Neighborhood effects, neighborhood problems and policy solutions

1
Neighborhood effects, neighborhood problems and
policy solutions

• Discussants Comments on Policy Responses to
  Neighborhood effects on Education, Work, Crime,
  and Health
• 7 April, 2011
• William A.V. Clark
• University of California Los Angeles/ CHR St
  Andrews

2
Introduction

• Places are different
• Neighborhood conceptions are both positive
  (Beverly Hills, Faubourg St Honore) and negative
  wrong side of the tracks, lower east side, dog
  town, pole town stigmatization
• Place based interventions are a response to
  difference, usually to negative differences

3
Basis of intervention

• Neighborhoods are (identifiable)
• Distinctive social worlds
• Territorially bounded
• Organization based on local institutions

4
But of varying form

• Nominal Neighborhoods named, but no precise
  limits
• Absolute Neighborhoods explicitly defined
  areas
• Functional Neighborhoods activity based
• Community neighborhoods interaction based

5
Three Seminars

• The question/s across two previous seminars have
  been about whether we have a theory of
  neighborhood effects and how do we understand
  dynamic neighborhoods -what is the role of place
• Now, can we intervene with ABIs what is the
  policy impetus and how sure are we that we can
  make a difference with ABI

6
(No Transcript)
7
Overview

• Four papers on substantive issues that have place
  relevance (presuming we have neighborhood theory)
• Education, worklessness, crime and health
• All are central issues for social well being
• Implicit assumption is that we know the effect
  and we can do something about it (take each paper
  in turn).

8
Education

• Response to poor education outcomes improve
  system, target neighborhoods
• ABI clearly linked to wider regeneration of areas
  (evidence?)
• Still only partially successful because meso
  level approach but need macro too
• Need to go from redistribution to recognition
  (but how, not spelled out)
• Solution give power back to schools and
  communities but how and do they want it?

9
Worklessness

• Persistent spatial concentration of worklessness
• Best route out of poverty is work - so deal with
  worklessness in deprived neighborhoods
• Causes usual suspects (econ restr., culture of
  worklessness, social capital, stigmatization,
  poor public transport)
• To intervene need to know the types of
  worklessness (can we distinguish and how?)
• But again, policies have limited impact (p.16)
  and do not do much (p.19)-no significant
  difference in gap between most and least deprived
• Interventions at local level poorly positioned to
  deal with wider change in labor markets

10
Crime

• A lot of interest from Sampsons work on
  collective efficacy and role of trust, by
  extension neighborhood capacity for informal
  control.
• BUT Really only one study and there are questions
  about the robustness of the results (Veysey and
  Messner)
• Still not a lot on HOW the neighborhood works,
  this paper tackles that question
• Specifically they show that people look beyond
  the immediate neighborhood
• That is the neighborhood is more than the
  neighborhood and it reiterates the issue of
  neighborhood definition.

11
Crime (2)

• Ok but what should we do, what is the policy
  implication of this research?

12
Health

• The paper examines two questions is poor health
  concentrated and do socially disadvantaged
  neighborhoods experience lower quality physical
  environments.
• A discussion of smoking outcomes is used as a
  springboard to discussions of environmental
  justice
• Sorting/ migration is identified as a major
  factor in the creation of difference who leaves
  who enters becomes determinant.
• But health behavior effects are elusive no or
  low association of access and outcomes
• What does this mean for policy?

13
Health (2)

• The smoking study two processes
• Selection effects in migration such that those
  prone to smoking are more likely to end up in one
  place than another
• And, smokers are more likely to move to areas
  where other smokers are more likely to live
• Question is the location decision made by
  smokers vis a vis non smokers because they are
  smokers or because they happen to be of low socio
  economic status and it is the status that is
  causing locational clustering?

14
A digression on Income and Health
15
Income distribution
16
Observations

• Neighborhood effect or income effect?
• And how to deal with the modifiable areal unit
  problem (MAUP)

17
Health by Neighborhood Type
Individual Logistic Blk Group Poisson Tracts OLS Community OLS

Satisfied 0.253 1.565 0.261
Safe 0.541 0.043 0.213 0.277
Latino -0.515
Female -0.458 0.027 0.526 0.575
Average Age -0.016
Married 0.248 0.016 0.229 0.246
BS/BA or greater 1.429
R-squared p lt 0.05 0.499 0.964 0.954
18
Life expectancy in Glasgow
14km apart 28 year difference in life expectancy
Lenzie Life expectancy82
Calton Life expectancy54
19
What is the question-from a policy perspective

• Are the areas different because of population
  composition
• Are they different from treatment effects - one
  area has superior facilities and staff
• Are they different BECAUSE there are area effects
  noxious industry, lead paint etc
• WHAT IS THE RELATIVE IMPORTANCE OF EACH

20
The underlying question

• If the bulk of the difference is composition
  effects then
• THE QUESTION IS HOW THESE AREAS ENDED UP WITH
  SUCH DIFFERENT COMPOSITIONS
• The answer residential sorting ( see Clark and
  Morrison -Residential sorting, neighbourhood
  effects and social mobility evidence from a
  large scale survey)
• but what are the processes which sorted people
  into these two neighborhoods?

21
Concluding observations

• Sorting out the sorting process is central to
  understanding the uneven concentrations of socio
  economic groups
• Unless we can sort out the sorting process we
  wont get close to sorting out the neighborhood
  effects

22
Review and Overview

1. They (neighborhood effects) may be mostly an area
   outcome not an area affect
2. If neighborhood effects exist, they are probably
   small, may be dependent upon your definition of
   neighborhood, and difficult to detect
3. Analysis of outlier neighborhoods may be more
   useful (e.g, poor neighborhoods with good health
   OR wealthy neighborhoods with poor health)
4. A sorting focus gives us both theory and testable
   measures.