Improving Statistics for Measuring Development Outcomes The need to accommodate rapid urbanization in the national statistical plan

1
Improving Statistics for Measuring Development
OutcomesThe need to accommodate rapid
urbanization in the national statistical plan

• UN-HABITAT, Nairobi
• 26 May, 2003

2
World Population Growth Will Be Mainly Urban
3
Almost All Growth Will Take Place in Cities of
Poor Countries
4
Especially in Their Smaller Cities
5
Developing Country Rapid Urbanizaton Leads to an
Increase in Informal Settlements

• Urban growth in developing countries comes
  primarily from individuals migrating from the
  rural areas (Nairobi 90 of recent arrivals to
  the slum areas came from rural Kenya).
• In the cities of developing countries there is
  restricted access to formal serviced land by the
  urban poor (limited formal land market
  activities, and limited access to credit)
• The urban irregular informal land market meets
  the demand of the urban poor (and is apparently
  both more profitable and easier to develop)
  Smolka, Lincoln Institute, ref Latin America.
• The result has been a rapid increase in the
  informal or slum areas. The formal serviced land
  market is not responding to the demand.

6
Need to sample the slum areas

• Slum areas are home health risks equivalent to
  those in the rural areas.
• National statistical practice and planning has
  been masking the problems of the urban poor.
• Poverty is still regarded as a predominantly
  rural problem.
• Sample sizes in the urban areas have been too
  small to disaggregate into slum areas .

7
Where intra-city differentials have been possible
the differences are stark.

• In Accra it was estimated that mortality rates
  were 3 to 5 times higher in high density poor
  areas (Songsore and McGranahan 1993). However,
  the perception by national planners is that those
  living in Accra are uniformly better off.
• Differentials in malnutrition are consistently
  larger in urban than in rural areas according to
  a ten country study using the DHS data.
• There are similar examples from Bangladesh,
  Philippines, Kenya.

8
SOCIOECONOMIC DIFFERENTIALS IN CHILD STUNTING ARE
CONSISTENTLY LARGER IN URBAN THAN IN RURAL
AREASPurnima Menon, Marie T. Ruel, and Saul S.
MorrisFood Consumption and Nutrition
DivisionInternational Food Policy Research
Institute

• Our study showed that children living in urban
  areas might be up to 10 times more at risk of
  being stunted if they are from poor households
  compared to children from households of higher
  socioeconomic status. The fact that there are
  consistently such strong socioeconomic gradients
  in urban areas of developing countries implies
  that reliance on global average statistics to
  allocate resources between rural and urban areas
  could be dangerously misleading, a point
  originally made by Basta (1977). We have
  previously shown that the average urban child
  is consistently less likely to suffer from
  stunting than the average rural child (Ruel et
  al. 1998), yet in virtually every case studied in
  the present analysis, there was a distinct group
  of highly vulnerable urban children that should
  be high on the list of national priorities for
  nutrition-oriented interventions.
• This piece of research demonstrates the dire
  need for program and policy attention to
  ameliorate the nutrition situation of the
  population living in poor urban areas.

9
Infant Mortality Higher in Dhaka Slum than in
Rural Bangladesh, 1991
10
SOME URBAN FACTS OF LIFE IMPLICATIONS
FORRESEARCH AND POLICYMarie T. Ruel, Lawrence
Haddad, and James L. Garrett

• The prevalence of diarrhea among the urban low
  SES group was also greater than among the rural
  low SES group in 7 of the 11 countries studied.
  Thus, overall diarrhea prevalence rates in urban
  areas rival those found in rural areas, and poor
  urban dwellers are often worse off than the rural
  poor in that regard.

11
SOME URBAN FACTS OF LIFE IMPLICATIONS
FORRESEARCH AND POLICYMarie T. Ruel, Lawrence
Haddad, and James L. Garrett
12
Rapid urbanization requires that we begin to
routinely provide data on intra-city
differentials

• Nearly all the projected urban growth for the
  next decades will be in developing countries.
  (WUP 2001)
• The current estimated 924 million slum dwellers
  could increase to 1.5 billion by 2020. Almost all
  of the increase in the urban areas of developing
  countries.
• The current estimate of the percentage of slum
  dwellers in Sub-Saharan Africa is more that 70
• We have no evidence that this proportion is
  decreasing to the contrary the evidence seems to
  point to an increase in the number of urban poor.

13
Summary

• Slum dwellers are paying in lives for
  urbanization that is not being addressed by
  national and local authorities. (mortality,
  morbidity, malnutrition)
• We need to advocate for and ensure that the urban
  differentials inform policy.
• Statistically this means that we must expand
  urban coverage and include slum strata in the
  sample design and encourage national statistical
  offices also to do this.
• Urgent Need for Spatially-Disaggregated Urban Data