly loading and unloading, minimal time for driver relief, and little or no traffi c congestion. Another hour would be consumed as the bus unloaded and then returned for the next run. If drivers had been willing to work 16 hour1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> days, 160,000 persons could have been evacuated in a day under ideal circumstances - i.e., drivers could be found for all runs, fuel was available for rapid refueling, no breakdowns occurred, and no traffic backups or other incidents slowed the trips. A more realistic plan would assume that speeds on freeways would be no more than twenty miles per hour during an evacuation, and perhaps as low as ten miles per hour. At these speeds a one hundred mile round trip would take five to six hours (assuming the inbound trip would be at posted speeds) and no more than two runs a day could be accomplished. This would have cut the New Orleans bus evacuation rate to only 40,000 per sixteen-hour workday.

Thus, with the benefit of hindsight, it is not too surprising that the transit evacuation strategy in place failed. The system capacity was small relative to the population without automobiles and serious unanticipated implementation problems (for example, the failure of drivers to show up) hampered the effectiveness of the effort. To be sure, the results from this one disaster does not indicate that public transit cannot partially or even completely fill the gap left by less than universal access to private 4 In the New Orleans Regional Transit Authority's 2/19/2004 report to the National Transit Database, the agency reports 364 buses and 40 demand-response vehicles available for maximum service (no spares, no vehicles down for repair). Elsewhere, the school bus fleet has been estimated to be about 100 vehicles (Litman, 2005.)

18transportation. In other disasters, transit has played a significant role in both evacuations and in post-disaster recovery. For example, transit was the principal means of evacuating hundreds of thousands of people from lower Manhattan after the attack on the World Trade Center. Transit also was used in New York to bring emergency aid into the area and in providing emergency assistance. Likewise in Washington, DC, Me tro evacuated several h undred thousand people after the attack on the Pentagon. Nonetheless, the difficulties encountered in the wake of Katrina suggest that more comprehensive (and perhaps creative) planning is needed to avoid such catastrophes in the future. For example, while the initial plan was to evacuate New Orleans residents from the city, an alternative was to move those without cars and those with nowhere to go only as far as the nearest "safe" location within the storm zone - the Superdome and Convention Center were intended to be such sights. While the Superdome proved to be a severely flawed destination, identifying a more suitable location within say ten to twelve miles would have greatly reduced bus travel distance and perhaps have permitted moving everyone to shelter in a day. Alternatively, reports indicate that Amtrak had equipment available for evacuation and this could have greatly increased New Orleans' overall evacuation capacity. A single train car could carry two buses worth of evacuees. Moreover, the exclusive right-of-way for rail means much less serious congestion problems. However, communications broke down and this option was not used. Similarly, buses were not brought in from other jurisdictions to help with the evacuation. Perhaps tapping such alternative so urces would provide a substantial boost to the transit evacuation capacity. The limitations to transit based evacuation stra tegies raises the question of whether higher car-ownership rates would increase the swiftness a nd coverage of a genera l evacuation. Clearly, 19low car ownership rates among poor and vulnerable populations will continue in nearly all American cities. Even the smallest metropolitan areas have literally thousands of residents who do not have a car. Moreover, a significant pr oportion of households wit hout a car also have mobility limitations or other physical or mental disabilities that would prevent them from driving. Access to a car, however, does not guarantee an easy evacuation. In New Orleans when Katrina hit, some poor families with cars chose to remain in their homes because they did not know anyone outside the city that could provide sh elter and did not have the money to stay in a hotel; others stayed because th ey did not want to leave pets, for whom evacuation plans had made no provisions (personal interviews, Jan. 2006). Lack of information regarding the availability of shelters further discouraged departure. Further, a study done two years before Katrina showed that there would not be enough ro adway capacity to evacuate everyone by car in any case; with only 60% evacuating by car, roads were predicted to be jammed with speeds lowered to five to ten miles per hour. (Wolshon, 2002). Other cities' experiences also underscore the limitations of auto-based evacuations. Following 9/11, many Washington D.C. workers left their cars in the city and evacuated by transit due to standstill traffic. Average speeds on both local streets and freeways dropped to less than ten miles an hour and some travel lanes were blocked by cars abandoned after crashes, breakdown, or running out of gas. The Houston experience, a few weeks after Katrina, shows that even a reasonably well-orchestrated car-focused evacuation strategy can fail. As Hurricane Rita bore down on the Houston-Galveston area, officials called for ea rly evacuation and planned to reverse freeway direction to increase capacity northbound, away from the coast. Plans assumed that as many as 201.2 million people would flee for points north. However, early notices recommended that everyone leave the metropolitan area, and some 2.5 million, about half the region's population, apparently attempted to do so. This led to massive traffic jams on all the freeways, with average speeds under five miles an hour on many routes.

5 Given the limitations of both transit as well as private transportation evacuation strategies, what then should localities do in planning for future catastrophes? We have shown that thousands of people in every metropolitan area of the US don't have cars to help them evacuate in the case of disaster. Many of them live in areas where transit could serve as a good alternative, but New Orleans shows that transit can fail badly and that even a sizeable transit capacity won't be enough in a major disaster. Moreover, many of the nation's metropolitan areas do not really have extensive transit systems, so there is no realistic possibility of relying on transit alone in those areas. Cars can evacuate thousands or even millions, but in a mass evacuation, cars can be caught in multi-hour traffic jams , with vehicle breakdowns, fuel shortages, and in the worst cases, road blockages that constrict movement in or out of the disaster area, exacerbating the

disaster. Cars won't work if the roads out have collapsed or are blocked, if fuel can't be found, or if the car itself is damaged in the disaster; cars may not be fast enough if everyone is evacuating at the same time. Tapping alternative sources of transit capacity is clearly an avenue worth exploring. Areas that face comparable yet imperfectly correlated risks (such as cities on the gulf coast) may 5 Cars overheated and broke down or ran out of fuel, blocking roads and cutting capacity further. Gas stations ran dry and others were closed for lack of staff as their operators joined the evacuation. Plans to reverse freeway lanes to increase outbound capacity were abandoned; officials stated publicly that the lanes were needed to get emergency vehicles and supplies into the danger zone, but privately they admit that a bigger reason was that the equipment and staff were not available to reverse direction of the lanes, a process that requires intervention at every on-ramp. In addition, they discovered that many motorists needed to head South to pick up family members before evacuating, and closures caused considerable motorist confusion and consternation. (personal interviews, November 2005.)

21 benefit from pre-arranged sharing agreements whereby inter-lo cality cooperation can expand the capacity of a city's transit system when needed. Sharing rides may also provide an alternative way to get more people out. To be sure, many drivers fill their cars with family members, pets and valuables; empty seats may be scarce.

However, hardly any area has organized to prov ide ridesharing in a disaster and we don't know what it would take to make it work. Future thinking and research in this area is needed. Other options need to be coordinated better than they were in New Orleans. Trains could have evacuated thousands but were left idle. In the Bay Area, ferries helped evacuate people in 1906; the Bay Area Regional Transit System kept going after the Loma Prieta earthquake. But the next quake could block or damage ferry access or damage the BART tube or tracks. It's hard to know in advance what will be accessible and usable and what will not, so scenarios should be played out.

22References Deakin, Elizabeth (1996), "Effects of the Loma Prieta Earthquake on Transbay Travel Patterns", UCTC Working paper 294.

Littman, T. "Lessons from Katrina and Rita: What Major Disasters Can Teach Transportation Planners", Victoria: Victoria Transpor tation Policy Institute, November 2005.

http://www.vtpi.org/katrina.pdf Raphael, Steven and Lorien Rice (2002), "Car Ownership, Employment, and Earnings," Journal of Urban Economics, 52: 109-130.

Raphael, Steven and Michael A. Stoll (2001) "Can Boosting Minority Car-Ownership Rates Narrow Inter-Racial Employment Gaps?," in William G. Gale and Janet Rothenberg Pack (eds), The Brookings-Wharton Papers on Urban Economic Affairs, volume 2, The Brookings Institution: Washington, DC: pp 99-145.

Raphael, Steven and Michael A. Stoll (2002), "Modest Progress: The Narrowing Spatial Mismatch Between Blacks and Jobs in the 1990s," The Brookings Institution: Washington, DC.

Wolshon, Brian (2002), "Planning for the Evacuation of New Orleans,"

ITE Journal (www.ite.org/itejournal/index.asp), February 2002.

23 Table 1 Percent and Number of Residents in Households with no Access to an Automobile, 2000 Top ten metropolitan areas Percent with no auto access Population with

no auto access New York-Northeastern NJ 42% 3,949,939 Jersey City, NJ 30% 182,252 Waterbury, CT 16% 17,731 New Orleans, LA 14% 175,289 Philadelphia, PA 13% 646,269 Baltimore, MD 12% 290,809 Newark, NJ 12% 243,673 San Francisco-Oakland-Vallejo, CA 12% 201,303 Chicago-Gary-Lake, IL 11% 877,429 Los Angeles-Long Beach, CA 11% 1,036,738 All metropolitan areas Abilene, TX 5% 6,476 Akron, OH 5% 35,049 Albany, GA 9% 10,975 Albany-Schenectady-Troy, NY 8% 62,644 Albuquerque, NM 5% 32,386 Alexandria, LA 9% 11,541 Allentown-Bethlehem-Easton, PA/NJ 7% 41,857 Altoona, PA 8% 10,082 Amarillo, TX 4% 7,863 Anchorage, AK 4% 11,125 Ann Arbor, MI 3% 15,610 Anniston, AL 6% 6,331 Appleton-Oshkosh-Neenah, WI 3% 9,606 Asheville, NC 5% 11,461 Athens, GA 5% 8,386 Atlanta, GA 6% 255,959 Atlantic City, NJ 11% 38,481 Auburn-Opelika, AL 4% 4,871 Augusta-Aiken, GA-SC 7% 30,915 Austin, TX 5% 52,757 Bakersfield, CA 8% 53,768 Baltimore, MD 12% 290,809 Barnstable-Yarmouth, MA 3% 4,326 Baton Rouge, LA 6% 39,183 24Beaumont-Port Arthur-Orange, TX 7% 26,692 Bellingham, WA 4% 7,390 Benton Harbor, MI 6% 9,366 Bergen-Passaic, NJ 9% 122,761 Billings, MT 4% 5,779 Biloxi-Gulfport, MS 5% 14,449 Binghamton, NY 8% 19,137 Birmingham, AL 7% 52,528 Bloomington, IN 4% 5,341 Bloomington-Normal, IL 4% 5,483 Boise City, ID 3% 13,922 Boston, MA 10% 334,172 Boulder-Longmont, CO 4% 8,061 Brazoria, TX 3% 7,725 Bremerton, WA 4% 10,105 Bridgeport, CT 9% 31,838 Brockton, MA 6% 14,643 Brownsville-Harlingen-San Benito, TX 9% 29,783 Bryan-College Station, TX 5% 8,051 Buffalo-Niagara Falls, NY 10% 122,645 Canton, OH 4% 17,479 Cedar Rapids, IA 4% 8,160 Champaign-Urbana-Rantoul, IL 6% 11,418 Charleston-N. Charleston, SC 9% 39,429 Charlotte-Gastonia-Rock Hill, SC 5% 72,754 Charlottesville, VA 5% 7,676 Chattanooga, TN/GA 5% 22,215 Chicago-Gary-Lake, IL 11% 877,429 Chico, CA 5% 11,088 Cincinnati OH/KY/IN 8% 115,193 Clarksville-Hopkinsville, TN/KY 4% 5,560 Cleveland, OH 8% 184,821 Colorado Springs, CO 4% 18,777 Columbia, MO 5% 6,595 Columbia, SC 6% 31,832 Columbus, GA/AL 9% 17,118 Columbus, OH 5% 79,196 Corpus Christi, TX 5% 12,422 Dallas-Fort Worth, TX 5% 177,464 Danbury, CT 3% 5,288 Danville, VA 10% 11,156 25 Davenport, IA Rock Island-Moline, IL 6% 15,215 Daytona Beach, FL 5% 20,458 Dayton-Springfield, OH 6% 59,374 Decatur, AL 4% 5,927 Decatur, IL 6% 6,955 Denver-Boulder-Longmont, CO 6% 111,526 Des Moines, IA 4% 16,708 Detroit, MI 7% 312,133 Dothan, AL 6% 7,819 Dover, DE 6% 7,105 Duluth-Superior, MN/WI 7% 13,568 Dutchess Co., NY 5% 14,377 Eau Claire, WI 4% 5,815 El Paso, TX 7% 49,802 Elkhart-Goshen, IN 7% 12,767 Erie, PA 7% 19,267 Eugene-Springfield, OR 5% 16,588 Evansville, IN/KY 5% 13,710 Fargo-Morehead, ND/MN 4% 5,386 Fayetteville, NC 5% 14,736 Fayetteville-Springdale, AR 3% 10,714 Fitchburg-Leominster, MA 7% 9,583 Flagstaff, AZ-UT 5% 5,947 Flint, MI 9% 22,122 Florence, AL 4% 5,712 Fort Collins-Loveland, CO 3% 5,949 Fort Lauderdale-Hollywood-Pompano Beach, FL 6% 105,401 Fort Myers-Cape Coral, FL 5% 20,280 Fort Pierce, FL 4% 14,406 Fort Smith, AR/OK 5% 8,874 Fort Walton Beach, FL 2% 4,165 Fort Wayne, IN 6% 25,801 Fort Worth-Arlington, TX 4% 69,076 Fresno, CA 10% 88,046 Gadsden, AL 6% 6,620 Gainesville, FL 5% 11,773 Galveston-Texas City, TX 6% 16,002 Gary-Hammond-East Chicago, IN 7% 41,805 Glens Falls, NY 5% 6,470 Goldsboro, NC 7% 7,462 Grand Junction, CO 4% 4,003 26 Grand Rapids, MI 4% 41,595 Greeley, CO 3% 6,227 Green Bay, WI 4% 8,340 Greensboro-Winston Salem-High Point, NC 6% 69,650 Greenville, NC 7% 9,507 Greenville-Spartanburg-Anderson SC 6% 44,295 Hagerstown, MD 6% 7,477 Hamilton-Middleton, OH 4% 14,568 Harrisburg-Lebanon-Carlisle, PA 6% 38,170 Hartford-Bristol-Middleton-New Britain, CT 10% 71,732 Hattiesburg, MS 5% 5,163 Hickory-Morgantown, NC 4% 13,658 Honolulu, HI 8% 71,373 Houma-Thibodoux, LA 7% 7,677 Houston-Brazoria, TX 6% 271,082 Huntsville, AL 4% 12,276 Indianapolis, IN 5% 80,229 Iowa City, IA 4% 4,313 Jackson, MI 5% 7,684 Jackson, MS 6% 28,334 Jackson, TN 8% 8,276 Jacksonville, FL 6% 63,967 Jacksonville, NC 4% 5,496 Jamestown-Dunkirk, NY 8% 11,210 Janesville-Beloit, WI 3% 5,059 Jersey City, NJ 30% 182,252 Johnson City-Kingsport-Bristol, TN/VA 4% 14,139 Johnstown, PA 6% 14,667 Joplin, MO 4% 6,336 Kalamazoo-Portage, MI 5% 21,581 Kankakee, IL 5% 5,168 Kansas City, MO-KS 5% 84,916 Kenosha, WI 4% 6,127 Kileen-Temple, TX 4% 11,367 Knoxville, TN 5% 26,109 Kokomo, IN 5% 4,574 LaCrosse, WI 4% 4,347 Lafayette, LA 7% 17,887 Lafayette-W. Lafayette, IN 4% 7,361 Lake Charles, LA 6% 10,688 Lakeland-Winterhaven, FL 5% 24,498 27Lancaster, PA 9% 41,469 Lansing-E. Lansing, MI 4% 18,255 Laredo, TX 8% 15,983 Las Cruces, NM 4% 6,652 Las Vegas, NV 7% 101,104 Lawrence-Haverhill, MA/NH 10% 25,762 Lexington-Fayette, KY 6% 15,256 Lima, OH 4% 6,729 Lincoln, NE 4% 9,180 Little Rock-North Little Rock, AR 6% 33,504 Longview-Marshall, TX 6% 10,067 Los Angeles-Long Beach, CA 11% 1,036,738 Louisville, KY/IN 7% 68,185 Lowell, MA/NH 5% 16,011 Lubbock, TX 5% 12,403 Lynchburg, VA 7% 14,208 Macon-Warner Robins, GA 7% 23,754 Madison, WI 5% 23,613 Manchester, NH 6% 6,839 Mansfield, OH 5% 6,900 McAllen-Edinburg-Pharr-Mission, TX 7% 42,237 Medford, OR 3% 6,220 Melbourne-Titusville-Cocoa-Palm Bay, FL 4% 17,223 Memphis, TN/AR/MS 9% 85,869 Merced, CA 9% 18,490 Miami-Hialeah, FL 10% 228,192 Middlesex-Somerset-Hunterdon, NJ 5% 61,028 Milwaukee, WI 9% 140,448 Minneapolis-St. Paul, MN 5% 154,771 Mobile, AL 6% 34,471 Modesto, CA 7% 29,312 Monmouth-Ocean, NJ 5% 60,251 Monroe, LA 11% 16,034 Montgomery, AL 6% 21,664 Muncie, IN 5% 5,503 Myrtle Beach, SC 5% 9,179 Naples, FL 4% 9,289 Nashua, NH 5% 5,999 Nashville, TN 5% 56,176 Nassau Co, NY 4% 120,914 New Bedford, MA 10% 17,524 28New Haven-Meriden, CT 11% 39,930 New Orleans, LA 14% 175,289 New York-Northeastern NJ 42% 3,949,939 Newark, NJ 12% 243,673 Newburgh-Middletown, NY 9% 30,939 Norfolk-VA Beach-Newport News, VA 7% 103,646 Oakland, CA 7% 169,166 Ocala, FL 4% 10,544 Odessa, TX 5% 12,022 Oklahoma City, OK 5% 44,908 Olympia, WA 5% 9,856 Omaha, NE/IA 5% 29,776 Orange County, CA 5% 152,799 Orlando, FL 5% 75,368 Panama City, FL 5% 6,935 Pensacola, FL 5% 21,380 Peoria, IL 5% 17,920 Philadelphia, PA/NJ 13% 646,269 Phoenix, AZ 5% 166,950 Pittsburgh-Beaver Valley, PA 9% 198,462 Portland, ME 5% 13,000 Portland-Vancouver, OR 5% 97,319 Providence-Fall River-Pawtucket, MA/RI 9% 87,841 Provo-Orem, UT 3% 9,183 Pueblo, CO 7% 9,620 Punta Gorda, FL 4% 5,160 Racine, WI 5% 9,833 Raleigh-Durham, NC 5% 56,984 Reading, PA 8% 27,727 Redding, CA 5% 8,130 Reno, NV 7% 23,300 Richland-Kennewick-Pasco, WA 4% 8,165 Richmond-Petersburg, VA 7% 67,107 Riverside-San Bernardino, CA 6% 203,710 Roanoke, VA 6% 13,864 Rochester, MN 4% 5,421 Rochester, NY 7% 72,861 Rockford, IL 5% 16,827 Rocky Mount, NC 9% 12,550 Sacramento, CA 6% 96,444 Saginaw-Bay City-Midland, MI 5% 19,880 29Salem, OR 5% 14,585 Salinas-Sea Side-Monterey, CA 6% 17,461 Salt Lake City-Ogden, UT 3% 45,122 San Antonio, TX 7% 106,999 San Diego, CA 6% 179,556 San Francisco-Oakland-Vallejo, CA 12% 201,303 San Jose, CA 4% 75,538 San Luis Obispo-Atascadero-P Robles, CA 4% 8,650 Santa Barbara-Santa Maria-Lompoc, CA 6% 23,078 Santa Cruz, CA 5% 11,994 Santa Fe, NM 3% 4,913 Santa Rosa-Petaluma, CA 4% 17,087 Sarasota, FL 4% 25,992 Savannah, GA 9% 21,350 Scranton-Wilkes-Barre, PA 7% 45,835 Seattle-Everett, WA 5% 123,523 Sharon, PA 7% 7,983 Sheboygan, WI 4% 4,098 Shreveport, LA 9% 36,052 Sioux City, IA/NE 6% 5,702 Sioux Falls, SD 4% 5,287 South Bend-Mishawaka, IN 6% 14,854 Spokane, WA 5% 21,624 Springfield, IL 8% 8,942 Springfield, MO 4% 14,322 Springfield-Holyoke-Chicopee, MA 9% 55,486 St. Cloud, MN 3% 4,840 St. Joseph, MO 6% 5,652 St. Louis, MO-IL 7% 173,574 Stamford, CT 5% 18,998 State College, PA 6% 8,084 Stockton, CA 7% 41,948 Sumter, SC 8% 8,554 Syracuse, NY 8% 58,100 Tacoma, WA 5% 32,222 Tallahassee, FL 6% 16,044 Tampa-St. Petersburg-Clearwater, FL 6% 135,858 Terre Haute, IN 5% 7,040 Toledo, OH/MI 6% 36,775 Topeka, KS 4% 7,534 Trenton, NJ 9% 31,239 30 Tucson, AZ 7% 58,110 Tulsa, OK 5% 37,376 Tuscaloosa, AL 6% 9,925 Tyler, TX 4% 7,665 Utica-Rome, NY 8% 22,703 Vallejo-Fairfield-Napa, CA 4% 23,372 Ventura-Oxnard-Simi Valley, CA 4% 29,415 Vineland-Milville-Bridg etown, NJ 10% 14,495 Visalia-Tulare-Porterville, CA 9% 31,867 Waco, TX 5% 11,194 Washington, DC/MD/VA 8% 396,047 Waterbury, CT 16% 17,731 Waterloo-Cedar Falls, IA 5% 6,025 Wausau, WI 3% 3,926 West Palm Beach-Boca Raton-Delray Beach, FL 6% 67,890 Wichita Falls, TX 5% 6,548 Wichita, KS 4% 21,568 Williamsport, PA 6% 7,484 Wilmington, DE/NJ/MD 6% 31,000 Wilmington, NC 5% 12,447 Worcester, MA 9% 26,112 Yakima, WA 6% 12,514 Yolo, CA 6% 10,916 York, PA 5% 18,595 Youngstown-Warren, OH-PA 5% 32,070 Yuba City, CA 6% 8,869 Yuma, AZ 6% 9,649 Tabulations from the 5 Percent Public Use Microdata Sample of the U.S. Census of Population and Housing.

31 Table 2 Percent and Number of Residents in Households with no Access to an Automobile U.S. Central Cities, 2000 Top ten cities Percent with no auto access Population with no auto

access New York, NY 48% 3,815,980 Newark, NJ 38% 101,835 Jersey City, NJ 34% 83,610 Baltimore, MD 32% 205,544 Hartford, CT 32% 38,975 Washington, DC 32% 182,927 Philadelphia, PA 30% 460,331 Boston, MA 28% 163,340 New Orleans, LA 26% 123,084 Buffalo, NY 26% 75,563 All cities Akron, OH 9% 20,635 Alexandria, VA 10% 12,697 Allentown, PA 15% 16,619 Anaheim, CA 8% 25,870 Anchorage, AK 4% 11,125 Ann Arbor, MI 5% 6,075 Bakersfield, CA 7% 18,287 Baltimore, MD 32% 205,544 Baton Rouge, LA 10% 23,830 Beaumont, TX 10% 11,534 Bellevue, WA 3% 3,791 Boise, ID 3% 6,329 Boston, MA 28% 163,340 Bridgeport, CT 19% 26,317 Brownsville, TX 10% 14,592 Buffalo, NY 26% 75,563 Burbank, CA 8% 7,769 Cambridge, MA 19% 19,719 Cape Coral, FL 2% 2,269 Charlotte, NC 7% 36,177 Chattanooga, TN 10% 14,744 Chesapeake, VA 4% 8,292 32Chicago, IL 23% 668,489 Cincinnati, OH 20% 67,740 Cleveland, OH 21% 98,461 Columbia, SC 12% 13,584 Columbus, GA 9% 17,118 Corona, CA 5% 5,898 Corpus Christi, TX 7% 19,815 Costa Mesa, CA 5% 5,952 Dayton, OH 16% 27,580 Des Moines, IA 6% 12,778 Detroit, MI 18% 171,495 Downey, CA 7% 7,913 El Monte, CA 15% 17,000 Elizabeth, NJ 21% 25,106 Erie, PA 13% 13,573 Evansville, IN 8% 10,434 Fayetteville, NC 8% 8,928 Flint, MI 13% 16,680 Fontana, CA 6% 7,206 Fort Collins, CO 3% 3,596 Fort Wayne, IN 7% 13,376 Fresno, CA 13% 53,958 Fullerton, CA 6% 7,754 Garden Grove, CA 6% 9,763 Gary, IN 15% 15,395 Glendale, CA 11% 20,611 Grand Rapids, MI 8% 16,699 Green Bay, WI 5% 5,605 Greensboro, NC 7% 16,373 Hampton, VA 7% 9,602 Hartford, CT 32% 38,975 Huntington Beach, CA 4% 7,006 Huntsville, AL 5% 7,653 Independence, MO 4% 4,595 Inglewood, CA 12% 13,405 Irvine, CA 3% 3,850 Irving, TX 5% 9,184 Jackson, MS 10% 17,982 Jersey City, NJ 34% 83,610 Kansas City, MO 10% 41,870 Knoxville, TN 8% 14,410 33Lafayette, LA 8% 8,824 Lancaster, CA 8% 9,809 Lansing, MI 8% 8,902 Lexington-Fayette, KY 6% 15,256 Little Rock, AR 8% 14,313 Livonia, MI 2% 2,298 Long Beach, CA 15% 68,720 Los Angeles, CA 15% 553,423 Lowell, MA 11% 12,239 Madison, WI 9% 18,731 Manchester, NH 6% 6,839 McAllen, TX 6% 6,662 Memphis, TN 12% 79,027 Milwaukee, WI 18% 109,656 Minneapolis, MN 15% 58,005 Mobile, AL 10% 19,307 Modesto, CA 7% 12,982 Montgomery, AL 8% 16,613 Moreno Valley, CA 5% 7,402 New Haven, CT 23% 28,996 New Orleans, LA 26% 123,084 New York, NY 48% 3,815,980 Newark, NJ 38% 101,835 Newport News, VA 9% 15,379 Norfolk, VA 14% 32,485 North Las Vegas, NV 8% 9,391 Norwalk, CA 6% 6,744 Oklahoma City, OK 6% 31,123 Ontario, CA 8% 12,314 Orange, CA 5% 6,449 Oxnard, CA 7% 11,517 Palmdale, CA 7% 7,882 Pasadena, CA 10% 14,277 Pasadena, TX 7% 9,523 Paterson, NJ 24% 36,268 Peoria, IL 10% 11,110 Philadelphia, PA 30% 460,331 Pittsburgh, PA 22% 71,588 Plano, TX 2% 3,530 Pomona, CA 9% 13,345 Providence, RI 17% 29,911 34 Provo, UT 3% 3,609 Raleigh, NC 5% 14,698 Rancho Cucamonga, CA 3% 3,805 Reno, NV 10% 17,394 Richmond, VA 19% 36,909 Riverside, CA 7% 17,946 Rochester, NY 21% 44,507 Rockford, IL 8% 12,086 Sacramento, CA 10% 41,076 Saint Louis, MO 22% 74,966 Saint Paul, MN 12% 33,949 Salinas, CA 6% 9,529 Salt Lake City, UT 8% 13,518 San Bernardino, CA 12% 23,137 San Buenaventura (Ventura), CA 4% 4,449 San Francisco, CA 21% 160,445 Santa Ana, CA 10% 33,992 Santa Clarita, CA 4% 5,452 Savannah, GA 15% 18,985 Seattle, WA 11% 62,269 Simi Valley, CA 2% 2,165 Sioux Falls, SD 4% 5,287 South Bend, IN 9% 9,786 Spokane, WA 8% 16,080 Springfield, IL 8% 8,942 Springfield, MA 19% 28,664 Springfield, MO 6% 8,982 Stamford, CT 8% 9,728 Sterling Heights, MI 3% 3,877 Stockton, CA 11% 25,747 Syracuse, NY 21% 30,807 Tacoma, WA 8% 14,894 Thousand Oaks, CA 3% 2,984 Toledo, OH 10% 30,329 Torrance, CA 4% 5,387 Tulsa, OK 7% 26,649 Vancouver, WA 6% 8,378 Virginia Beach, VA 3% 14,698 Warren, MI 4% 5,579 Washington, DC 32% 182,927 Waterbury, CT 16% 17,731 35 West Covina, CA 5% 5,117 Wichita, KS 5% 17,093 Winston-Salem, NC 9% 17,394 Worcester, MA 14% 23,152 Yonkers, NY 23% 46,171 Tabulations from the 5 Percent Public Use Microdata Sample of the U.S. Census of Population and Housing.

36 Table 3 Percent of Non-elderl y and Elderl y Poor In Households without Access to a Car by City Percent of non-elderl y poor residin g in a household with no carPercent of elderl y poor in a household with no car Top Ten Cities New York, NY 68% 64% Jersey City, NJ 58% 52% Newark, NJ 54% 52% Baltimore, MD 53% 53% Yonkers, NY 53% 42% Philadelphia, PA 50% 49% Washington, DC 48% 47% Buffalo, NY 47% 36% New Orleans, LA 47% 44% Hartford, CT 44% 50% All Cities Akron, OH 26% 33% Alexandria, VA 23% 18% Allentown, PA 32% 30% Anaheim, CA 19% 17% Anchorage, AK 10% 25% Ann Arbor, MI 7% 29% Bakersfield, CA 19% 18% Baltimore, MD 53% 53% Baton Rouge, LA 21% 31% Beaumont, TX 27% 29% Bellevue, WA 13% 18% Boise, ID 9% 8% Boston, MA 38% 48% Bridgeport, CT 39% 41% Brownsville, TX 20% 36% Buffalo, NY 47% 36% Burbank, CA 22% 38% Cambridge, MA 14% 66% Cape Coral, FL 5% 8% Charlotte, NC 20% 26% Chattanooga, TN 23% 32% Chesapeake, VA 16% 21% Chicago, IL 43% 48%

37 Cincinnati, OH 41% 37% Cleveland, OH 39% 42% Columbia, SC 15% 40% Columbus, GA 25% 29% Corona, CA 23% 14% Corpus Christi, TX 21% 26% Costa Mesa, CA 5% 26% Dayton, OH 29% 38% Des Moines, IA 15% 17% Detroit, MI 33% 38% Downey, CA 15% 8% El Monte, CA 28% 16% Elizabeth, NJ 37% 43% Erie, PA 29% 30% Evansville, IN 21% 22% Fayetteville, NC 22% 29% Flint, MI 29% 31% Fontana, CA 12% 13% Fort Collins, CO 5% 10% Fort Wayne, IN 16% 19% Fresno, CA 25% 30% Fullerton, CA 12% 9% Garden Grove, CA 17% 16% Gary, IN 25% 25% Glendale, CA 22% 30% Grand Rapids, MI 19% 19% Green Bay, WI 11% 16% Greensboro, NC 15% 24% Hampton, VA 12% 25% Hartford, CT 44% 50% Huntington Beach, CA 13% 4% Huntsville, AL 11% 23% Independence, MO 13% 21% Inglewood, CA 26% 25% Irvine, CA 6% 17% Irving, TX 11% 22% Jackson, MS 22% 29% Jersey City, NJ 58% 52% Kansas City, MO 25% 29% Knoxville, TN 14% 24% Lafayette, LA 20% 30% Lancaster, CA 21% 20% Lansing, MI 20% 34%

38Lexington-Fayette, KY 13% 29% Little Rock, AR 22% 24% Livonia, MI 6% 7% Long Beach, CA 33% 24% Los Angeles, CA 30% 32% Lowell, MA 30% 31% Madison, WI 20% 28% Manchester, NH 16% 23% McAllen, TX 15% 25% Memphis, TN 29% 35% Milwaukee, WI 38% 34% Minneapolis, MN 29% 25% Mobile, AL 24% 24% Modesto, CA 20% 19% Montgomery, AL 23% 25% Moreno Valley, CA 18% 17% New Haven, CT 35% 51% New Orleans, LA 47% 44% New York, NY 68% 64% Newark, NJ 54% 52% Newport News, VA 24% 27% Norfolk, VA 23% 39% North Las Vegas, NV 21% 16% Norwalk, CA 17% 14% Oklahoma City, OK 16% 19% Ontario, CA 15% 25% Orange, CA 14% 3% Oxnard, CA 14% 16% Palmdale, CA 19% 15% Pasadena, CA 26% 21% Pasadena, TX 19% 21% Paterson, NJ 42% 51% Peoria, IL 25% 23% Philadelphia, PA 50% 49% Pittsburgh, PA 37% 41% Plano, TX 6% 17% Pomona, CA 15% 12% Providence, RI 26% 47% Provo, UT 4% 22% Raleigh, NC 10% 22% Rancho Cucamonga, CA 6% 11% Reno, NV 22% 23% Richmond, VA 35% 48%

39 Riverside, CA 17% 20% Rochester, NY 38% 28% Rockford, IL 21% 14% Sacramento, CA 19% 28% Saint Louis, MO 39% 43% Saint Paul, MN 25% 27% Salinas, CA 10% 18% Salt Lake City, UT 16% 37% San Bernardino, CA 27% 23% San Buenaventura (Ventura), CA 12% 25% San Francisco, CA 35% 54% Santa Ana, CA 19% 15% Santa Clarita, CA 9% 19% Savannah, GA 29% 41% Seattle, WA 20% 41% Simi Valley, CA 8% 15% Sioux Falls, SD 6% 17% South Bend, IN 20% 26% Spokane, WA 21% 28% Springfield, IL 26% 19% Springfield, MA 39% 35% Springfield, MO 10% 23% Stamford, CT 27% 29% Sterling Heights, MI 10% 10% Stockton, CA 23% 23% Syracuse, NY 35% 36% Tacoma, WA 18% 24% Thousand Oaks, CA 6% 18% Toledo, OH 24% 30% Torrance, CA 10% 24% Tulsa, OK 19% 19% Vancouver, WA 13% 23% Virginia Beach, VA 10% 19% Warren, MI 6% 13% Washington, DC 48% 47% Waterbury, CT 39% 33% West Covina, CA 20% 11% Wichita, KS 13% 23% Winston-Salem, NC 23% 25% Worcester, MA 28% 21% Yonkers, NY 53% 42% Tabulated from the 5 percent Census PUMS files, 2000.

40 Table 4 Percenta ge of Residents without Household Access to Automobile b y Race and Ethnicit y: The Nation as a Whole, the New Orleans Metr opolitan Area, and the City of New Orleans Panel A: All U.S. Residents All Residents White, non-Hispanic Black, Non-Hispanic Other, non-

Hispanic Hispanic All 7.8% 4.6% 19.0% 9.6% 13.7% Poor a 20.4% 12.1% 33.4% 20.9% 25.0% Near Poor b 12.4% 9.0% 21.4% 14.6% 14.0% Non Poor c 4.0% 2.7% 9.9% 5.5% 7.8% Panel B: The New Orleans Metropolitan Area All Residents White, non-Hispanic Black, Non-

Hispanic Other, non-

Hispanic Hispanic All 14.1% 4.7% 27.3% 6.8% 10.2% Poor a 35.8% 13.3% 46.7% 17.8% 22.1% Near Poor b 18.9% 10.5% 27.2% 7.5% 8.5% Non Poor c 4.9% 2.5% 10.7% 2.7% 7.1% Panel C: The City of New Orleans All Residents White, non-Hispanic Black, Non-

Hispanic Other, non-

Hispanic Hispanic All 25.5% 9.7% 32.7% 9.1% 20.9% Poor a 46.7% 17.4% 52.4% 22.1% 40.2% Near Poor b 30.2% 22.1% 33.4% 3.5% 21.0% Not Poor c 10.0% 5.9% 13.2% 4.9% 11.5% Tabulations from the 5 % Public Use Microdata Sample of the U.S. Census of Population and Housing. a. Individuals in households with incomes under 100 percent of the federal poverty line. b. Individuals in households with incomes between 100 and 200 percent of the federal poverty line. c. Individuals in households with incomes greater than 200 percent of the federal poverty line.

41 Table 5 Basic Demographic and Socioeconomic Characteristics of Residents in Households With and Without Access to Au tomobiles in 2000 The Nation as a Whole, the New Orleans Metr opolitan Area, and the City of New Orleans Panel A: All U.S. Residents Share of All ResidentsShare of Residents in Households with a Car Share of Residents in Households with no Car White, non-Hisp.

69.1% 71.5% 40.6% Black, non-Hisp.

12.5% 11.0% 30.5% Other, non-Hisp.

4.6% 4.5% 5.7% Hispanic 12.5% 11.7% 22.0% 17 and Under 25.6% 25.7% 24.8% 65 and Over 11.8% 11.1% 19.9% Homeowners 67.3% 70.6% 28.2% Poor a 14.6% 12.6% 38.4% Near poor b 16.9% 16.0% 26.7% Panel B: The New Orleans Metropolitan Area All Residents In Household with a Car In Household with no

Car White, non-Hisp.

52.2% 57.9% 17.8% Black, non-Hisp.

39.6% 33.5% 76.9% Other, non-Hisp.

2.8% 3.0% 1.3% Hispanic 4.7% 4.9% 3.5% 17 and Under 26.6% 25.5% 33.2% 65 and Over 10.8% 10.2% 14.7% Homeowners 63.0% 69.2% 25.0% Poor a 20.8% 15.5% 53.0% Near poor b 19.3% 18.2% 26.0% Panel C: The City of New Orleans All Residents In Household with a Car In Household with no

Car White, non-Hisp.

26.9% 32.6% 10.2% Black, non-Hisp.

67.0% 60.5% 86.1% Other, non-Hisp.

2.4% 3.0% 0.9% Hispanic 3.1% 3.3% 2.6% 17 and Under 26.8% 24.1% 34.5% 65 and Over 11.1% 10.6% 12.8% Homeowners 47.6% 56.8% 20.6% Poor a 30.5% 21.8% 55.9% Near poor b 21.5% 20.1% 25.4% Tabulations from the 5 % Public Use Microdata Sample of the U.S. Census of Population and Housing. a. Individuals in households with incomes less than 100 percent of the federal poverty line. b. Individuals in households with incomes between 100 and 200 percent of the federal poverty line.

42 Figure 1 Fr eq uen cy of U.S. M e t r op ol i t an A r eas b y t h e P e r cent o f t h e R e s i den t P o p u l a t i o n w i t h N o A ccess t o a H o useh ol d A u t o m o b i l e 0 20 40 60 80 100 120 140 160 180 200 0 t o 4%4% t o 8%9% t o 12%1 3% t o 16%Gr eat er t h an 16%Pe r c e n t o f R e s i d e n t s w i t h N o A c c e s s t o a n A u t o m o b i le# o f M e t r o p o l i t an A r eas N e w O r l eans Figure 2 Fr equ e ncy o f U.S. C i t i es by t h e P e r cent o f t h e R esi den t P o p u l a t i o n w i t h N o A ccess t o a H o us e h ol d A u t o m o b i l e 33 66 21 10 11 44 2 0 10 20 30 40 50 60 70 0 t o 5%6% t o 10%11% t o 15%16% t o 20%21% t o 2 5%25% t o 30%3 0% t o 35%G r eat er t han 35%P e rc e n t of R e s i d e nt s w i t h n o A c c e s s t o a H o us e h o l d A u t o m o b i l e# o f C i ti es Ne w O r l e an s 43 Figure 3 D i st r i bu t i o n o f U.S.B l ack R esi d e n t s A c r o ss M e t r o p o l i t an A r eas b y S c o r e s o f t h e B l ac k-W h i t e D i ssi m i l a r i t y I n d ex, 2 000 0.16 0.99 1.22 4.85 4.83 7.3 4 13.66 22.25 13.41 18.63 8.78 3.86 0 5 10 15 20 25 0.26 t o 0.30 0.31 t o 0.35 0.36 t o 0.40 0.41 t o 0.45 0.46 t o 0.50 0.5 1 to 0.55 0.56 t o 0.60 0.61 t o 0.6 5 0.66 t o 0.70 0.71 t o 0.75 0.76 t o 0.80 0.81 t o 0.85 B l a c k-W h it e D i s s im il a r i t y In d e x Percent of Metropolitan Area s M edi an = 0.645 N e w O r l eans = 0.687 Figure 4 P r op or t i on of H ous e hol ds w i t hout a C a r i n t h e Ty pi c a l N e i g hbor h ood of A l l R e s i de nt s , W h i t e R esi d e n t s, an d B l ack R esi d e n t s f o r t h e E n t i r e U.S. an d f o r N e w O r l e a n s, 200 0 0.10 1 0.078 0.197 0.154 0.087 0.26 0 0.05 0.1 0.15 0.2 0.25 0.3 Al l W h i t e Bl a c k Proportion of neighbors with no c a Nat i onal A v er age New O r leans C M S A 44 Figure 5 A v e r a g e N u m b er of Per s ons pe r N e i g hbor h ood C a r f o r A l l R e s i de nt s , W h i t e R e s i den t s, and B l a ck R esi de n t s f o r t h e E n t i r e U.S. an d f o r N e w O r l e a n s , 2 000 2.2 1.9 3.5 2.3 1.7 3.3 0 0.5 1 1.5 2 2.5 3 3.5 4 A l l W hi t e B l ac k N e ighborhood Persons per C a N a t i onal A v er age N e w O r l eans CM S A 45 Figure 6 Percent Bla c k by Cens us Tract for the City of New Orlea n s 46 Figure 7 47 Figure 8 S c a t t e r P l o t of t h e P r opor t i on of H ous e hol d s w i t h N o C a r i n t h e C e ns us Tr a c t of t h e A v e r a g e B l ack R e s i d e n t A g ai n s t t h e D e g r ee o f B l ack-W h i t e S e g r eg a t i o n b y M e t r o p o l i t a n A r ea, 200 0 y = 0.5447 x 2 - 0.270 1x + 0.1201 R 2 = 0.4709 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.20.3 0.40.5 0.60.7 0.80 D i s s im i l a r it y In d e x , W h it e R e la t i v e t o B l a c k P r oport i on of Carl ess Househol ds i n t h e Tract of t h e Average Bl a Resi dent.9 Figure 9 D i f f e r e n ce i n t h e P r o p o r t i o n o f H o us eh o l d s w i t h N o C a r i n t h e T r act s o f t h e A ver ag e B l a ck a n d W h i t e M e t r o p o l i t an R esi d e n t s A g a i n s t t h e B l a ck-W h i t e D i s s i m i l ar i t y I n d e x , 2 000 y = 0.4907 x 2 - 0.2293 x + 0.0457 R 2 = 0.5595-0.0 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 D i s s imi l a r it y In d e x , W h i t e R e la t i v e t o B l a c k Difference in the Proportion of Households Carless, Black-Wh i