Neighborhoods, communities, and collective goods and bads

1
Neighborhoods, communities, and collective goods
and bads
Beate Völker Dept. of Sociology, Utrecht
University, The Netherlands
2
Todays message

1. Communities in Dutch neighborhoods exist, but
   they consist of weak relationships, which are
   nevertheless often connected.
2. The association between community and
   social/physical order is only indirect the
   mechanism is informal control.
3. Context effects of social diversity on trust and
   participation are weak and not robust. Relational
   similarity and structural embeddedness are much
   more important.

3
Why sociological research in neighborhoods?

• Shift from ascribed to achieved also within
  networks?
• Assumption trend towards less cohesion and
  community implies less contacts among neighbors
• People cannot avoid neighbors what does come
  out of these opportunities for contact?
• Social integration depends largely on having
  weak ties more than on ties to family or close
  friends

4
Ongoing research projects/interests of BV

• Community and solidarity behavior in
  neighborhoods
• Economic and social conditions for individual
  well being in neighborhoods
• Community failure social and physical disorder
  in neighborhoods (1)
• Community failure Troublesome neighborhood
  relations (2)
• Trust and collective good production
  neighborhoods

4
5
Ongoing research projects/interests of BV

• Community and solidarity behavior in
  neighborhoods
• Economic and social conditions for individual
  well being in neighborhoods
• Community failure social and physical disorder
  in neighborhoods (1)
• Community failure Troublesome neighborhood
  relations (2)
• Trust and collective good production
  neighborhoods

5
6
What is a neighborhood?    

• Geographical area
• Administrative area zip code
• What people consider to be a neighborhood
• Neighbors who interact with each other more than
  with other people (who live also close by)

7
Neighborhoods in the SSND (the Survey of the
Social Networks of the Dutch)     

• Sample of 160 neighborhoods in 40 municipalities
  taking urbanization and region into account
• Neighborhood 5-position zip code area (i.e. 230
  addresses on average) this resembles the route of
  a postman
• 6-8 respondents in each neighborhood
• Neighborhood characteristics partially via
  respondents, partially via national bureau of
  statistics (CBS, wijken en buurten)
• Analyses multilevel analyses respondents nested
  in neighborhoods

8
The Survey of the Social Networks of the Dutch
(SSND) municipalities where we collected data
-
9
Facilities in the neighborhood

• Supermarket
• Butcher
• Bakery
• Green grocery
• Fish grocery
• Cinema
• Shop for building equipment
• Shop for clothes
• Market for fresh vegetables etc.
• Flowery shop
• Snack bar
• Physician/general practitioner
• Police station
• Church
• Garage/gas station

• 16. Sport field
• 17. Cafe
• 18. Restaurant
• 19. Day care center
• 20. Neighborhood center
• 21. School
• 22. Park
• 23. Swimming center
• 24. Sport- or fitness center
• 25. Post office
• 26. Bus station
• 27. Train station
• 28. Theatre, opera, concert hall
• 29. Public library
• 30. Playground

10
Who are neighbors? Exchange method name
generating questions partially standard,
partially focused on own research interests step
1

• Who did help you get your current home?
• Who has the keys to your house?
• Who do you ask for helping with odd jobs
  in/around your house?
• With whom do you discuss personal matters?
• Who are your direct neighbors?

• How did you get your current job?
• Who do you ask for advice concerning matters at
  your job? Whom do you give advice concerning
  these matters?
• With who do you have a problem?
• Who is your boss?

Neighbors enter the network in two ways via name
generating questions and via the direct question
11
Characteristics of network members and the
relation between network member and respondent
( step 2)

• Characteristics of Alter
• Sex, age, education, occupation, religion, family
  situation
• role relation with ego
• Characteristics of relationship Ego-Alter
• Intensity, trust, liking
• Duration of relationship
• Where met first, where meeting currently
• Frequency of contact
• Geographical distance

12
What are activities among neighbors and what is
the quality of neighbor relationships?

• Neighbors who are directly delineated do only
  rarely have any additional function
• Neighbors are in particular important for odd
  jobs, one does visiting neighbors and they are
  also sometimes member in ones core discussion
  network
• Segregation between working and dwelling, private
  and public one rarely discusses work matters
  with neighbors
• Neighbor relations belong to the weakest
  relationships in ones network

13
Activities among neighbors
Network-neighbors Direct neighbors All other relationships
Job 2.4 0.2 2.7
Advice (giving) 1.5 0.3 13.4
Advice (asking) 2.4 0.2 16.6
Work matters 0.3 0.1 3.6
Cooperation 1.5 0.4 13.5
Taking care of house 7.4 1.4 2.5
Odd jobs 37.9 16.9 12.5
Key 32.2 19.3 10.5
Visiting 37.9 12.2 34.7
Core discussion 13.3 2.9 22.6
Source SSND, 2000, example of all network
neighbor relationships, 2.4 are mentioned being
important for getting a job
14
Strength of neighbor relationships in the
Netherlands
Network neighbors Direct neighbors All other relations
Strength (1-5) 3.9 (.85 3.4 (1.0) 4.3 (.94)
Trust (1-5) 4.1 (.96) 3.7 (1.1) 4.4 (.94)
15
Strength of neighbor relationships in the
Netherlands
Network neighbors Direct neighbors All other relations
Strength (1-5) 3.9 (.85 3.4 (1.0) 4.3 (.94)
Trust (1-5) 4.1 (.96) 3.7 (1.1) 4.4 (.94)
16
Strength of neighbor relationships in the
Netherlands
Network neighbors Direct neighbors All other relations
Strength (1-5) 3.9 (.85 3.4 (1.0) 4.3 (.94)
Trust (1-5) 4.1 (.96) 3.7 (1.1) 4.4 (.94)
70 of the direct neighbors which is asked for
directly do not have any other relational
function besides being just direct neighbors!
17
Local communities in the Netherlands
18
When does a community exist?

• If people realize a number of important goals
  within the same group of other
• This does not imply that one needs many
  relationships for experiencing a community
• But is does imply that a community offers
  something for the individual and not the other
  way around
• Hence community joint production of wellbeing
• Note this is very efficient!

19
Conditions for the creation of local communities

• Chance to meet (e.g. much facilities)
• Mating motivation social capital
• Interdependency
• Few alternatives

20
Meeting opportunities opportunities of joint
production

• No mating without meeting (Lois Verbrugge)
• Depending on
• Time spent in the neighborhood
• Degree to which one is bounded to the
  neighborhood (e.g. because of having young
  children)
• Places and facilities enforce meeting (places,
  parks with benches, shops etc.)
• Synchronic rhythms of life (e.g. when do you and
  your neighbor leave your house?)
• Residential stability probably on both, micro
  and macro level

21
Mating motivation- Social capital
motivation of joint production (i)

• Depending on
• Shadow of the future (e.g. the intention to stay
  in the neighborhood)
• Shadow of the past (investments in specific
  others in the neighborhood)
• Resources (e.g. education, social status of ego
  and alter)
• Similarity concerning relevant characteristics,
  e.g. social and marital status, family situation

22
Alternatives motivation of joint production (ii)

• A neighborhood is not the only setting where one
  can achieve his or her goals, also at work or in
  a voluntary club important goals can be realized.
  
• Not only relational alternatives are of
  importance here but also material property can
  constitute an alternative for starting
  relationships (in the neighborhood) e.g. one can
  derive status from having a luxury car.

23
Interdependencies -ease of joint production

• Different forms of dependencies
• Structural network embeddedness
• Cognitive common frame of reference, e.g.
  belonging to the same culture, religion or
  neighborhood
• Functional dependency on others for achieving a
  goal, e.g. writing a petition, making an
  arrangement on parking cars etc.
• Note dependency is highest if these different
  forms coincide in the same relationships

24
Measurement of Wellbeing/Community

• Combination of items in a Cobb Douglas function
• Community stimulation2comfort2status2affectio
  n2
• Cobb Douglas function (a production function in
  economics) allows to model diminishing returns of
  scale. If all exponents are equal to 1, there are
  constant returns to scale. If they are smaller
  than 1, returns are diminishing. It also allows
  to model substitution effects, i.e. one does not
  have status but lots of affection and therefore
  experiences community.

25
Local communities (1) goal achievement
Goal Item example Complete agreement Partial agreement Disagreement
Stimulation In this neighborhood are a lot of things going on 19.9 19.2 60.9
Comfort I feel safe here 93.1 6.1 0.8
Affection Contacts in this neighborhood are generally good 88.0 6.7 5.2
Status I enjoy respect in this neighborhood 63.9 28.7 7.4
26
Local community (2) Combination of goals
Comfort Affection Status Stimulation
1 goal X 75.9
X 14.8
X 5.6
X 3.7
2 goals X X 73.9
X X 12.4
X X 7.2
X X 5.2
X X 1.3
3 goals X X X 87.9
X X X 11.1
X X X 1.0
27
Local communities (3) Multilevel Analysis
Community
Meeting opportunities
Residential stability
Facilities in the neighborhood
Children in household
Interdependencies
Collective action/common activities
Contact among direct neighbors
R knows who resides in neighborhood
Mating motivation (social capital)
Education
Homogeneity income
Intention to leave
Alternatives
Network members outside neighborhood
Explained variance neighborhood level
Individual level
Note in this analysis it is controlled for sex,
age cohort, being married, length of residence,
urbanism and number of foreigners in the
neighborhood
28
Local communities (3) Multilevel Analysis
Community
Meeting opportunities
Residential stability .079 (.031)
Facilities in the neighborhood .011 (.004)
Children in household .126 (.061)
Interdependencies
Collective action/common activities .050 (.011)
Contact among direct neighbors .127 (.050)
R knows who resides in neighborhood .194 (.036)
Mating motivation (social capital)
Education -.037 (.012)
Homogeneity income .075 (.028)
Intention to leave -.152 (.032)
Alternatives
Network members outside neighborhood -.003 (.001)
Explained variance neighborhood level 55
Individual level 26
Note in this analysis it is controlled for sex,
age cohort, being married, length of residence,
urbanism and number of foreigners in the
neighborhood
29
Conclusion, so far

• Community in neighborhoods depends on a number of
  conditions
• In particular, interdependencies are very
  important
• Facilities, meeting opportunities do matter also,
  facilities have not only an economic function but
  also a social one! Yet, there is a differential
  effect of meeting places, not all work in the
  same direction
• Residential stability is as always - of
  importance, yet in the Netherlands this effect
  seems ot be smaller than, e.g. in the US
• Effect of relational alternatives is only weak

30
Furthermore

• Relations with neighbors are weak, they belong to
  the weakest relationships individuals have
• It is not necessary to have many neighborhood
  relationships for developing a sense of community
• E.g., higher educated people have more relations
  with neighbors yet experience less community in
  their neighborhood
• Effects of urbanization and migrants vanishes, if
  controlled for (a.o.) education

31

• Community failure?
• Social and physical disorder in neighborhoods

32
Background

• Studies and arguments by Sampson, e.g. Sampson et
  al. 1997, Sampson and Raudenbush, 1999
• Popular idea (in criminology) social order is a
  consequence of physical order e.g. broken window
  hypothesis (Wilson Kling, 1982)
• Sampson c.s. the correlation between physical
  and social order is spurious, it is influenced by
  another common condition, i.e. social control.
  Networks and community are a condition for
  informal control

33

• Social and physical order in the neighborhoods
  are collective goods.
• Who sanctions those who do not contribute to the
  production of collective goods?
• This is known as a second order collective good
  problem (Coleman, 1990, 266 en passim).

34
Model of assumptions on neighborhoods and social
and physical order
35
Analyses

• Description of all elements of the model
• Association between the different elements
• Multivariate multilevel regression model
  (controlling for age, sex, education, family
  situation)

36
Informal social controlDo you expect that
people in your neighborhood do something, if they
observe the following in the neighborhood
Yes, sure Probably yes Probably not Definitely not
Children hanging around and skipping school 20.6 27.6 35.2 16.7
Adolescents spraying graffiti on the walls 4.2 6.9 31.6 57.3
Tough arguing 11.6 32.3 34.3 21.8
Burglary 1.6 4.7 26.4 67.4
Vandalizing cars of neighborhood inhabitant 1.4 3.2 22.9 72.5
37
Physical disorder(asked to the interviewers)
What do you thing about the neighborhood of the
respondent?
Much/quite much disorder 9.0
In between 37.6
Tidy neighborhood 36.9
Absolutely tidy neighborhood 16.4
38
Social disorderIs there sometimes vandalism in
the neighborhood?
Yes 26.8
No, at least not often 13.1
No, not that I noticed at any time 60.1
39
.133 (.011)
.200 (.034)
.225 (.102)
.083 (.029)
.399 (.072)

• Note
• Low income neighborhoods create more community
• No strong correlation between social and physical
  order .30
• Ecometric analyses are partially done, no
  difference in conclusion

40
Most important conditions/associations (yes
significant association, while controlling for
all other conditions)
Neighbors in network Community Control Social order Physical Order
Meeting opportunities Yes yes yes, different effects green vs pubs! yes Yes, much green space
Residential stability yes yes yes yes yes
N of neighbors in network --- no no no no
Community --- --- yes no no
Control --- --- --- yes yes
41
Conclusion, so far

• Actual relations are not a necessary condition
  for community, as long as ties are not negative,
  weak ties are sufficient
• Conditions in neighborhoods and
  facilities/meeting opportunities are interesting
  conditions for relations, community, control as
  well as social/physical order

42
Furthermore

• The effect of pubs, snack bars etc. on social and
  physical disorder is much greater than the effect
  of migrants!
• Community has no direct effect on disorder, the
  effect is indirect via social control and
  interventions of residents
• Why? Possibly, because members of a community do
  not necessarily conform to general norms,
  community norms can be different from general
  social norms

42
43
Therefore

• Policy measures should not focus on
• Creation of cohesion and strong ties in
  neighborhoods
• Rather, they should aim at
• Stimulating social control and intervening on
  behalf of common goods in public places

43
44
Collective good production, trust and diversity
in neighborhoods- the turtle effect
45
Three important contributions by Putnam

• Putnam on Making democracy work (1993).
• Putnam on The strange disappearance of social
  capital in America (1995) and Bowling alone
  (2000).
• Putnam (2007) on Diversity and community in the
  twenty-first century.

46
E pluribus unum

• Three contributions of E pluribus unum
• Shifting the problem from bowling alone to
  bowling apart. It matters who is bowling with
  whom!
• New data allow for testing hypotheses on bridging
  and bonding, which could not yet be tested in
  bowling alone
• Not only data on macro-phenomena but also micro
  level data in individual characteristics and
  behavior.

47
Bridging and bonding

• Different forms of social capital
• Bridging and bonding
• ties to different others ties to similar
  others
• Both might have different consequences for
  individual action.
• Bridging and bonding show a positive
  correlation, according to Putnam

48
Claims of E Pluribus Unum (1)
1) Ethnic diversity will increase substantially
in virtually all modern societies over the next
several decades. Increased immigration and
diversity are not only inevitable, but over the
long run they are also desirable. Ethnic
diversity is an important social asset.
49
Claim 2 and 3
2) In the short to medium run, however,
immigration and ethnic diversity challenge social
solidarity and inhibit social capital.
3) In the medium to long run, successful
immigrant societies create new forms of social
solidarity and dampen the negative.
50
Evidence presented is merely on claim 2 the
undesirable consequences of diversity
51
Evidence (1)
Source Putnam (2007)
52
Evidence (2)
Source Putnam (2007)
53
Evidence (3)
Source Putnam (2007)
54
Other evidence for claim 2

• Diversity leads to
• Lower confidence in local government
• Lower political efficacy
• Lower frequency of registering to vote
• Less expectations regarding collective action
• Less likelihood of giving ot charity and
  volunteering
• Less close friends and confidants
• Less happiness and lower perceived quality of
  life
• More tv watching

55
Putnams explanatory idea constrict theory

• Diversity triggers anomie and isolation - not
  ingroup/outgroup division
• There is no positive correlation between ingroup
  favoritism and outgroup rejection, but a positive
  one between bridging and bonding
• Diversity leads to hunkering down, this is the
  turtle effect

56

• Analysis provides us with a number of alternative
  explanations (eg. age, education, sex, Rs race,
  ownership, incoem, satisfaction with income) all
  of them not mentioned by Putnam, yet empirically
  considerably more important than ethnic diversity
  in a given area!

57
Analysis
Source Putnam (2007)
58
Problems with Putnams arguments and analyses

• Turtle effect not analyzed! No analysis
  presenting trust against similar others!
• Macro/micro conclusions and micro/macro data on
  which level are the hypotheses
• Herfindahl index knows problems

59
Microhypotheses in e pluribus unum are not
tested

• If Putnam would have tested the implicit
  hypotheses about bridging capital, he would have
  done the following
• a) trust of Whites to Whites, Blacks, Asians and
  Latin people,
• b) trust of Blacks in Whites, Blacks, Asians and
  Latin people,
• c) trust of Asians in Whites, Blacks, Asians
  and Latin people,
• d) trust of Latin people in Whites, Blacks,
  Asians and people.
• Instead he has chosen a general dependent
  variable, which is not testing any of these
  hypotheses.

60
Micro/macro implications of arguments

• 1. more ethnic homogeneity more trust
• This is not
• 2. people of a certain ethnicity prefer to
  trust others who are ethnically similar to
  themselves (aggregation of 2 leads to 1 but 1
  does not lead to 2)

61
Analyzing diversity

• Herfindahl index 1-?pi2
• IQV index for qualitative variation
• 1-?pi2 /(1-1/k)
• Where, kn of categories and
• pproportion of observations in category i
• Source Agresti Agresti 1977 Voas et al. 2002

62
Herfindahl Index

• Imagine four groups with the following
  distribution of social categories (in percent)

Category Red Green Yellow Blue
Group 1 25 50 10 15
Group 2 10 15 50 25
Group 3 50 25 15 10
Group 4 10 15 25 50
63
Herfindahl Index

• is the same for every distribution, yet
  socially this matters a lot!

Category Red Green Yellow Blue Herfindahl index
Group 1 25 50 10 15 .655
Group 2 10 15 50 25 .655
Group 3 50 25 15 10 .655
Group 4 10 15 25 50 .655
1-((.25.25)(.50.50)(.10.10)(.15.15))
64

• Absolute group size in the population - resp.
  whether the number is on majority and or minority
  groups - is important and not taken into account
  in the index
• AND it is important to which group an actor
  belongs, in order to determine whether a tie
  provides bridging or bonding social capital

65
How to proceed ? (1)

• Technically
• Do not use the index, but separate proportions
• Include characteristics of ego as well as alter
  (who trusts who)
• Calculate cross level interaction effects
  individuals with a certain ethnicity in an area
  with a certain percentage of another ethnicity
  etc.
• Take more characteristics than ethnicity into
  account

66
How to proceed? (2)

• Theoretically
• Bridging and bonding are relative to group size
  and individual characteristics
• who puts trust in whom and how does this depend
  on context characteristics? E.g. do minority
  group members trust minority group members in
  areas where many majority group members live?
• Take into account that identities are multiple,
  cleavages need be salient and this depends on,
  e.g. goals, tasks, interdependency

67

• Multiple identities, faultlines and earthquakes
• Lau and Murninghan (1998) argue that group
  conflict becomes much more likely if a group
  shows more faultlines, that is, the coincidence
  of certain categories.
• E.g

Member 1 Member 3 Member 3 Member 4
Group 1 White, male, salesperson, Age 50 White, Male, Sales person Age 55 Black Female, Staff member Age 30 Black Female Staff member Age 34
Group 2 White Male Manager, age 60 Black Female Manager, age 20 Black male Secretary, age 40 Asian Female Sales, age 30
Group 1 knows more faultlines, but less diversity
than group 2
68
Problem with own analyses

• Majority of respondents is Dutch (93)
• Even more has mentioned only Dutch neighbors
• Therefore similarity with regard to religion is
  used.
• However religion not yet available as
  neighborhood characteristic
• Therefore at the neighborhood level ethnicity in
  combination with income is used

69
Multilevel analysis trust in neighbors (SSND
2000 and macrolevel information)
Fixed part Empty model M1 M2 M3
Constant 3.751 (.037) 3.774 (.036) 3.790 (.033) 3.790 (.033)
NEIGHBORHOOD
Urbanization .124 (.035) .076 (.036) .067 (.035)
Stability .144 (.074) .130 (.072)
Lowest income quintile -.066 (.101)
Second lowest income quintile .054 (.101)
Moroccans in neighbourhood -.106 (.033)

Moroccans in lowest income quintile .089 (.082)
Moroccans in second lowest income quintile -.204 (.084)
Note it is controlled for respondents ethnicity
70
Multilevel analysis conted
NEIGHBORHOOD
Urbanization .026 (.034) .024 (.034) .037 (.034) .035 (.034)
Stability .038 (.073) -.036 .073) -.009 (.073) -.009 (.073)
Moroccans in lowest income quintile .050 (.080) .044 (.080) -.045 (L.085) -.036 (.086)
Moroccans in second lowest income quintile -.166.062) -.162 (.081) -.082 (.086) -.091 (.087)
EGO
Sex (female) -.071 (.032) -.062 (.032) -.040 (.033) -.040 (.033)
Age .150 (.032) .147 (.032) .118 (.032) .116 (.032)
Education -.083 (.031) -.066 (.032) -.060 (.030) -.060 (.030)
Religion (no) -.102 (.032) -.112 (.032) -.091 (.033) -.086 (.032)
Married .052 (.032) .043 (.032) .033 (.032) .022 (.032)
ALTER
Married (alter) .047 (.024) .065 (.024) .065 (.024)
Same sex .002 (.022) -.006 (.022) -.007 (.022)
Same education .087 (.023) .067 (.024) .065 (.023)
Same age (/-3 years) .023 (.021) .020(.022) .017 (.022)
same religion .137 (.054) .154 (.055) .155 (.055)
CONTACTS AMONG DIRECT NEIGHBORS .171 (031) .200 (.036)
SAME RELIGIONCONTACTS -.119 (.058)
71
Interaction between similarity in religion and
contact


72
Additional analyses does contact among neighbors
depend on similarity? NO!

• Odds for contacts among direct neighbors,
  depending on (a.o.)
• Educational similarity 1.22
• Religious similarity 1.048
• Same age 1.084
• Same sex 1.25
• Note odds are highest for Catholics (1.5)!

73
Conclusion

• Context effects on trust are weak
• Most important are contacts among direct
  neighbors. These contacts do not depend on
  similarity/diversity.
• Trust is predicted through individual
  characteristics - age and education, and through
  relational characteristics
• If context effects matter, however, segregation
  lines are more important than separate conditions

74
Thanks for your attention!

• ????...questions????
• B.VOLKER_at_UU.NL

75
How to build social capital?
Hold a neighborhood barbecue
Fix it even if you didnt break it
Say hello to strangers
Ask neighbors for help and reciprocate
Participate in political campaigns
Join a gardening club
Register to vote and vote
Join or start a babysitting cooperative
Call an old friend
Hold a neighborhood barbecue
Donate blood
Have family dinners and read to your children
Sing in a choir
Avoid gossip
Audition for community theater
Play cards with friends or neighbors
Get to know your children's teachers
Bake cookies for new neighbors or work colleagues
Mentor someone of a different ethnic or
religious group
Be real. Be humble. Acknowledge others'
self-worth
Get to know the clerks and salespeople at your
local stores
Volunteer your special skills to an organization
Stop and make sure the person on the side of the
highway is OK
Hire young people for odd jobs
http//www.bettertogether.org/150ways.htm