Download - Olena nizalova ppt
1
Social Assistance System Modernizationand Participation of the Poor
Hanna Vakhitova, Olena Nizalova, Denys NizalovKyiv School of Economics
GDN conferenceJune 2013, Manila
2
Motivation: Post-communist countries and poverty
1989, Communist regimes in Europe 9% of the World population 14% of the World land area
Poverty (less than $4 a day) 1989: 14 million people (out of 360) 1998: > 140 million people.
3
Motivation: Importance of welfare system
“The last public policy instrument to prevent extreme poverty” (Cerami, 2009)
Ukraine: 19% of all families; 13000 employees Low effectiveness of poverty reduction programs
in transition countries (Verme, 2008, Moldova; Van de Walle, 2004, Vietnam; Milanovic, 2000, Latvia) Amount of assistance (too high, too low) Poor take-up (best performers - 36%) Lack of capacity (Lokshin and Ravallion, 2000, Russia)
4
Our question
Impact of the welfare system re-engineering on three outcomes:
Share of applicants among the poor; Share of recipients among the poor; Share of recipients among the poor who
receive multiple types of assistance.
5
Our contribution
Focus on the poor-oriented system intervention
Look at particular activities Include large set of controls Account for non-randomness of Project
implementation
6
Literature
Academic literature: poverty spell (Okrasa, 1999a, 1999b, Poland) probability of exit from poverty (Okrasa, 1999a, 1999b,
Poland; Ravallion et al., 1995, Hungary; Van de Walle, 2004, Vietnam)
protecting from poverty (Van de Walle, 2004, Vietnam; Lokshin and Ravallion, 2000, Russian crisis)
Policy papers (WB, 2005, 2009): Poor targeting, complicated application procedure, poor
employee motivation, inadequate service quality, clients’ dissatisfaction – factors that prevent take-up.
77
2.5 mln families by the end of 2009 (19%) Total annual expenditures = 17 bln UAH (310 mln USD) -
6.3% of the State Budget MLSP, 27 regional and 754 local departments, 13,000
employees 15 national and a number of local social assistance
programs 6 major programs cover 97% of recipients
Brief description of the system
88
Modernization of the System
System re-engineering started in 2005:program based ►process based Functional divisions (separation of the application admission, case
processing, and money transfer procedures) Single Window/ ”one-stop shop” Other administrative changes (appointments by phone, control over
transfer of cases , archive) Renovation of offices
Modernization and improvement of the analytical capacity of the local and regional departments Computerization Employee training Unified software Unified data-base
Information campaign
9Project Impact on Household/ Population Behavior 9
Increase in productivity of employees
Population awareness about various types of
assistance
Population attitude towards system of social assistance
Number of processed applications
Activity: Information campaign
Poverty reduction
Number of applications
Activity:New computer
equipment
Activity: Facilities
renovation
Activity: Single application
Social assistance system targeting
Application timeOpportunity to submit single application for
several types of assistance
Psychological discomfort during the application
process
Readiness of applicants for application process
Number of mistakes in the assistance
assignment
Program Impact Model
10
Project activities implementation
Treatment indicatorsYear
Difference2008 2009
Renovation of premises and office equipment A
0.0 0.287 0.287***
Computer hardware was purchased during the last year B
0.111 0.006 -0.095***
Single application procedure B 0.761 0.851 0.09***
Informational campaign A 0.0 0.544 0.544***
Data: A – Administrative data; B – Employees Survey.
11
Poverty-related outcomes
Project outcomes Year Among
non-poorAmong
poorOverall in the
population
Applied for social assistance
2008 19.4% 30.6% 23.6%
2009 18.0% 33.3% 23.4%
Both years 18.7% 31.9% 23.5%
Receive any social assistance
2008 18.6% 29.3% 22.5%
2009 17.0% 28.8% 21.1%
Both years 17.8% 29.1% 21.8%
Receive multipleassistance
2008 6.0% 9.8% 7.4%
2009 4.6% 9.9% 6.5%
Both years 5.3% 9.8% 7.0%
12
Methodology
Y – Project outcome (application or participation);Z – Project activities vector; P – Poverty indicatorХ – Household controls; V – selection factors (district and office characteristics )Т – timei – individual j – district
District-level clustering
jitjjtittititjtjtjit DVXTPPZZY 76543210
13
Surveys: (Wave 1 - Fall 2008; Wave 2 - Fall 2009) Households Employees
Administrative data of MOL Program implementation log (2005-2009) Census of Social Assistance Departments (2005-2009)
General Statistics by State Statistics Committee
Data
14
Results
15
Basic specifications
Applied for social
assistance
Receive any social
assistance
Receive multiple
assistance
Renovation of premises and office equipment 0.046 0.058 -0.008
Modernization of computer networks -0.201*** -0.189*** -0.056*
Single application procedure 0.016 0.035 -0.01
Informational campaign 0.008 0.011 0.031
Poor * Renovation -0.026 -0.002 0.005
Poor * Computer 0.391*** 0.411*** 0.197**Poor * Single application 0.057 0.037 0.013
Poor * Informational campaign 0.044 0.033 -0.004
Poor 0.053 0.048 0.027
Pseudo R2 0.032 0.03 0.023
Observations 2,142 2,142 2,142
Probit marginal effect reported. *** p<0.01, ** p<0.05, * p<0.1.
16
Adding controls: household
Applied for social
assistance
Receive any social
assistance
Receive multiple
assistance
Renovation of premises and office equipment 0.065 0.076 0.002
Modernization of computer networks -0.204*** -0.189*** -0.036*
Single application procedure 0.016 0.035 -0.008
Informational campaign -0.007 -0.003 0.027
Poor * Renovation -0.045 -0.019 -0.013
Poor * Computer 0.507*** 0.529*** 0.249***Poor * Single application 0.062 0.042 0.005
Poor * Informational campaign 0.058 0.038 -0.013
Poor -0.023 -0.026 -0.001
Pseudo R2 0.102 0.102 0.205
Observations 0.102 0.102 0.205
Probit marginal effect reported. *** p<0.01, ** p<0.05, * p<0.1.
17
Adding controls: districts
Applied for social
assistance
Receive any social
assistance
Receive multiple
assistance
Renovation of premises and office equipment 0.250** 0.278*** 0.012
Modernization of computer networks -0.198* -0.193** -0.082
Single application procedure 0.09 0.119** 0.07
Informational campaign 0.089 0.078 0.014
Poor * Renovation 0.006 0.024 -0.017
Poor * Computer 0.353*** 0.378*** 0.309**Poor * Single application 0.114** 0.118** 0.039
Poor * Informational campaign 0.022 0.01 0.009
Poor 0.054 0.026 0.03
Pseudo R2 0.118 0.119 0.126Observations 2,013 2,008 1,385
Probit marginal effect reported. *** p<0.01, ** p<0.05, * p<0.1.
18
Adding controls: household + district
Applied for social
assistance
Receive any social
assistance
Receive multiple
assistance
Renovation of premises and office equipment 0.248** 0.283*** 0.02
Modernization of computer networks -0.190* -0.181** -0.035
Single application procedure 0.084 0.117** 0.034
Informational campaign 0.054 0.045 0
Poor * Renovation -0.004 0.019 -0.023
Poor * Computer 0.501*** 0.537*** 0.279***Poor * Single application 0.122** 0.123** 0.009
Poor * Informational campaign 0.022 0.003 -0.013
Poor -0.051 -0.071* 0.007
Pseudo R2 0.205 0.213 0.359Observations 1,998 1,993 1,377
Probit marginal effect reported. *** p<0.01, ** p<0.05, * p<0.1.
19
Final specification (with household, district and office controls)
Applied for social
assistance
Receive any social
assistance
Receive multiple
assistance
Renovation of premises and office equipment 0.262** 0.314** -0.028
Modernization of computer networks -0.208 -0.193* -0.01
Single application procedure 0.104* 0.146** 0.041*
Informational campaign 0.05 0.055 -0.031
Poor * Renovation -0.045 -0.022 -0.023*
Poor * Computer 0.430*** 0.470*** 0.205**
Poor * Single application 0.151** 0.147*** 0.014
Poor * Informational campaign -0.03 -0.047 -0.020*
Poor -0.043 -0.064 0.013
Pseudo R2 0.2 0.21 0.357Observations 1,665 1,660 1,099
Probit marginal effect reported. *** p<0.01, ** p<0.05, * p<0.1.
20
Social Assistance System re-engineering does improve targeting of the system towards poor and simplifies the application process:
Renovation of premises and purchase of new office equipment increases both the application rate and the participation rate of the population in the system.
Conclusions
21
Single Window Application Procedure increases the participation rate of both non-poor and poor, with the latter effect being twice larger.
Single Window Application Procedure also encourages application among the poor, without increasing the system load among non-target group.
Conclusions
22
Purchase of Computers and Modernization of Computer Networks discourages the application and participation in the system among the non-poor, while having the opposite, and quite large in magnitude, effect on the poor.
Informational campaign has no significant impact on any of the studied outcomes.
Conclusions