35
Drivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities Daria Zubareva, Leonid Polishchuk National Research University – Higher School of Economics, Moscow Abstract Social capital, broadly understood as the capacity for collective action based on trust, norms, and networks, is shown to be instrumental for the quality of public sector governance and service delivery, and for the efficient use of open access resources (“commons”). Both of these reasons make one to expect that social capital should have a tangible impact on safety and comfort of urban driving: drivers’ courtesy, mutual respect and compliance with traffic rules relieve congestion and prevent accidents, whereas civic culture and political activism improve accountability of municipal governments and hence bode well for urban planning, road maintenance, and traffic management and control. These hypotheses are tested in the paper for a sample of twenty Russian cities, all of which show signs of increased traffic congestion and accident rates caused by steep rise of car ownership in the country. We find evidences that higher stocks of social capital are indeed associated with lesser traffic jams and fewer accidents, and hence with cheaper vehicle insurance. Pro-social norms and values have a tangible impact on drivers’ behavior on the roads, putting the existing transportation networks – “urban commons” – into better use. Accident rate shows significant sensitivity to social capital, whereas traffic congestion’ elasticity to social capital is less pronounced. Traffic jams in Russian cities result primarily from the capacity bottlenecks, rather than from how this capacity is used by motorists. Therefore an answer to congestion problems should be sought in improved public policies, infrastructure investments, and road planning, and

 · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

  • Upload
    others

  • View
    3

  • Download
    0

Embed Size (px)

Citation preview

Page 1:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

Drivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities

Daria Zubareva, Leonid Polishchuk

National Research University – Higher School of Economics, Moscow

Abstract

Social capital, broadly understood as the capacity for collective action based on trust, norms, and networks, is shown to be instrumental for the quality of public sector governance and service delivery, and for the efficient use of open access resources (“commons”). Both of

these reasons make one to expect that social capital should have a tangible impact on safety and comfort of urban driving: drivers’ courtesy, mutual respect and compliance with traffic rules relieve congestion and prevent accidents, whereas civic culture and political activism improve accountability of municipal governments and hence bode well for urban planning,

road maintenance, and traffic management and control.

These hypotheses are tested in the paper for a sample of twenty Russian cities, all of which show signs of increased traffic congestion and accident rates caused by steep rise of car

ownership in the country. We find evidences that higher stocks of social capital are indeed associated with lesser traffic jams and fewer accidents, and hence with cheaper vehicle

insurance. Pro-social norms and values have a tangible impact on drivers’ behavior on the roads, putting the existing transportation networks – “urban commons” – into better use.

Accident rate shows significant sensitivity to social capital, whereas traffic congestion’ elasticity to social capital is less pronounced. Traffic jams in Russian cities result primarily

from the capacity bottlenecks, rather than from how this capacity is used by motorists. Therefore an answer to congestion problems should be sought in improved public policies, infrastructure investments, and road planning, and key to such problems is therefore in the

hands of citizens, rather than drivers. Our data show however that the Russian society has yet to made its voice heard in addressing urban transportation problems – we found no significant

relationship between civic culture and awareness in Russian cities, on the one hand, and driving conditions in these cities, on the other.

The paper shows that social capital is indeed a valuable resource in resolving increasingly severe transportation problems, and that both “transmission channels” – horizontal and

vertical – between social capital and economic outcomes need to be engaged. The horizontal channel improves the situation at the grassroots through better use of “urban commons”, whereas the vertical one requires political collective action to ensure proper public policy

responses to mounting road congestion.

1. Introduction

Proverbial problems of Russian roads have become more acute in recent years, when steeply

increasing traffic and car ownership stretched to the limit the capacity of inter- and intra-city

Page 2:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

road networks. Presently roads are one of the most severe bottlenecks to urban development

in Russia, causing massive air pollution, losses of time and increasingly of human life.

Traffic jams are now common features of major Russian cities, and with the number of

victims of car accidents over 70 per 100,000 cars a year, Russia is in a tie with countries

notorious for deplorable and unsafe traffic conditions.

Measures proposed to deal with this problem include expansion of road networks, toll roads,

greater reliance on public transit, various investments in road infrastructure, stricter

enforcement of traffic rules, etc. In this paper we’re interested in the role of the ‘human

factor’ in traffic jams and road accidents. More precisely, our goal is find out to what degree

Russian road problems are associated with and could be explained by norms, values and

behavioral patterns in the Russian society, comprising what is known as social capital.

Social capital’s main function is to facilitate collective action when individual choices and

behavior need to be coordinated for the sake of common good. Two types of collective action

are relevant for road conditions. The first is proper use of open access collective resources

known as the commons. Lack of coordination in commons leads to their congestion when

individual users crowd out each other; traffic jams and accidents illustrate this general pattern

in the case of urban commons – inter-city road networks. The second type is participation in

democratic processes to ensure government accountability and proper delivery of public

services, in the present context – urban planning and management, including construction and

maintenance of roads, and effective enforcement of traffic rules. Political awareness and

activism are powered by a special type of social capital known as civic culture.

The first type of social capital is essential for the proper use of the existing road networks and

operates on the demand side of urban traffic systems; it affects behavior of drivers. One

should expect that pro-social norms result in responsible driving, consideration for others,

mutual help and respect among motorists, compliance with traffic rules and joint efforts to

Page 3:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

resolve traffic problems if they should occur. The second type is required for adequate supply

of urban traffic infrastructure, and shapes behavior of citizens.

In this paper we analyze and measure the impact of social capital on road conditions in

Russian cities. Main outcomes of social capital are convenience and safety of urban road

transportation; opposite measures of these outcomes are resp. severity of traffic jams and

various indicators reflecting accident rates. Social capital of drivers and citizens could be a

priory relevant for both of those outcomes; our task hence is to establish whether this is

indeed the case.

The paper contributes to the growing strand of literature studying the impact of social capital

on economic development and welfare, government performance, social service delivery, etc.

Surprisingly, urban traffic so far has not been properly represented in such studies, and our

paper partly fills this gap. One of the very few contributions on this subject is by Fishman and

Miguel (2007) who established an association between violations of parking rules by

diplomats based in New York City with corruption levels and ultimately cultural traits in their

home countries. In another paper Incla et al. (2005) reveal a link between traffic accidents

and casualties thereof, on the one hand, and propensity for cooperation in local communities

in Mexico, on the other. It is argued in the paper that due to a lack of reciprocity and capacity

for collective action residents fail to perceive high traffic mortality as a common problem and

hence are unable to jointly take measures to resolve it.

The main findings of our paper are as follows. Based on empirical data, we show that social

capital based on pro-social norms and values indeed reduces accident rates in Russian cities,

and that such conclusion is robust to different accident measures and estimation strategies.

Here social capital works at the grassroots through a ‘horizontal channel’ At the same time

the impact of such social capital on traffic jams is much milder and often insignificant. This

leads us to conclude that while drivers’ behavior (and norms and values that it reflects) is a

Page 4:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

major cause of accidents, the roots of traffic jams are primarily on the supply side of urban

transportation in Russia, and keys to this problem are in improved urban management and

planning, i.e. better municipal governance. Solving this problem thus requires a different type

of social capital, which is based on civic culture and operates through a ‘vertical channel’

linking social capital and governance. While a sister paper (Menyashev, Polishchuk, 2011)

shows that such traits exhibit significant variations from one Russian city to another and have

substantial impact on life satisfaction and governance in Russia cities, in this particular

instance we do not observe a statistically significant relation between civic culture and

behavior on the one hand, and traffic jams, on the other. A plausible explanation is that

immature Russian democracy does not generate pressure on municipal governments strong

enough to counteract vested interests that push for lucrative development projects and neglect

everyday needs of ordinary citizens. To show that under more democratic conditions the

expected link exists, we use US data and observe a strong relationship between traffic jams in

US cities and participation in elections of these cities’ population, which is a telltale sign of

civic culture.

2. Data

The main source of empirical information for this study was a survey of drivers in 20 Russian

cities, 80 respondents per city, conducted in the summer 2010. City level is an appropriate

unit for our analysis, since this is a natural confine for urban traffic. Furthermore, since road

maintenance and development is the responsibility of urban governments, cities internalize

the link between citizens and governments with regard to road conditions. Finally, using

cities rather than countries in this sort of analysis makes it easier to meet the ‘ceteris paribus’

requirement. Cross-country studies would suffer from an obvious omitted variable bias, for

which control variables would offer only partial remedy. Cross-country differences in road

outcomes are obviously driven by road institutions, traffic rules and their enforcement,

Page 5:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

composition of the traffic, role of public transit, etc. Studying various cities in a single

country alleviates this problem.

Traffic outcomes

We measure two traffic outcomes, accidents and road congestion. Our measure of road

congestion is based on survey questions where we ask drivers about relative increases of their

travel time from work to home in rush hours as opposed to off-pick hours.

We employ several measures of road accidents. First, we collect official statistics of road

accidents and road injuries from the State Inspectorate for Road Traffic Safety (GIBDD).

Specifically, we collect regional data on the number of car accidents per 10000 cars

(ac_10_cars) and the number of people killed in car accidents per 100000 people

(ac_100_pop). The potential problem with these official statistics is that they may be biased

because of incomplete reporting by police officers who might be trying to improve their

performance records by concealing accidents and/or underestimating their severity. To

complement official data by an independent and potentially more reliable source, we also

used regional car insurance premiums which reflect market assessment of the real risk of road

accidents. Both the official statistical data and the automobile insurance premiums are

available only for regions (Russia’s subnational units) rather than cities, but this does not

pose a serious problem, since most Russian motorists live in the capital cities of their region

and the lion’s share of road trips takes place within the confines of city borders. As a yet

another measure, we also use individual assessments by survey respondents who were asked

about the number of accidents they personally witnessed over a period of time. While being

subjective and prone to potential biases, this is the only measure that we have which is

available at the city level.

Table 1: Summary statistics for the traffic outcomes

Page 6:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

Variable Obs Mean Std. Dev Min Maxac_10_cars 20 25.2 6.078608 15.7 39.3ac_100_pop 20 100.57 24.11368 55.9 142.4Kasko 20 31018.75 4319.777 25060 41895Observed accidents 20 3.218855 0.25202 2.833333 3.6875congestion 20 1.687454 0.415723 1.180132 2.596372

The above table shows that there is a wide variation across cities in traffic outcomes. Market

prices of car insurances reflect these variations and the highest premium is almost twice as

high as the lowest one. Overall the table reveals highly unsatisfactory traffic conditions in

Russian cities; the number of car accidents and of accident victims are very high by

international standards.

Traffic behavior

We measure traffic behavior by asking drivers to assess the behavior of other drivers in the

same city and averaging their answers at the city level. Here’s a sample of driving behavior

patterns registered by the survey

running yellow and red lights (s_slip_yell)

illegal parking obstructing traffic (s_park_proh);

bypassing traffic jam on sidewalk or road shoulder (s_jam_side);

hit and run (s_leave_acc)

causing traffic jams by blocking intersection (s_crossr_m~e)

yield to a pedestrian (s_skip_ped~r)

yield to a car which changes lines or entering highway traffic (r s_skip_byway)

help other drivers when their car is stuck in snow or mud (s_help_pull)

help other drivers when they experience technical problems (s_help_rep)

A summary statistics for city averages is presented below.

Page 7:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

Table 2: Summary statistics of driving culture

Variable Obs Mean Std. Dev Min Max

s_slip_yell 200.48039

30.08511

30.22530

90.58950

6s_park_proh 20

0.498956

0.099947 0.16358

0.604651

s_jam_side 200.41552

20.08995

50.13888

9 0.625

s_leave_acc 200.37520

40.08140

80.10915

50.50632

9s_crossr_m~e 20

0.424189

0.094001

0.166667

0.587209

s_skip_ped~r 20

0.590299

0.087597

0.444767

0.777778

s_skip_byway 20

0.438041

0.063901

0.324074 0.63141

s_help_pull 200.36891

80.05638

20.28947

40.50316

5

s_help_rep 200.29718

80.05625

50.21202

5 0.42284

Variations of city averages of the above indexes is surprisingly large; for some measures the

ratio of maximum to minimum could be several hundred percent. Some of the measures are

significantly correlated with each other, and we perform factor analysis to compress the list.

The first two factors (principal components) explain 84% of the total variation and are as

follows.

Table 3: Factor analysis of driving culture

Variable Factor1 Factor2Uniqueness

s_slip_yell 0.8975 0.2386 0.1376s_park_proh 0.8914 0.1003 0.1953s_jam_side 0.8268 0.2947 0.2295s_leave_acc 0.8728 0.2821 0.1587s_crossr_m~e 0.9204 0.0911 0.1446s_skip_ped~r -0.4612 0.551 0.4837s_skip_byway -0.4595 0.3674 0.6538s_help_pull -0.2223 0.9359 0.0747s_help_rep -0.3094 0.7961 0.2706

Page 8:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

The loadings of the first factor suggest that it is a proper measure of ‘free-riding’ (pun

intended) behavior among drivers. Indeed, it loads strongly and positively on questions that

capture disregard of interests of fellow motorists and damping costs of such behavior on

others.

In Figure 1 we plot the answers to the questions that load strongly on the first factor. It shows

considerable differences in the incidence of ‘free-riding’ across cities in our sample.

Figure 1: Motorists’ free-riding

0.2

.4.6

Arkhang

elsk

Barnau

l

Ekater

inburg

Kalinin

grad

Kazan

Kemerov

o

Krasnod

ar

Krasnoy

arsk

Lipets

k

Novos

ibirsk

Perm

Rostov-o

n-Don

Saratov

Tolyatt

i

Tomsk

Tver

Vladivo

stok

Volgog

rad

Vorone

zh

Yarosla

vl

run yellow or red lights illegal parking obstructing traffic

by passing traffic jam on sidewalk hit and run

blocking road intersection

The second factor in table 2, according to its loading, can be interpreted as road courtesy; it

also varies in a broad range in our sample.

Figure 2: Motorists’ courtesy

Page 9:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

0.2

.4.6

.8

Arkhang

elsk

Barnau

l

Ekater

inburg

Kalinin

grad

Kazan

Kemerov

o

Krasnod

ar

Krasnoy

arsk

Lipets

k

Novos

ibirsk

Perm

Rostov-o

n-Don

Saratov

Tolyatt

i

Tomsk

Tver

Vladivo

stok

Volgog

rad

Vorone

zh

Yarosla

vl

yield to a car which changes yield to a pedestrian

help other drivers when their car is stuck help other drivers when they experience technical problems

Personal values

Behavioral patterns revealed by our survey could be expected to reflect drivers’ values. The

survey does not register respondents’ values, and we proxy those by data from another survey

(GeoRating2007) with a focus on values and attitudes of residents of Russian cities. It

appears that values which are commonly associated with social capital, such as trust, mutual

respect, help to each other and solidarity in the society have significant correlations of the

expected positive sign with the factor representing motorists’ courtesy, while their correlation

with the free riding factor, although also of the expected (this time negative) sign, is much

less pronounced.

Table 4: Values and road behavior

Courtesy traffic

behaviorFree riding traffic

behaviorRespect1 0.38 -0.16Consensus and cohesion 2 0.28 -0.11

1 Respect was measured by the question “How common is respect towards people in your city?”2 Consensus and cohesion was measured by the question “Is there more consensus and cohesion or more

Page 10:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

Willingness to help3 0.25 -0.03

The table points out to normative roots of motorists’ behavior, although more research is

required to better visualize such links.

Political attitudes and associational activities

To find evidence of a vertical channel at work linking social capital with traffic outcomes in

Russian cities, we used the following data characterizing political attitudes and behavior:

participation in elections;

membership in automobile associations ;

willingness to participate in protest rallies against construction of a new shopping mall at a

place which already suffers from acute traffic problems;

willingness to go to court in case of violation of motorists’ rights by traffic police or road

service providers

The first two measures are drawn resp. from official electoral statistics and automobile

associations sourcses, whereas the two remaining ones are based on our survey data.

The above measures reveal low levels of political activities and association membership in

Russia, which are consistent with the general perception of the modern Russian society as a

largely apolitical one. Thus, 50% of respondents are not willing to take part in any protest

action, and only 15% express willingness to take part in public rallies (Figure 3). Less than

3% of respondents are members of automobile associations (although associational activity

among Russian car owners is presently on the rise, in response to worsening road conditions,

abuse by road police , and driving privileges of bureaucracy).

disagreement and disunity in our country nowadays?”3 Willingness to help was measured by the question “How often one can observe the willingness to help people around?”

Page 11:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

Figure 3: Political attitudes

3. Empirical results

In this section we present estimations of econometric models which relate traffic outcomes to

the above described indicators of social capital. Unless explicitly stated otherwise,

observation units in our sample are cities. Small size of the sample (20 cities) poses an

estimation problem; cognizant of this problem, we include in our regressions not more than

three explanatory and control variables, changing the composition of those to check

robustness.

Road accidents and the horizontal channel

We begin with estimation of regression equations

traffic outcomes=α+β1 Horizontal+γi controli+εi

which relate traffic outcomes to road behavior. We use three indexes of road accidents, the

first – automobile insurance premiums (Table 4), the second – official accidents statistics

Page 12:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

(Table 5), and the third – accident rates reported by survey respondents (Table 6). Our main

explanatory variable is the free riding indicator as the prime ‘suspect’ in automobile

accidents. For measuring free riding traffic behavior we use the variable “free_ride”

constructed by factor analysis (higher value of the variable indicates more free riding on the

road). For measuring courtesy we use the second factor derived by factor analysis (higher

value of the variable indicates grater incidence of pro-social behavior on the roads).

We use in regressions various control variables. Some of them are obtained from our survey,

such as drivers’ assessment of road surface quality (est_pavmnt), urban planning

(est_plan_inters) and the quality of road police work (gai_viol). Other control variables are

obtained from official statistical sources (GosKomStat for 2009). We use bank deposits per

capita (depos) and budgetary funding per 1000 people (budg_1000) as proxies for economic

well-being. The last control variables are the number of cars in the city (cars) and the number

of cars per 1000 people (cars_on_1000_pop); these date are provided by the Autostat agency.

Table 4: Impact of driving behavior on car accidents (dependent variable – car insurance premium)

  (1) (2) (3) (4) (5) (6) (7) (8)                 free_ride 2,672*** 2,108*** 2,863*** 2,706*** 2,670*** 2,715*** 2,584*** 2,397**

(885.0) (699.8) (903.7) (877.0) (893.1) (901.3) (855.8) (851.3)respect -442.6

(896.5)cars_on_1000_pop -1.271 -1.478 -2.341 -0.423 -1.504 -1.139 1.345 1.316

(9.866) (7.606) (9.771) (9.856) (9.938) (9.973) (9.793) (9.451)depos 0.455

(0.265)budg_1000 -0.151

(0.183)child_high_1000 -22.24

(29.47)est_pavmnt -24.54

(2,079)est_plan_inters -780.6

(2,532)gai_viol 5,559

(4,671)bus_1000 42,227

(28,115)Constant 31,295*** 26,665*** 32,295*** 33,229*** 31,420*** 33,297*** 18,757 28,568***

Page 13:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

(2,807) (3,138) (2,982) (3,708) (4,941) (6,846) (10,939) (3,229)

Observations 19 18 19 19 19 19 19 19R-squared 0.385 0.476 0.402 0.397 0.375 0.379 0.429 0.456

Table 5: Impact of driving behavior on car accidents (dependent variable – number of car accidents per 1,000 cars)

  (17) (18) (19) (20) (21) (22) (23) (24)

                 free_ride 2.792** 2.735** 2.529** 2.903*** 2.766** 2.508** 2.816** 2.408**

(1.050) (1.124) (1.087) (0.961) (1.068) (0.994) (1.083) (1.000)respect 0.991

(1.064)cars_on_1000_pop 0.0120 0.0115 0.0136 0.0156 0.0127 0.0102 0.0119 0.0165

(0.0117) (0.0122) (0.0117) (0.0108) (0.0119) (0.0110) (0.0124) (0.0111)depos 0.000361

(0.000425)budg_1000 0.000209

(0.000220)child_high_1000 -0.0646*

(0.0323)est_pavmnt 1.559

(2.485)est_plan_inters 5.003*

(2.791)gai_viol -1.198

(5.910)bus_1000 60.39*

(33.02)Cars

Constant 21.23*** 17.96*** 19.78*** 26.46*** 18.02*** 8.730 23.78 17.05***(3.332) (5.039) (3.585) (4.065) (5.906) (7.548) (13.84) (3.792)

Observations 19 18 19 19 19 19 19 19R-squared 0.377 0.364 0.378 0.479 0.357 0.457 0.342 0.461

Table 6: Impact of driving behavior on car accidents (dependent variable – accidents reported by respondents)

  (25) (26) (27) (28) (29) (30) (31) (32)

                 free_ride 0.135** 0.133** 0.153** 0.129** 0.134** 0.140** 0.137** 0.145**

(0.0542) (0.0566) (0.0540) (0.0527) (0.0512) (0.0542) (0.0536) (0.0539)respect 0.00136

(0.0584)depos -0.068

Page 14:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

(0.0209)budg_1000 -0.0125

(0.0108)child_high_1000 0.00185

(0.00182)est_pavmnt -0.170

(0.128)est_plan_inters -0.0777

(0.150)gai_viol -0.168

(0.284)bus_1000 -1.470

(1.731)cars 5.51e-07 5.15e-07 6.55e-07 3.32e-07 2.06e-07 5.30e-07 4.89e-07 5.16e-07

(6.70e-07) (6.37e-07) (6.01e-07) (6.35e-07) (6.40e-07) (6.15e-07) (6.20e-07) (6.07e-07)Constant 3.117*** 3.275*** 3.153*** 2.983*** 3.514*** 3.322*** 3.488*** 3.195***

(0.148) (0.237) (0.137) (0.189) (0.327) (0.417) (0.639) (0.164)

Observations 19 18 19 19 19 19 19 19R-squared 0.302 0.301 0.360 0.347 0.375 0.315 0.318 0.334

In all estimations free riding has the expected positive coefficient and is almost invariably

significant at 0.01-0.05 level; given small size of the sample, this effect is pronounced quite

clearly. It is particularly strong when we use the market-based accident index, i.e. car

insurance premium (Table 4), when free riding in all but one specification is significant at

the 0.01 level. Various controls do not diminish this significance and only modestly change

the coefficient for free riding, which attests to robustness of the established effect (Figure 4).

Figure 4: Automobile insurance premiums and road behavior

Page 15:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

-500

00

5000

1000

0A

utom

obile

insu

ranc

e pr

emiu

ms

-3 -2 -1 0 1free riding traffic behavior

coef = 2108.3457, se = 699.83445, t = 3.01

*) Control variables – number of cars per 1,000 people and bank deposits per capita

We obtain qualitatively similar results for other measures of accidents (Tables 5 and 6),

albeit with lower significance, which could be ascribed to noise in used measures due to

either imprecision and/or incompleteness of survey data, or falsification in official data.

The second factor (principal component) of road behavior – respect – has in our estimations

the expected sign, but its significance is considerably smaller than the first one. More

research is required to explain this observation.

Given the importance of the link between accidents and driving behavior powered by social

capital, we have performed a yet another robustness check, this time using individual data

rather than city averages. An important advantage of this specification strategy is a vastly

larger number of observations that considerably improves the odds to get statistically

significant estimators. We can also include in regression equations a much larger number of

control variables. Furthermore, there could be inter-city variations of driving patterns and

accident rates that would not be captured in regressions based on city averages. In the

regressions reported below standard errors are clustered per region. Instead of using the

Page 16:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

aggregate free riding index, we enter in regression equations as dependent variables specific

indicators of ‘free riding’.

Table 7: Impact of driving behavior on car accidents (individual responses)

  (1) (2) (3) (4)VARIABLES Accidents Accidents Accidents Accidents         s_crossr_move 0.526**

(0.196)s_slip_yell 0.433***

(0.139)s_park_proh 0.543**

(0.205)s_jam_side 0.302

(0.217)elect_yes 0.109 0.0924 0.0939 0.0930

(0.0696) (0.0748) (0.0756) (0.0767)sc_respect -0.0470 -0.0493 -0.0532 -0.0654

(0.0575) (0.0564) (0.0557) (0.0548)est_pavmnt -0.0932* -0.0880* -0.0897* -0.0851

(0.0494) (0.0465) (0.0480) (0.0495)est_plan_inters -0.0214 -0.0235 -0.0163 -0.0266

(0.0552) (0.0548) (0.0534) (0.0559)gai_viol 0.0973** 0.0952* 0.0831* 0.0865*

(0.0464) (0.0495) (0.0479) (0.0471)c_sex -0.0398 -0.0747 -0.0757 -0.0779

(0.0579) (0.0631) (0.0675) (0.0667)

c_age-

0.000347-

0.000234-

0.000607 0.000430(0.00504) (0.00454) (0.00474) (0.00451)

c_educ -0.0361** -0.0332** -0.0299* -0.0303*(0.0146) (0.0157) (0.0152) (0.0151)

c_inc 0.0351 0.0377 0.0381 0.0385(0.0250) (0.0248) (0.0248) (0.0258)

c_expr 0.000665-

0.000111 0.000381 -0.00129(0.00480) (0.00451) (0.00456) (0.00444)

c_freq -0.140** -0.149** -0.153** -0.146**(0.0575) (0.0559) (0.0538) (0.0582)

car_price 0.0312 0.0414 0.0347 0.0326(0.0228) (0.0240) (0.0223) (0.0235)

c_need 0.0441* 0.0492* 0.0456* 0.0539**(0.0242) (0.0247) (0.0233) (0.0252)

Constant 3.137*** 3.160*** 3.143*** 3.283***(0.449) (0.442) (0.432) (0.461)

Observations 1,180 1,217 1,221 1,214R-squared 0.079 0.074 0.084 0.067Robust standard errors in parentheses

Page 17:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

*** p<0.01, ** p<0.05, * p<0.1

Once again free riding indexes have the expected positive sign, and in three out of four

specifications are significant at the 0.01-0.05 levels. Control variables also have expected

signs. It is noteworthy that some of these controls become significant, such as quality of the

road surface (est_pavmnt) and the performance of road police.

Traffic jams and the horizontal channel

While driving behavior has a strong impact on road accidents which is consistently robust

across all o our specification, its influence, if any, on traffic jams is harder to discern. In our

estimations (which we do not report here) with traffic jams as the dependent variable driving

behavior (free-riding and respect measured as results of the factor analysis fact) has the

expected sign but is never significant.

Our analysis thus indicates that driving behaviour, erratic and uncooperative as it may be, is

not a major cause of traffic jams plaguing Russian cities. This finding agrees with the

general dictum of the urban and transportation sciences that traffic jams require government

intervention through bottleneck removal, improved planning, maintenance, control, etc. (see

e.g. Downs, 2004). Road usage pales in its significance in comparison with the above

factors, and driving behaviour cannot be blamed for traffic jams when roads are in poor

conditions and urban transportation networks do not match traffic volumes and patterns.

As it was argued earlier in the paper, social capital can still be instrumental in such cases, but

this time it should work through the vertical channel and power not so much driving, but first

and foremost political behaviour. We now turn to search for evidence that such channel is

indeed at work in Russian cities.

Page 18:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

Traffic accidents and the vertical channel

To find out whether civic culture and associational activity have an impact on Russian road

conditions, first and foremost traffic jams, we estimate regression models

traffic outcomes=α+β1 Horizontal+β2Vertical+γ icontroli+εi,

where traffic outcome as dependent variable is measured by road congestion. We allow in

the model both types of social capital – those affecting driving behaviour and those

underlying political actions – to find out if there is any substitution between the two. As

before, we use in our specifications various control variables that could be relevant for traffic

jams, including geographic location of the city (latitude and longitude) and the year the city

was established (one could argue that in older cities road networks could be less suitable for

modern traffic needs and this could be a factor contributing to traffic jams).

Estimation results for various specifications of the above model are presented below.

Table 8: Road congestion and vertical channel

  (1) (2) (3) (4) (5) (6) (7) (8)                 free_ride 0.116 0.114 0.106 0.111 0.138 0.106 0.123 0.132

(0.0886) (0.108) (0.0948) (0.0922) (0.103) (0.0938) (0.0790) (0.0885)cars_on_1000_pop 0.00170 0.00175 0.00179 0.00190* 0.00165 0.00188 0.00110 0.00192*

(0.000986) (0.00106)

(0.00106) (0.00105)

(0.00104) (0.00109)

(0.000914)

(0.000980)

est_pavmnt -0.287(0.206)

associations 0.0831(1.220)

sit_court_gai 0.127(0.476)

sit_court_serv 0.263(0.334)

elections_yes -0.528(0.835)

moll_const 0.179(0.399)

est_plan_park -0.567**(0.225)

est_plan_inters -0.385(0.249)

Constant 1.797*** 0.916 0.971 0.655 1.512** 0.453 2.636*** 2.185***

Page 19:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

(0.490) (4.724) (1.040) (0.795) (0.525) (1.774) (0.608) (0.673)

Observations 19 19 19 19 19 19 19 19R-squared 0.304 0.214 0.218 0.245 0.235 0.225 0.448 0.322

In specifications 2 to 6 we introduce all measures of social capital that is expected to be

transmitted through the vertical channel. Their coefficients are never significantly different

from zero. Therefore we can reject the hypothesis the social capital through vertical channel

alleviates urban traffic problems in Russia. The above regressions however once again

confirm that more accountable municipal governance could be a potent resource of

improvement of urban traffic conditions: in specification (7) availability of parking is shown

to significantly (at 0.05 level) contribute to the alleviation of traffic jams. If social capital (in

the form of civic culture and political participation) is in sufficient supply, municipal

authorities would be more visibly and effectively involved in improving traffic conditions in

their cities.

The likely reason that we do not observe such relationship on our data is the low level of

civic culture in the Russian society, which could be ascribed to Russia’s political history

(Tabellini, 2008) and also perhaps to the recent political trends in the country which put

direct subordination of regional and municipal administrations to governments of higher tiers

ahead of direct accountability of subnational governments to people.

To show that the vertical channel could be working under more democratic conditions and

stronger civic culture we turn to US data and find a strong and robust relationship between

traffic jams in US’s major metropolitan areas4 and political activity and mobilization of local

population measured by election turnout. The latter exhibits strong statistically significant

negative association with incidence and severity of traffic jams, even after controlling for

4 We use road congestion measures of the Texas Transportation Institute (http://tti.tamu.edu/)

Page 20:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

population, employment rates, income, inequality and educational attainments5. Vibrant

democracy indeed converts higher stocks of civic culture into better road conditions.

Table 9: Political participation and traffic jams in US metropolitan areas

  (1) (2)VOTE -0.00941***

(0.00335)GROUP -0.00666*

(0.00358)

POPUL-9.27e-05***

(2.73e-05)unemployed 0.000397*** 0.00210***

(0.000139) (0.000490)income_p_c 5.22e-06 9.98e-06**

(3.67e-06) (3.83e-06)ed_bachelor -0.00208 -0.00236

(0.00317) (0.00301)etn_black 0.000511 -0.000270

(0.00140) (0.00127)ineq_below_poverty -0.0119** -0.0121**

(0.00500) (0.00483)Constant 1.617*** 1.097***

(0.232) (0.197)

Observations 51 51R-squared 0.504 0.563

Figure 5: Vertical channel at work in the US

5 Source: US Census Bureau

Page 21:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

-.2-.1

0.1

.2.3

cong

estio

n in

dex

-10 -5 0 5 10 15actual vote rate

coef = -.00721923, se = .00347157, t = -2.08

Endogeneity concerns

Our main results that free riding behavior is related to road accidents may be subject to an

endogeneity problem: it is plausible that on less safe roads people behave in a more

aggressive and self-interested way. The same argument can conceivably be applied to road

congestion: it cannot be ruled out that in cities with longer traffic jams local driving culture

deteriorates since incentives for co-operative behavior are greatly diminished. We observe

however no statistical relation between free riding and road congestion, which makes the

hypothesized reverse causality less likely for road congestion and therefore a fortiori also for

road accidents that are much less frequent and observable and hence less likely to affect the

local driving culture. Furthermore the quality of roads which in theory could affect driving

norms and behavior is found to be not related to the free riding behavior. All of the above

suggests that local driving cultures are exogenous to traffic outcomes and road quality and

have other, perhaps deeper, roots.

Page 22:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

4. Conclusion

The paper confirms that social capital matters for Russian road conditions, although its

performance is uneven across traffic outcome and transmission channels.

We posited that traffic outcomes are influenced by drivers through their behavior (the

horizontal channel) or by citizens through their ability to hold local governments accountable.

We find that free riding behavior on the roads is indeed related to traffic accidents. We have

several indications that this finding is not driven by reverse causality. This lends support to

the hypothesis that the horizontal channel is indeed at work in affecting traffic outcomes. The

vertical channel is however is not manifesting itself in our data, but we find strong evidence

of its presence in a sample of US cities. The observed difference should be ascribed to

different levels of civic culture and democratic participation in the two countries.

Our study shows that improved road behavior of drivers (through internalization of pro-social

norms or perhaps stricter enforcement of traffic rules) could significantly improve road safety

in Russia, but is unlikely to alleviate mounting traffic jams. Improved urban governance and

planning seems to be the key to this second problem, which requires more civic culture and

involvement that is currently observed in Russia.

References Downs A., Still stuck in traffic: coping with peak-hour traffic congestion , Brookings

Institution Press, Washington D.C., 2004

Durlauf S., Fafchamps M. Handbook of Economic Growth, Elsevier, edition 1, volume 1,

number 1, 2005.

Incla´n C., Hijar M., Tovar V. Social capital in settings with a high concentration of road

traffic injuries. The case of Cuernavaca, Mexico Social Science & Medicine 61 (2005) 2007–

2017

Page 23:  · Web viewDrivers and Citizens: The Impact of Social Capital on Traffic Conditions in Russian Cities. Daria Zubareva, Leonid Polishchuk . National Research Universi. ty – Higher

Fisman R., Miguel E. Cultures of Corruption: Evidence From Diplomatic Parking Tickets

Journal of Political Economy, 2007, vol. 115, no. 6 1020-1048

Menyashev R., Polishchuk Does Social Capital Have Economic Payoff in Russia? Working

paper WP10/2011/0, center for institutional studies, HSE.

Ostrom E. Collective Action and the Evolution of Social Norms // The Journal of Economic

Perspectives, Vol. 14, No. 3 (Summer, 2000), pp. 137-158

Putnam R., Leonardi R., Nanetti R. Making Democracy Work: Civic Traditions in Modern

Italy, Princeton University Press, Princeton (1993).

Shinar D., Traffic safety and human behavior, Elsevier, 2007

Tabellini G. Institutions and Culture // Journal of the European Economic Association, 6(2-

3), 2008, 255–294.

Woolcock, M., Narayan, D. 2000. Social capital: Implications for development theory,

research and policy. World Bank Research Observer, 15: 225-250.

World Health Organization, Global Status Report on Safety: time for action, 2009