impact of cash transfers on poverty and well-being of … · v. child welfare and social transfers...

Study: Welfare impact of transfer programs – 16/11/2011

1

UNICEF Moldova

Impact of cash transfers on poverty and well-being of the

most vulnerable families in the Republic of Moldova,

within the context of transition from category-based to

means-tested social assistance

Final Report

November 16, 2011

By Thomas Otter and Natalia Vladicescu1

DISCLAIMER:

The views expressed by the authors in this publication do not necessarily reflect the policies or views

of UNICEF

1 Thomas Otter is the lead author and team leader of the assignment on behalf of ECI; Natalia Vladicescu, with

the support of Gloria Jigau on behalf of CBS/AXA, implemented the qualitative research component; and Ivan

Velasquez, on behalf of ECI, supported part of the statistical analysis. Editing by Richard Elsam for the English

and Gloria Jigau for the Romanian text.


2

Content:

Executive Summary

I. Introduction

II. Background

III. Methodology

IV. Changes in coverage of social transfers between 2009 and 2010

V. Child welfare and social transfers – real life stories

a. Parent’s principal preoccupations in daily life

b. Children’s needs and parent’s capacities to respond to these

c. Income sources and coping strategies

d. Use given to social transfers in households with children and their importance

VI. The transition to Ajutor Social – people’s perception

VII. Poverty impacts of Ajutor Social

VIII. Conclusions and Policy Options

List of Tables

Table 1: Implemented focus groups

Table 2: Implemented in-depth interviews

Table 3: Mean per capita transfers of Ajutor Social per number of children for each household

Table 4: Cost of raising a child throughout the different stages of childhood

List of Figures

Figure 1: All social transfers to all households with children

Figure 2: Social transfers to households with children – Quintile 1

Figure 3: Targeting of Ajutor Social for households with children of Q1 and Extreme Poverty (2010)

Figure 4: Expected poverty decrease after an increase of Ajutor Social by an additional 16,000

households (percentage points)


3

Acronyms

ECA Europe and Central Asia

EU European Union

F Female

GoM Government of Moldova

HBS Household Budget Survey

HH Household

LPA Local Public Administration

M Male

MDG Millennium Development Goal

MDL Moldovan Lei

MDPI Multi Dimensional Poverty Indicator

MLSPF Ministry of Labour, Social Protection and Family

NBS National Bureau of Statistics

OECD Organization for Economic Co-operation and Development

OPM Oxford Policy Management

Q1 The lowest quintile (20%) of the income distribution

UNDP United Nations Development Programme

UNICEF United Nations Children‘s Fund

WB World Bank


4

Acknowledgements

The authors of this study are very grateful for the contributions made by the Ministry of Labour

Social Protection and Family, the National Bureau of Statistics, the World Bank and OPM/Every

Child to this report. The team also benefited from discussions with the UNICEF regional office in

Geneva.

The team appreciates the generous time and insights offered by interviewees and focus group

participants, all of whom have provided highly relevant information, which enriched the team's

understanding of the reform of the social transfer system in Moldova.

All opinions expressed in this report, and any omissions and mistakes, remain the sole responsibility

of the authors and should not be attributed to other parties.


5

Executive Summary

This study was undertaken at the request of UNICEF and in close consultation with the

Ministry of Labour and Social Protection and the World Bank, with the purpose of identifying

policy options to assist in the further development of the on-going social protection reform

process in the Republic of Moldova. The study explores the effectiveness of the new, targeted

Ajutor Social2 programme in reaching the most disadvantaged children and families. It aims to

identify opportunities to further enhance achievements, at a strategic and critical stage of the reform

process, where adjustments can still be easily included in the programme. In order achieve this, the

research examines the experience to date, and analyses targeted transfers. Based on this evidence, a

number of policy options are proposed which can help ensure further reforms to help reach the policy

objectives of ensuring protection of those most in need, including children.

The increase in coverage of Ajutor Social in 2010 was of considerable scale and the achieved

improvement in targeting the payments to vulnerable households has been remarkable. The

expansion of Ajutor Social was supposed to compensate for a contraction of nominal compensations,

but in fact, Ajutor Social expended stronger than the compensations decreased. This success,

however, is partially offset by a decrease in the payment of allowances resulting in the overall change

coverage being negative for any kind of social transfers to households with children in 2010. Hence,

we have a positive trend for households with children who are receiving Ajutor Social, while at the

same time a negative trend for households with children who are not receiving Ajutor Social, in what

refers to coverage with other social transfers. Additionally, in 2010, more (adult) people lived in

households with children than in 2009, which reduces the per capita benefit. More importantly, in

2010, some 45% of households with children from the poorest quintile (Q1) did not receive any social

transfer at all. In spite of the considerable increase of Ajutor Social, the overall balance of coverage

with any kind of social transfers between 2009 and 2010 for households with children is negative,

both for all households with children and for households with children from the poorest quintile.

In spite of the impressive achievements in improving the targeting of Ajutor Social between

2009 and 2010, ensuring that this programme covers the poorest of the poor households with

children remains a challenge to be addressed in the coming years. This is not specifically a

Moldovan problem, but it frequently occurs with means-tested transfer programmes, since this

method has technical limitations. After the important improvement of the targeting mechanism in

2010, its technical limits seem to have been already reached; only small gaps for further technical

improvement are left. Poverty impacts of Ajutor Social are already visible but could even be

improved, without requiring considerable additional financial resources, targeting the poorest

households again on nominative definitions, since the means test fails to identify them properly. By

doing so, many of the perceived inequities identified in this report could be rectified; the programme

would induce more distributive justice and, more importantly, reach its objective of benefiting the

poorest.

Additional risks for child welfare, in spite of the success of Ajutor Social, might arise during the

transition period from a scheme of nominal compensations to the new programme unless these

are carefully planned. For example, children from some poor households may be excluded due to

2 Social aid


6

unavoidable exclusion and inclusion errors, which occur in any household targeting mechanism.3 In

particular, much experience was gained during the greatly increased coverage process of Ajutor

Social between 2009 – when a total population of 46,184 persons was covered – and 2010, when

coverage reached 112,927 persons4; this offers important lessons learned on the challenges of the

transition period regarding the impact on child welfare.

Social transfers to households with children, of any nature and amount, are usually spent for

the benefit of children, even if the amount is small. Hence, even if intra-household expenditures

cannot be controlled, transfers to households with children in most cases seem to achieve their

objective of being spent in benefit of children. According to the findings of our study, food, shelter,

clothes, education, health and living environment, in this order of priorities, are the main concerns of

parents regarding their children. Available financial resources in households with children are

distributed to these needs in the same ranking, first giving priority to cover basic needs at home and

then giving priority to access social services (education and health), which support development

opportunities for children. Social transfers of any nature, if available, frequently represent the only

permanent source of income. Relying on them is one of the principal coping strategies, together with

relying on networks of social capital (family members, friends, neighbours), labour migration, child

labour, taking loans and selling assets. Out-of-pocket expenditure for education and health are a

burden, which frequently cannot be afforded by poor households with children. The poorer the family,

the more helpful the social transfers are, especially when their reception is guaranteed (secure),

monthly (in order to smooth cash flow), and with a low opportunity cost for getting access to the

benefit. Many households who benefit from Ajutor Social do not seem to have a coping strategy at all,

in case the payment would be discontinued. Some focus group participants were sure to find a way

back to their living conditions prior to Ajutor Social, but other participants showed concern for being

able to handle such a situation.

In spite of the successful expansion of Ajutor Social, there are substantial opportunities to

improve further - many of which have already been independently identified and acted upon by

the Ministry prior to finalization of this study. People seem to be better informed regarding how to

gain access to Ajutor Social, but not regarding eligibility, reapplication or the estimation of the

transfer amounts. As it stands today, the system of Ajutor Social creates unavoidable inequalities,

which are clearly perceived as ―unfair‖ by the population. In fact, even access conditions are not

equal, since opportunity costs for the bureaucratic application process are relatively higher for worse-

off households, to the extreme point of being unaffordable for the most vulnerable ones. Experiences

in other countries show that the implementation of Ajutor Social on the ground could be enhanced

through a deeper involvement of LPAs. These, however, are not currently in conditions to offer this

wider support, given their own resource limitations. While households get very helpful technical

support to access Ajutor Social (through social assistants), it would be important to consider what

support they require to design an exit strategy in case payments are discontinued.

3 The targeting system for the selection of beneficiaries is based on econometric models, which predict available

household income based on what is known in econometric literature as ―proxy means‖. By nature, proxy means

do not perform perfectly and commit inevitable but to certain degree reducible errors of including households

which, considering their real (not their estimated) income, should not receive the transfers (because their income

is too high to qualify as a low-income household); likewise, households which according to their real (not their

estimated) income should benefit from transfers (because they are low-income households) are not being

selected by the proxy mean test. 4 Data according to NBS HBS 2009 and HBS 2010


7

At present, Ajutor Social only reaches approximately half of the overall eligible households, but it can

already be foreseen that even reaching a complete coverage of some 70,000 households (according to

current selection criteria) will not necessarily ensure that extremely poor households, or a high share

of households with four and more children, will be covered.

These lessons learned clearly show that even if Ajutor Social manages to make social transfers to a

certain degree more cost- and more result-efficient, efficiency gains remain limited to beneficiary

households. Meanwhile, poor non-beneficiary households are at risk of falling back into deeper

poverty. It is important to recognize that Ajutor Social is a good and a right thing whose benefits

could be improved and complemented through complementary activities in social policies.

The study shows also that any transfer or benefits, particularly those predictable like nominal

compensations or 50 mdl allowance, regardless their amount, count for the poorest quintiles. The

withdrawal of these benefits without coverage of Ajutor Social or any other social payments would

increase the poverty gap and severity.

This study shows that specific nominal targeting mechanisms are still the most powerful mechanism

for reaching the poorest population groups and proposes it as an option for the GoM to consider for

the poorest households with children. The study shows that certain levels of assets in households and

access to social services exist, but that monetary resources are required to in order to make these

assets and services work as a source of welfare for children. At the same time, at least one quarter of

households with children lack a minimum level of access to services and assets and a minimum level

of available financial resources. These households could be targeted for monetary transfers via a

nominative scheme, and would need additional support to overcome their simultaneous structural,

social and monetary exclusion.

Principal Policy Options

Continue the expansion of Ajutor Social and target the poorest households (e.g. rural

households with four or more children) through allowances, since they remain widely

excluded from means tested targeting and are disadvantaged when it comes to the definition

of transfer amounts by the current mechanism.

Hand over these transfers preferably to mothers or female carers (where possible), according

to worldwide evidence showing that women use additionally available financial resources for

children more than men. Additionally, the management of such kinds of funds has proved to

be an important mechanism for women‘s empowerment.

For defining the total amount of transfers per household, consider establishing ceiling per

program and per capita (for example, 100 MDL per adult and 150 MDL per child) and

maximum amounts per household (e.g. a maximum of 900 MDL, even if the sum of per

capita payments would be higher). Doing so would allow the government to set clearly

predictable expenditure amounts and beneficiaries would know for sure how much money

they could count on. In order to maintain the purchase power of such kinds of payments, they

can be adjusted to the consumer price index on a periodical basis.


8

For very poor households, complementing Ajutor Social with other social and public services

such as ensuring pre-school education, support to ensure school attendance and periodic

health checks during an extended period of childhood (for example, up to 15 years).

During the transition period from one model of social transfers to another, accompany any

withdrawal transfers with up-scaling of Ajutor Social, especially considering the expected

impact of withdrawal on poor households with children.

Implement a new information campaign regarding eligibility criteria, the determination of

transfer amounts and the unavoidability of inequities generated during the transition period.

Provide additional assistance to the most vulnerable potential beneficiary groups of Ajutor

Social who are not in conditions to accomplish bureaucratic access conditions to the transfers

- e.g. Roma families without personal identification documentation (this could be a role for

the upcoming Roma mediators planned by the MLSPF). This could happen within Ajutor

Social with special (exceptional) administrative rules, or even more efficiently with

nominative allowances as in the proposed case of rural households with four or more

children.

Reduce application opportunity costs of Ajutor Social for the poorest households (e.g.

through a onetime first payment subsidy). Current reform proposals from the MLSPF and the

WB already address the speed of application approval. In fact, this will reduce the time aspect

of opportunity costs; however, it does not solve the financial part of opportunity costs.

Knowing, for example, that some travel costs would be covered, independently from the

approval or disapproval of the application, could encourage a higher number of poor families

to initiate the application process.

Involve LPAs and Social Assistants deeper in monitoring and orientation for the use of Ajutor

Social. Social inspection has already been created. Social Inspection staff could receive

training on how to monitor Ajutor Social beneficiary households and how to best convince

and motivate them into spending additional resources in favour of children within the

households.

Develop clear procedures for intervention and additional social support in cases of improper

(e.g. drug or alcohol abuse) use of funds received via Ajutor Social. Therefore, a kind of

referral system for social emergency cases could be considered. Moldova has already

experienced successful referral systems, for example, in domestic violence and human

trafficking.


9

I. Introduction

While important achievements were made through the development and implementation of social

policies and in reaching the Millennium Development Goals (MDG) throughout the last years,

children remain among the most vulnerable and unprotected population groups in Moldova. In recent

years, good progress has been made in the health sector regarding child and maternal mortality but

still more efforts are needed for HIV/AIDS and contagious diseases (such as Tuberculosis), as well as

regarding access to safe water. In terms of education, important improvements have been made

regarding pre-school education; more efforts are required to reach the MDGs and national targets in

primary and secondary education. Infant mortality rate fell by about 6.6 percentage points between

2000-2009, while the under-5 mortality rate fell by about 8.1 percentage points. The compulsory

health insurance system introduced in 2004 covered about 78.6% of the population in 2009. The basic

package of healthcare services was reviewed and extended, with a greater emphasis being placed on

children, women, elderly people and the vulnerable groups of the population who are insured by the

state. As a result, all children and pregnant women are covered by free medical insurance.

Nevertheless, in 2009, about 23.2% of households were still outside of the insurance system, the

greater portion of them being from the rural sector (28.5%). Thus, about 46.5% of farmers, 34.4% of

agricultural sector employees and a quarter of households with children still have no medical

insurance policy.5 The share of underweight new-born children even increased from 4.3 in 2000 to 5.3

in 2008.

These findings suggest that even if the strategy of focusing social policies on vulnerable groups

generates an overall positive impact, there is still a need for more specific attention to these

vulnerable groups. Children are already vulnerable by the mere condition of being young and hence

requiring an adult to take care of them. The natural bond between children and their parents is already

severely weakened in Moldova, since about 17% of Moldovan children - more than 130,000 young

boys and girls - are currently living without one or both parents, due to labour migration; despite

significant progress in the last five years, some 7,000 children live in residential institutions.

Since 2008, Moldova has begun implementing a new, targeted system of social transfers, which

apparently is already overcoming difficulties of its early stage of implementation. Results from 2010,

to be discussed throughout this report show significant achievements, which go beyond expectations,

as well as space for improvement. Since the on-going reform of social transfers represents a major

change in the approach and in the implementation of social policies, they are expected to mark a new

long-term path for the future. If this is the case, it is of extreme importance to assess the current policy

performance regarding its impact on child welfare and propose measures of adjustment, which could

be implemented during the current reform period.

It is planned that the up-scaling of Ajutor Social should go hand-in-hand with the downsizing of other

cash-benefit programmes to households, mainly nominative compensations, such as utilities, child

benefits and heating allowances. The Nominative Compensations program includes discounts on

payments for gas, electricity, heating, and community services. This is a categorical program that

provides entitlements to certain groups of people (e.g. disabled, veterans, single pensioners,

Chernobyl victims etc.) regardless of their welfare status.6

5 All data included in this introduction based on UNDP Moldova National Human Development Report 2010/11

on social exclusion. 6 In the Republic of Moldova, the system of payments of social assistance includes 18 types expressed in:

compensations, allocations, allowances and aids regulated by various legislative-normative documents,


10

Since the targeted cash transfers are not per definition a specific instrument for child protection, it

cannot be expected to perform perfectly at the current stage, regarding its impact on families with

children. This study takes a look at the current performance (2010) of the targeted transfer regarding

child welfare, it analyses the improvement path of this performance between 2009 and 2010,

considering qualitative and quantitative information, and proposed policy options and adjustments

aiming at improving the performance of the targeted transfer system in the near future and fine tuning

its implementation mechanisms for a successful long-term change in Moldova‘s social policy.

II. Background

Childcare system in Moldova

OPM/Every Child (September 2009) provides the following overview regarding childcare policies in

Moldova since the early 1990s, mentioning that country has been reforming its childcare system for

many years, particularly after its accession to the United Nations Convention on the Rights of the

Child in 1993. Childcare has long been seen as part of many broad policy debates, included in sectors

ranging from migration and trafficking to education, health and social assistance. There have been

two significant changes in the pace and scope of childcare reform, first around 2002 and then again

around 2006.

In 2002, the government developed a national concept paper on child and family protection, and the

following year it approved the National Strategy on Child and Family Protection and its Action Plan

for 2003-2008. It was the first document to outline the fundamental strategic directions to improve the

child protection system in Moldova. It proposes legal and institutional reform, including the

elaboration of the Law on Social Assistance and secondary regulations on standards; the development

of human resources working in childcare, through the training of social assistants; and the

development of community and specialised social services for children and families in difficulty. In

2006, Moldova officially launched its childcare reform, which aimed to establish a network of

community social assistants, develop family support services and alternative family placement

services, and reorganise residential childcare institutions.

All these policies took place in an environment of declining population, due to both a low birth rate

and a high migration rate, and fluctuations in economic growth and associated fluctuations of

resources available for households and for the government (social expenditure). The proportion of

however, often insufficiently interconnected (due to a segmented treatment in the policies of social protection

and external interventions). Eligibility for payments is mainly determined by the categorical principle, and the

payments have a rewarding character. The eligibility mechanisms of examination of solicitors‘ incomes

(allowances for childcare until 1.5 - 3 years or until 16 years and material aids under the framework of the

RFPSS and the social aid) are applied only in case of three payments. The system of payments is broken down

into four large groups according their material content: a) Universal payments – (i) unique allowance at child

birth; (ii) monthly allowances for childcare until 1.5 years; b) Unique or sporadic payments – (i) unique

allowance at child birth; aid in case of death; (ii) compensations and annual allowances and (iii) humanitarian

and material aids; c) Permanent payments – (i) nominative compensations; (ii) monthly allowances for childcare

from 1.5 - 3 years until 16 years; (iii) allowances for adopted or guarded child; (iv) state monthly allowances;

(v) state nominal allowances for special merits; (vi) allowances for care; (vii) state social allowances and (viii)

social aid; d) Payments in kind – (i) material aids with money or goods (ii) humanitarian aids. The social

assistance payments are divided into two groups according their character, with and without income testing.

(MLSPF 2009 p. 77)


11

children in the population is rapidly declining while the number and proportion of elderly people in

the population is greatly increasing. Between 2009 and 2020, the number of children in Moldova is

expected to diminish from 773,000 (out of a total population of 3.57 million) to about 665,000.7

Today‘s children, being tomorrow‘s adults, will have to sustain a growing proportion of elderly. They

will require the best possible human development right from the first days of their lives. The

challenge of an on-going childcare and social policy reform is to ensure that sufficient provision

continues to be made for children, to make child welfare possible.

In 2009, allowances paid to families with children represented the main economic support from the

state through unique or periodic payments to families in the case of childbirth and care according

children age and family income level. Unique birth allowances represented 1400 MDL at birth for the

first child and 1700 MDL at birth for every subsequent child (both for insured and uninsured persons),

with some 40,000 beneficiaries in 2009. Monthly allowance for childcare until 3 years for insured

persons was 479 MDL, for some 29,000 beneficiaries, while the monthly allowance for childcare up

to 1.5 years for uninsured persons was 150 MDL, for some 40,000 beneficiaries.8 These are the two

largest categories apart from an important number of benefits for special categories (albeit small

categories regarding their number of beneficiaries), such as payments for disabled children of

households who take care of disabled children and others.

Ajutor Social

A relatively new element - and of growing importance within the context of an ongoing reform of

social policies - is the targeted cash transfer program called the Ajutor Social9 launched on pilot basis

in 2008, scaled-up in 2009, and implemented with the support of the World Bank, EU and other

donors. A recent World Bank study (World Bank, May 2011) shows that compared to previous social

transfers (nominative compensations), the new transfer modality was performing extremely well in

2010. ―Data from the (…) 2010 Household Budget Survey (HBS) show that 86.4 percent of benefits

paid under the (Ajutor Social) program went to those in the poorest quintile (20 percent) of the

population. This compares with the Nominative Compensations Program, which channelled only 27

percent of its benefits to those in the poorest quintile in 2010. Moreover, a huge share of the benefits

from the Ajutor Social Program goes to the very poor: 72.7% of all benefits paid under the Program

went to those in the poorest decile (10 percent) of the population. This puts Moldova‘s Ajutor Social

Program among the top performers in the Europe and Central Asia (ECA) region for poverty-targeted

last resort programs (…) and on par with those in OECD countries.‖10

―The coverage (of) (…) the

program is still small, currently covering 14.3 percent of households in the poorest quintile, of which,

most are among the poorest (65 percent of all beneficiaries are households in the poorest decile; 21.9

percent of those in the poorest decile are covered)‖.11

The highly positive achievements regarding overall targeting to low-income groups is expected to

have an automatically positive impact on families with children and provide support according to

children‘s needs, by the designers of Ajutor Social. There is a considerable possibility that this might

in fact occur; however, there is no automatic guarantee for it. This study takes a closer look at the

degree of positive (financial) impact of Ajutor Social on households with children and on the

7 Without taking the Transnistrian population into consideration

8 MLSPF 2010, page 118 - 119.

9 Social aid

10 World Bank, May 2011, page 5.

11 ibid


12

contribution this new programme already provides and is expected to provide in the future, to a

sufficient provision of support to children and their development.

In spite of the success of Ajutor Social, challenges have been identified during the transition period

from a scheme of nominal compensations to the new programme, as well as regarding exclusion and

inclusion errors of the household selection mechanism.12

In particular, the considerable increase in

coverage of Ajutor Social between 2009, when a total population of 46,184 persons was covered, and

2010, when the coverage reached 112,927 persons13

, offers important lessons learned on the transition

period regarding the impact on child welfare.

The basic intention of the new Ajutor Social is to redirect funds towards the most needed population

segments. In 2010, some 32,000 households14

were already receiving this type of social assistance, as

the institutional registry of beneficiaries confirms. The analysis of this study, however, is based on

information from the HBS. Since HBS are sample-based surveys, which are not able to reproduce

coverage percentages of small groups of population with 100% accuracy, the HBS of 2010 reports

26,809 households covered at the end of 2010, against 10,582 at the end of 2009. Hence, the increase

of coverage of Ajutor Social throughout 2010 was considerable.

Traditionally, categorical programs dominated the composition of Moldova‘s safety net. Moldova‘s

largest spending programs are the Nominative Compensations15

(30 percent of overall social

assistance spending in 2008, but decreasing over time to 20 percent in 2010). The WB (2011) reports

that, compared to other countries in the ECA, ―Moldova‘s social assistance programs were among the

weaker performers in terms of targeting accuracy (…) with only 10-20 percent of benefits reaching

the poorest decile of the population. (….) Coverage of social assistance has been modest, despite high

overall public spending levels.‖16

One of the most prominent objectives of the Ajutor Social programme is to achieve a higher cost-

efficiency; meaning that a better targeting of the most needed people and a more appropriate

determination of the transferred amounts of monthly payments increases welfare outcomes, while

reducing overall costs for the government. Increased cost-efficiency on the government side could

then be used for reducing budget deficit or for increasing coverage of the non-contributory social

safety net. Hence, after a transition process, Ajutor Social is expected to widely replace previously

existing nominative compensations.

In the process of phasing out the nominative compensations, the Government of Moldova took the

decision to freeze the new entries in the 11 categories and the amount of nominative compensations,

12

The targeting system for the selection of beneficiaries is based on econometric models, which predict

available household income based on what is known in econometric literature as ―proxy means‖. By nature

proxy means do not perform perfectly and commit inevitable but reducible errors of inclusion (household which

considering their real (not their estimated) income should not receive the transfers (because the income is too

high to qualify as a low income household) and households which according to their real (not their estimated)

income should benefit from transfers (because they are low income households) but are not being selected by

the proxy mean test. 13

Data according to NBS HBS 2009 and HBS 2010 14

WB 2011, page 30 15

The Nominative Compensations program includes discounts on payments for gas, electricity, heating, and

community services. This is a categorical program that gives entitlements to certain groups of people (e.g.

disabled, veterans, single pensioners, Chernobyl victims etc.) regardless of their welfare status. As a result, the

targeting accuracy of the program is weak. 16

WB 2011, page 3


13

which were at more than 60% energy and heating compensations, but had to deal with several

difficulties so far, during the transition from the old (nominative compensations) to a new system

(Ajutor Social). Increase in prices for energy, gas and heating, as occurred for example in May

201017

, eroded the value of Ajutor Social, and the Government discussed the approval of a new type

of compensations called ―social compensations‖ to help some of the vulnerable groups to survive

during the winter. Creating new types of compensations might seem counterintuitive during a

transition process, which is expected to reduce nominative payments, but it might well be socially

required.

Experiences from the transition process from nominal compensations to Ajutor Social in 2010 showed

that, in spite of its overall success in general, it has also negatively affected a small group of

households with children. There are cases, for example, of families who did not qualify for Ajutor

Social, but did not receive nominative compensations either, a situation which creates difficulties

particularly during winter or in a context of growing unemployment in the rural areas of Moldova.

These families are at risk of falling into deeper poverty, jeopardizing the development of the children.

In this context, the value of universal family and child allowances is particularly important for the

families with modest revenues. Without an effective social assistance system, many families are

having difficulties ensuring decent living conditions for their children, particularly during the winter

and are forced to use different coping strategies that can be dangerous for child development, such as

reducing the quality and quantity of food, child labour, a reduction of expenditure for health services,

amongst others.18

In close consultation with the Ministry of Labour, Social Protection and Family, UNICEF

commissioned the present assessment to analyse the impact of a partial or total withdrawal of

nominative compensations within the context of a transition to Ajutor Social and their impact on the

wellbeing of vulnerable children and families, as well as to propose policy options and adjustments

for the future. The assessment considers qualitative and quantitative information.

III. Methodology

The fieldwork for collecting data and information was carried out in two separated processes. The

quantitative analysis is based on results from the HBS 2009 and 2010 and the qualitative analysis is

based on a process of focus group meetings and in-depth interviews with Moldovan experts from the

social protection sector.

Qualitative Analysis

For this qualitative study, seven focus group meetings were implemented with the following

categories of participants: parents from families benefiting from Ajutor Social; families that

previously benefited from the monthly child allowances of 50 MDL, families with many children,

monoparental families and Roma families.

17

Prices for gas increased by 15%, electricity by 16% heating by 16%; we chose the example of 2010 energy

costs because the study takes a look at the transition between two systems of social transfers during 2010. 18

UNICEF, UNDP, UNFEM research on crisis impacts at a community level


14

Participants were recruited through the ―snow-ball‖ method,19

based on a couple of criteria (sex,

living environment, number of children), thus assuring the heterogeneity of the group. Participants

were selected from all three regions of the country (Centre, North and South). Data were collected

with the help of an interview guide (see annex). Focus group meetings lasted on average 90 to 135

minutes. Discussions were videotaped, with the consent of the participants.

Table 1: Implemented focus groups

Category of respondents Number of focus

groups

Number of

participants

Single parents 1 11 persons

Parents with many children 1 9 persons

Parents from Roma families 1 10 persons

Poor families that do not receive social aid 1 9 persons

Persons that currently benefit from social aid 1 9 persons

Persons that previously benefited from the monthly allowance

for growing children from 1.5 years and older20

2 20 persons

Period of data collection: February-March 2011

An additional 12 in-depth interviews were carried out, with experts in the social protection domain,

with social assistants working at the local level (villages, cities), at rayon directions and with

representatives of local public administrations.

Table 2: Implemented in-depth interviews21

Category of respondents Number of

respondents

Moldovan experts in the social protection domain 2

Social assistants 7

Representatives of the local public administration 3

Period of data collection: February-June 2011

Quantitative Analysis

The quantitative analysis is based on descriptive statistics, comparing results from the HBS 2009 and

the HBS 2010 regarding coverage and targeting or nominative compensations, allowances and Ajutor

Social. Special attention is paid to households with children. Attention is paid to the impacts on levels

of poverty, extreme poverty and the poverty gap caused by the on-going transition from nominative

compensations to Ajutor Social. The study also takes a look at effects related to allowances. This is

done in order to understand if any different interesting alternatives would exist on how to implement a

smooth transition between the two systems, maximizing positive effects and minimizing the still-

existing negative effects on households with children. The quantitative analysis also pays attention to

19

Snowball sampling relies on referrals from initial subjects to generate additional subjects. 20

The amount of the allowance was 50 MDL per month. Its payment was suspended once the payment of

means-tested social aid was established. The suspension of this monthly allowance was selected as an example,

in order to be able to capture the significance and importance, which the loss of an even small amount of

monthly payments can represent for child development in vulnerable families. 21

Interviewees were promised anonymity


15

how much space could still exist for technical improvement, regarding the use of different models of

household selection, beyond the already highly positive achievements, obtained after the adjustments

of the mean testing household income prediction variables and coefficients occurred in 2010.

Further below, this study will refer to simulations of poverty impacts of nominative compensations

and allowances. Since the analysis of this study is based on HBS, and HBS do not offer disaggregated

information regarding the different types of compensations and allowances a household receives, all

simulations consider available lump sum amounts of whatever allowance of compensation. However,

it is highly probable that most of the allowances recorded in HBS going to households with children

are in fact child-related allowances.

IV. Changes in coverage of social transfers between 2009 and 2010

The increase in coverage of Ajutor Social in 2010 was of considerable scale and the achieved

improvement in targeting the payments to vulnerable households is remarkable. The expansion

of Ajutor Social was supposed to compensate for a contraction of nominal compensations, and

in fact Ajutor Social expended stronger than the compensations decreased. This success,

however, is partially offset by a decrease in the payment of allowances so that the overall change

coverage by any kind of social transfers in 2010 to households with children is negative. Hence,

we have a positive trend for households with children who are receiving Ajutor Social, but at

the same time a negative trend for households with children who are not receiving Ajutor

Social, in what refers to the coverage with social transfers, since the overall coverage of

transfers other than Ajutor social decreased. Additionally, in 2010, more (adult) people lived in

households with children than in 2009, which reduces the per capita benefit. More importantly,

in 2010, some 45% of households with children from the poorest quintile (Q1) did not receive

any social transfer at all. In spite of the important increase of Ajutor Social, the overall balance

of coverage with any kind of social transfers between 2009 and 2010 for households with

children is negative, for all households with children and for households with children from the

poorest quintile.

As already mentioned in Chapter II, the coverage of Ajutor Social increased considerably during

201022

and the targeting of poorer income groups improved. Since this increased coverage took place

within the context of a transition from a non-targeted to a targeted system of social payments, the

overall coverage of compensations and allowances decreased.23

UNICEF fully recognizes that the overall objective of the creation of social welfare through the

ongoing reform goes far beyond child welfare and that targeting goals of course also include poor

households without children, especially in what refers to households with elderly people. However,

since UNCEF‘s primary objective addresses child welfare, all following quantitative results refer only

22

By roughly 16,000 households according to HBS data 23

Even if the reform project for social transfers only addresses the reduction of compensations in exchange for

Ajutor Social, the overall financial situation of the public sector also lead to a de facto decrease in the coverage

of allowances. Such a decrease necessarily has to also have a negative impact on households with children,

which might or not be compensated by other payments and can hence imply important welfare impacts. Only

considering the complete picture of changes related to all three types of payments, Ajutor Social, compensations

and allowances, permits the assessment of the overall welfare impact of occurred changes on households with

children.


16

to households with children. Most of the following statistical information is expressed in total number

of persons (including adults) living in households with children.

In 2010, four out of ten households24

in Moldova were home to children. Remember that in 2010,

Ajutor Social included some 16,000 additional households, some 14,000 of which were households

with children; hence, the coverage of households with children by Ajutor Social roughly tripled.25

Figure 1: All social transfers to all households with children

Source: compiled by the authors, based on NBS (HBS) data.

Figure 1 represents the entire population living in households with children, independently of their

welfare level. We have three important findings here: i) the coverage of the number of persons living

in households with children covered by Ajutor Social has more than doubled, ii) the decrease in

coverage of compensations is not even one third of the increase in Ajutor Social; however iii) there is

a strong decrease in the coverage of allowances and additionally an important increase of the overall

number of people living in households with children.26

In spite of the important increase of Ajutor

Social, the overall balance of coverage with any kind of social transfers between 2009 and 2010, for

households with children, is negative.27

24

38%, or 438,347 out of 1,283,751 households 25

From 8,616 households in 2009 to 22,654 households in 2010 26

An additional 65,000 persons between 2009 and 2010; one possible hypothesis here could be that, by

envisaging economic or social problems, people are moving together (since the overall number of children de

facto is decreasing). 27

The sum of additional people living in 2010 in households with children and the decrease of compensations

and allowances surpasses the increase of Ajutor Social by 155,000 persons.

43,050 105,931

326,597 310,034

356,171 296,228

1,075,095 1,154,026

0

200,000

400,000

600,000

800,000

1,000,000

1,200,000

1,400,000

1,600,000

1,800,000

2,000,000

2009 2010

Per

son

s

Year

Ajutor Social Compensations Allowances No payments


17

Figure 2: Social transfers to households with children – Poorest Quintile


Figure 2 shows the same results, but only for the lowest 20%28

of consumption levels of the

population – again only households with children. The increase of coverage with Ajutor Social is

much stronger than compared to the entire group of households with children. Population in poor

households with children covered by Ajutor Social29

increased from some 23,000 to 80,000 persons.

However, in spite of this very positive result the overall balance between the increase of Ajutor Social

against the decrease of compensations and allowances for Q1 households with children also produces

a negative general balance.30

V. Child welfare and social transfers – real life stories

Social transfers to households with children, of whatever nature and amount, even if the

amount is small, are usually allocated for the benefit of children. Hence, even if intra-household

expenditures cannot be controlled, transfers to households with children in most cases seem to

achieve their objective of being spent in benefit of children. According to the findings of our

study, food, shelter, clothes, education, health and living environment, in this order of priorities,

are the main concerns of parents regarding their children. Available financial resources in

households with children are distributed to these needs in the same ranking, first giving priority

to cover basic needs at home and then giving priority to access social services (education and

health), which support development opportunities for children. Social transfers of whatever

28

The still unofficial poverty headcount rate at the time of elaborating the statistical results for this report was

established by NBS at 21.4%. Poverty headcount in Moldova is estimated based on consumption levels. The

lowest quintile (Q1) represents the 20% of the total population with the lowest consumption (equal to

expenditure) levels. Since Q1 (20%) is close to the expected poverty rate (21.4%), we use Q1 as a proxy for the

poverty level in Moldova 2010. 29

An increase by 240% in households with children of Q1, compared to a 140% increase for all households with

children 30

Even if the total number of persons living in households with children of Q1 decreased by some 5,000 persons

between 2009 and 2010, the sum of decrease of compensations and allowances surpasses the increase of Ajutor

Social by 69,000 persons. More importantly, the number of persons in households with children of Q1 without

any coverage increased by some 3,600 persons.

23,075 80,188

116,742

103,919

141,430 88,487

226,976 230,630

0

150,000

300,000

450,000

600,000

2009 2010

Per

son

s

Year

Ajutor Social Compensations Allowances No payments


18

nature, if available, frequently represent the only permanent income source. Relying on them is

one of the principal coping strategies, together with relying on networks of social capital (family

members, friends, neighbours), labour migration, child labour, taking loans and selling assets.

Out of pocket expenditure for education and health are a burden, which frequently cannot be

afforded by poor households with children. The poorer the family, the more helpful the social

transfers are, especially when their reception is guaranteed (secure), monthly (in order to

smooth cash flow) and with a low opportunity cost for getting access to the benefit. Many

households who benefit from Ajutor Social do not seem to have a coping strategy at all, in case

the payment would be discontinued. Some focus group participants were sure to find a way

back to their living conditions prior to Ajutor Social, but other participants showed concern for

being able to handle such a situation.

In order to be able to better understand the practical importance of social transfers for poor

households with children, the team carried out a qualitative study, which focussed on the following

topics:

Parents‘ principal preoccupations in daily life

Children‘s needs and parents‘ capacities to respond to them

Income sources and coping strategies

Use given to social transfers in households with children

a. Parents’ principal preoccupations in daily life

All participants of our qualitative study revealed that financial problems are the most pressing ones,

since they cannot assure the required minimum income for covering the daily cost of decent

livelihoods for their children. The lack of employment or opportunities for income generation

(especially in the rural area) and the low level of wages compared to consumer prices are at the heart

of financial problems. Prices increasing faster than wages and income and the loss of purchase power

constitute serious difficulties for paying monthly utilities. For example, two of the focus group

participants were disconnected from electricity due to the lack of payment. Others, in order to avoid

disconnection from services or to reduce expenditure, ceased using gas heating or telephone.

“We installed autonomous heating, went abroad and earned some money, in order to improve living

conditions and now the gas price has increased and I am just looking at this installation as if it was a

museum. I do not dare to put it on.” (Female (F), 38 years old, maternity leave, non-beneficiary of Ajutor

Social)

Parents are worried about the future of their children, especially regarding how they are going to

ensure their education. This is a permanent concern and persists through all age groups of children,

from parents with children of preschool age to parents with children who are already about to

graduate from high school.

The living environment of children is at the top of the list of parents‘ preoccupations. Parents wish for

a decent life for their growing children and are concerned that the present apparently presents greater

and different challenges to the youth, compared to previous generations. Parents are worried about the

fact that there are very few options for the young people for spending their free time and that today

young boys and girls are easily exposed to alcohol, as for example in bars and discos. Parents

consider that children left behind by parents living abroad are the most vulnerable to adopting certain


19

deviant behaviours. In general, labour migration is considered a problem of the Moldovan society,

regarding family relationships and general social cohesion, due to the discrepancies regarding income

levels generated by migration and remittances. For certain, the income of migrant families is higher

than for non-migrant families. Many respondents from different focus groups mentioned that working

abroad and receiving remittances frequently represents the only feasible solution to financial

problems. The outreach of this solution goes beyond the migrant household (direct beneficiary), since

there are other indirect beneficiaries in communities, as for example other persons working for

families who receive remittances.

b. Children’s needs and parents’ capacities to respond to these

Food

The majority of the focus group participants stated that they were only able to assure food security in

their households thanks to self-consumption of what their own households (or other wider family

member households) produce through agriculture and livestock (small animals) activities. However,

their diet is consequently less diversified than it should be and is perceived as being less tasty. Parents

declare that they would like, at least occasionally, to be able to provide to their children certain

specific foods or products, requested specifically by the children.

“The day before yesterday my son said to me: “Daddy I want sausage.” I looked at him and swallowed

hard: “Adrian, we are going to receive the pay check and we will buy sausages.” I will go and get

everything that they want, at least to make them taste from fruits and oranges.” (Male (M), 39 years old,

unemployed, beneficiary of Ajutor Social)

However, there are families who occasionally suffer situations of a total lack of food. Several

respondents, mainly from the urban area (single parents and parents with many children) mentioned

having already faced situations when they did not have anything to give to their children.

“I had such a situation. I really had nothing to eat and I had small children, I was just crying and at noon

my brother showed up with flour, onions and potatoes.” (F, 38 years old, nurse, non-beneficiary of Ajutor

Social, 3 children, single mother)

“Friends gave us some food in the form of a loan from the shop. I don’t have any land and at the

apartment I can’t raise chicken. My children have already forgotten when was the last time they ate meat

or eggs. “ (F, 34 years old, nurse, non-beneficiary of Ajutor Social, 4 children)

“During three months we ate only mamaliga three times a day” (F, 46 years old, unemployed, non-

beneficiary of Ajutor Social)

“During one week, the children ate only mamaliga with scraps and after that they started vomiting; they

got sick.” (F, 32 years old, unemployed, beneficiary of Ajutor Social)

Health

Health problems in low-income families are a true challenge from a financial point of view, because

they have to make financial efforts in order to be able to consult a doctor, and if they manage to do so,

it is usually quite late, considering the deterioration process of their health conditions. The great


20

majority of focus group participants revealed that when one of their children is sick, they do

everything they can to guarantee medical treatment, by selling goods from the household or trying to

borrow some money. If the illness is not too serious, such as flu, parents explained that they are

treating these themselves and that they are only consulting a doctor for complicated cases.

Some parents mentioned being shocked by the behaviour and attitude of some doctors towards poor

families. Parents feel that sometimes doctors can do more harm than good. In rural localities, but also

in the regional centres, access to health services is perceived as unsatisfactory.

―The family doctor comes on Fridays to our village. We went there at 9am but the doctor was not there.

We stayed until 1pm, waiting. Then we called her; she said she was on her way. I stayed there for four

more hours but she did not come and after waiting a whole day, me and the children (twins 4 months)

returned back home. Finally she came on Monday and called me upon her arrival. We went there

immediately but even then we had to wait again for one hour and a half. We even had a little cold. She

gave me some syrup, which the next day provoked diarrhoea for the babies. So I bought other drugs on my

own which worked better for my children. However, such illnesses represent a big problem for us,

because they represent high expenses. Usually we were using four pampers a day per child and now, with

the diarrhoea it was 15 for each of them. There are also others in the neighbourhood who say that doctors

are doing worse. I am not even going to bother having our regular check-up anymore.” (M, 29 years old,

unemployed, non-beneficiary of Ajutor Social)

Parents are aware that a child, regardless of age, can receive emergency medical help in public

hospitals. However, they still do not trust in this new policy and fear that a financial cost still exists

and that their children would not be treated without paying. Some parents mentioned being in

situations when they were aware that their baby should be treated in a hospital, but due to lack of

money, they treated them at home.

“My little girl had chicken pox two weeks ago with 40 degrees temperature. I had no money to go and buy

certain treatments. I called the ambulance. They came, took a look and said my girl has to go to the

hospital, but I did not do so, because I had no money. I finally treated her with what I had at home, just

some drugs for the fever and that was all.” (F, 54 years old, unemployed, non-beneficiary of Ajutor

Social)

Nevertheless, several respondents recognized that there are also different kinds of doctors and

different kinds of experiences. There are also cases when doctors did everything they could for the

baby and they did not ask for money, even though parents were predisposed to offer, knowing the

general practice.

“I was with my boy in the hospital, because of an infection. He had a free surgery; they did not ask for

anything. I asked how much money I should borrow, in order to be able to pay. They explained that they

take care of it for free until 18 years of age, and they told me that it is enough just to say thank you.” (F,

34 years old, unemployed, beneficiary of Ajutor Social)

In households with disabled family members, who require permanent medication and attention, the

situation is very difficult because it places parents in a dilemma of having to choose whether to

purchase food or medicine.


21

In two of the focus groups, mothers on maternity leave suggested that it would be good if the mothers

could have free health insurance, at least for the period while they are taking care of their children,

because they cannot financially afford any treatment. Additionally, a respondent with six children

noted that it is very good that mothers with four or more children can benefit from free health

insurance, but it would be even better to also extend this service to the fathers, since they provide the

income for the family.

Education

Parents are generally satisfied with the conditions their children experience in preschool institutions.

They affirm that the same amount of money they have to pay for food in the preschool would not be

sufficient for covering the cost of cooking at home. Additionally, they perceive that preschool

institutions offer many other benefits, such as assuring children with a dietary and a sleeping routine,

as well as socializing and learning.

“For about 70 to 100 MDL, I can allow my son to be together with other children and to receive

education. Maybe I don’t have time to learn that poem with him, but in the evening I see he knows a poem

and that makes me feel happy. He eats three times a day, in the morning tea with bread and butter; even

the food is quite good, at noon a soup. I cannot offer such things to him at home.” (F, 22 years old,

maternity leave, non-beneficiary of Ajutor Social)

However, there is a group of parents who are not able to permanently pay for the kindergarten.

Especially during the winter, when there is no income from agriculture, they give up taking their

children to the kindergarten and keep them at home for a certain period of time, in order to reduce

household expenditure.

“I saw that the money was not going to be enough for the whole month, so I kept her at home.” (M, 39

years old, unemployed, beneficiary of Ajutor Social)

Some parents fear that their children could ‗lose their place‘ in the kindergarten, in case they do not

attend regularly. They mentioned that, from time to time, they are borrowing some money to pay the

kindergarten, and they never remove their children from the kindergarten for financial reasons.

Another reason for non-assistance to preschool institutions is the lack of adequate clothing and shoes

suitable for the season, especially when it rains or it is cold. Parents perceive that the access of

children from vulnerable families is restricted by the lack of places in the preschool institutions in

some localities, in the sense that children whose parents have a certain social and/or financial position

are received with preference. Also, several parents from Roma families declared that they face

discrimination problems against their children. However, many of them recognized that they are used

to taking their small children with them when they are travelling for work. Consequently, most of

their children only attend the kindergarten during winter. In many localities, the study participants

mentioned that LPA supports families with limited financial resources and offers full kindergarten

attendance, free of charge.

Access to primary education is less critical; however, with some exceptions, the great majority of

interviewed parents are concerned with assuring that their children have sufficient conditions for

studying. Exceptions are families whose children are involved in seasonal or occasional agricultural

work and some Roma families who are not sending their children to school in winter, because of


22

being afraid of them getting sick, since they do not have proper shoes, clothes, and because schools

are sometimes far away.

“My boys are still at school. Sometimes we interrupt their classes to take them to work. This can happen

once or twice a week.” (F, 45 years old, nurse, non-beneficiary of Ajutor Social)

“When it was freezing I did not allow them to go to school, because the water was entering their shoes.

Their shoes were old and I did not have money to buy some new ones and the children were staying at

home for one or two days, or sometimes even for a week.” (F, 54 years old, unemployed, non-beneficiary

of Ajutor Social)

“My boy easily falls sick in winter. Last winter, when it was very cold, I did not allow him to go to

school.” (F, 39 years old, unemployed, beneficiary of Ajutor Social)

“Out of the two children of mine who go to school in the village, I have a boy in the 7th grade and a girl in

4th

grade. Last winter it so happened that the girl did not go to school, because it was cold and I did not

have proper shoes for her, but the boy kept going. He was taking my pair of shoes and was going. ” (M,

35 years old, non-beneficiary of Ajutor Social, blacksmith)

Parents feel a financial pressure to ensure non-schooling expenditure, which is related to school

assistance. This kind of expenses, for example, seeks to provide children with clothes and food, that

according to their perception must comply with certain standards: ―to be nicer”, “to be of better

quality”, “the child shouldn’t be ashamed” amongst others. In this regard, parents have higher

expenses at the beginning of the year.

“At the beginning of September I calculate what I have worked all summer and then I purchase clothing

to last even until next spring. I try to purchase shoes of better quality so they can withstand the winter.

This way, I know that their clothes will last until spring. Summer is not a problem. I know they will wear

shorts and different T-shirts and everything seems a little bit easier.” (F, 38 years old, maternity leave,

non-beneficiary of Ajutor Social)

Other expenditure arises from paying enrolment, renovation of classroom conditions and other

financial needs for repairing or covering costs of events carried out at school. In some localities, the

parents have to pay the transportation for children that attend school in neighbouring villages.

“Food, the school, and transport to school need to be paid every month. It is a lot of money parents

are paying every month; some 60 MDL per month.” (Social assistant)

The financial contribution of parents for enrolment, and other contributions, differs importantly

between institutions, especially in the case of families with many children. For example, in some

institutions, children from vulnerable families are receiving free enrolment, and in secondary

education, enrolment is paid per family and not per child. In other institutions, families with three or

more children are not paying enrolment at all.

Many parents recognize that they have to purchase many things in order to make sure their children

can better understand the content of their studies; however, they are unsatisfied when they are

investing money in things their children will end up using only occasionally.


23

“One month I pay half of the sum and the next month I pay the full amount, representing 30 MDL for the

first month and 55 MDL for the orchestra the following month. The instruments are a problem; my

children are supposed to have instruments, but they do not, so they are using the instruments from the

music school. My children want to go there but I almost want them to stop. I am constrained to pay every

month and have to feed them, but they don’t want to give up… they have a musical ear.” (F, 46 years old,

maternity leave, beneficiary of Ajutor Social)

“I spoke with the driver, the one who takes the children to school. When I have money I pay him, when I

don’t have money I do not pay him. He knows me and waits. I need 200 MDL per month for both children

for the music school. The school sometimes waits to get paid because I cannot always pay. But when I

can, I do pay for several months in advance. Since I have such situations, when I have work, I have

money. When I don’t have any work, I have no money.” (M, 35 years old, blacksmith, non-beneficiary of

Ajutor Social)

When asked what would they do if their monthly income hypothetically increased by 500 MDL, the

majority mentioned giving priority to buying food and other things to satisfy the basic needs of

children in the household.

c. Income sources and coping strategies

Income sources

For the majority of the interviewed persons, the variety of income sources is quite limited. If one

income source fails there is hardly any alternative opportunity for income generation. When prices

increase, but wages do not and agriculture does not perform well, small livestock activities are

frequently the only alternative.

“I am a teacher and my husband works on our land; we live more from the income from agriculture than

from my wage. We are additionally raising a pig and a cow.” (F, 41 years old, teacher, non-beneficiary of

Ajutor Social)

Even more importantly, for some families the social transfers represent their only sources of income.

In 2010, social transfers (18.7% of aggregated household income) represented the second most

important income source after wages (42.6% of aggregated household income).31

“I have 4 children at school, they need clothes, shoes, but we never have enough money for buying all

these things. My wife only receives Ajutor Social of 1200 MDL.” (M, 40 years old, beneficiary of Ajutor

Social)

“Per month, I receive compensations of 110 MDL for a child and I am registered for the reception of

Ajutor Social. I receive 425 MDL per month during half a year.” (F, 25 years old, unemployed,

beneficiary of Ajutor Social)

Respondents consider seasonal work, usually in form of daily labour in agriculture or construction,

mostly in the summer, as an important additional income source. There are very limited opportunities

31

NBS, Income and expenditure of the population in 2010


24

for employment as a daily labourer during the cold period of the year. Some persons manage to get

hired as: ―selecting the pulp from walnuts”, “cutting wood”, or ―taking care of persons” (children or

elderly people).

For some people, income from agriculture covers up to nine months of the year, but in winter social

transfers represent the only sources of income.

“During the summer, I went to the market as a seller or as a daily labourer, but during the winter I

remained with what I had. Thank God that they are giving us some transfers, which we receive during

winter. We apply again each year when winter comes. It’s hard, but still, we live with the money we

receive.” (F, 46 years old, 6 children, non-beneficiary of Ajutor Social)

Periodic labour migration represents an important source of income for many households, and often

the only chance for solving problems whose solution requires a greater amount of money.

“Together with my children, we go to Moscow and we work in construction during vacations. Sometimes I

send the boys home and I stay over winter. We have some animals at home” (M, 51 years old,


Social assistants, and indeed also some of the interviewed experts, mentioned that the situation of the

families with migrants is confusing when it comes to the decision of whether to include them under

the coverage of Ajutor Social. It is difficult to properly monitor their household income. Social

assistants mentioned that they have no mechanism of control for income from abroad, or even to

detect if some family members migrated, resulting in a possible income from remittances. Only

exceptionally do families self-declare their changing condition, when a family member becomes a

migrant. Most of these cases are reported by neighbours and not by the families themselves, even if

social assistants openly explain the scope of existing penalties and sanctions.

Coping strategies

- Using second hand clothes. Some focus group participants revealed that they are giving

second hand clothes to their children that have been offered by relatives, neighbours or other people.

A large portion of participants mentioned that they are even buying second hand clothes for their

children. For some respondents, even used clothes are inaccessible. Parents buying second hand

clothes mentioned doing so also because used clothes are frequently of a better quality, when

compared to the new merchandise currently offered in the markets.

“We have a second hand shop where we go and buy clothes. If required we even can pay later. They give

us credit. From time to time we are not dressed properly, but at least we are dressed…”(F, 32 years old,


“I also buy second hand clothes; we cannot think of new clothes, it is impossible. (F, 32 years old,


- Selling animals and agricultural products in prejudice of family consumption. For example,

in some households, the family consumes milk only in small quantities, in order to gather and sell it.

Fruit and vegetables of a higher quality are sold, and only the remaining ones are consumed within the


25

household. There are also cases where high quality products are being sold, in order to use the money

for buying the exact same product, but cheaper and of lower quality and in larger quantities. This

happens, for example, when chickens raised at home are sold in order to use the money for buying

chicken legs.

“I sell the chicken for 100 MDL and I buy chicken legs for 20 MDL for preparing the meal. This way I

can prepare at least three meals, but by consuming our own chicken we can only have one meal out of it.

Only if the chicken are sick, or something like this, do we decide to eat them ourselves because we cannot

sell them.” (F, 46 years old, maternity leave, 6 children, non-beneficiary of Ajutor Social)

- Taking loans during the cold period of the year. Many respondents mentioned that they

borrow money or food from relatives or neighbours, and then work during the summer in order to

repay the money or work for free for their creditors.

“When the summer comes, I can earn some money (in construction) until the New Year. After the New

Year, I start borrowing money again; at the moment, I already have a debt of 600 MDL. During the

summer I return everything.” (M, 29 years old, unemployed, beneficiary of Ajutor Social)

- Involving children in different types of work

“I have three children and I have been a single mother for 7 years now, raising them on my own. My

mother helps me from time to time. I take my children with me and all of us work as daily labourers. I am

ashamed of myself and cry sometimes. The little boy is herding cows and makes 70-80 MDL. Then he buys

a pair of used sport shoes for the sports class at school. He is only 11 years old and the poor boy has to

make his own money for his clothes and his shoes.” (F, 34 years old, unemployed, beneficiary of Ajutor

Social)

“My children are 9 and 10 years old and work as daily labourers. If they can hold the hoe they are ready

to work.” (F, 45 years old, unemployed, non-beneficiary of Ajutor Social)

“So, during the summer as soon as they are on vacation the boys leave to work. We have a relative in

Moscow who lives there with his family and the boys go there and work and solve some of their issues.”

(M, 46 years old, retired because of illness, non-beneficiary of Ajutor Social)

Limited incomes or the lack of income usually generates conflicts within families and between

partners. In particular, women are pushing their husbands to find financial resources for supporting

the children and paying utilities.

“900 MDL is enough for half of a month after we have already paid the utilities, half of the month we are

all right this way. During the second part of the month, the fighting begins because of the lack of money

and what this means to us, for example that our children need to be dressed... ” (M, 39 years old,


“I have three children; I have 250 MDL per month that I receive for the girl. My husband manages to

work from time to time with the trolley. This is all our income. He might be able to get some freight to

Chisinau; maybe to Criuleni. But when you sit and think: a small baby, the other two children already at

school, you are feeling stressed day by day. The month passes and you have to pay all these utilities:


26

water, gas, telephone and electricity and I don’t know, I want to cry and I even have quarrels with my

husband.” (F, 38 years old, maternity leave, non-beneficiary of Ajutor Social)

d. Use given to social transfers in households with children and their importance

Without exceptions, all social transfers (Ajutor Social, nominative compensations and allowances) are

used in households with children to firstly pay for all the utilities: electricity, gas, water, and

telephone. Money left over after paying utilities is used for buying food and then other primary needs

such as clothes and shoes.

“It’s been two weeks since we received Ajutor Social (1020 MDL). When we received it, the next day we

divided it: food, milk (for small kids), coal, electricity, telephone and then, there were only 40 MDL left.”

(M, 29 years old, unemployed, beneficiary of Ajutor Social).

In families who have members with health problems, priority is frequently given to health

expenditures. Many respondents declared that they live on credit from one month to another and that

even receiving Ajutor Social would not solve this problem.

“Now I receive 1300 MDL. I’m not sure that it will be enough for even one week. I have to pay all the

debts from last month and if I need food again or something else I will borrow once more from my

brothers and they borrow from others in order to be able to give some money to me. After paying all debts

I pay for gas and electricity. That leaves me again with nothing. So I start borrowing again and so on.”

(F, 46 years old, maternity leave, 6 children, non-beneficiary of Ajutor Social).

Other needs, which imply bigger investments, such as repairs to the house or the purchase of durable

goods or animals, are usually carried out in cases when beneficiaries receive payments of Ajutor

Social for several (two or three) months in only one disbursement. In fact, these higher sums, in case

there are no debts to cover, allow them to solve certain problems related to higher expenditure.

Focus group participants perceive that allowances and compensations (except allowance from

contributory social insurance) are very small, and frequently only symbolic, such as for example the

ones for participation in armed conflicts in Afghanistan and Transnistria. People perceive these

allowances as a form of ―mockery‖, although some admit that these payments deserve some trust,

since they have really proved to be paid regularly. Within this group of low amounts of monthly

payments, nominative compensations for families with four or more children, for the payment of

utilities and the one offered to the families with small incomes for heating during the cold period of

the year (130 MDL every 2 months) are perceived as being the most useful ones. Interviewed experts

mentioned that the unique allowance for birth and the allowances for care of children up to 1.5 years

old (for families without social insurance and up to 3 years old for families with coverage of social

insurance) are also very important for supporting families with children. At the same time, mothers

with small children (focus group participants) considered that these allowances are not even enough to

cover the cost of diapers.

The security and permanency of continuously receiving even a small but relatively secure payment, is

highly appreciated by focus group participants, as shown by the example of the already discontinued


27

payment of allowances for children of 1.5 years and older.32

Many parents expressed dissatisfaction

about the fact that they no longer receive ―those 50 MDL per child‖. Although this may have been a

very small amount of money, it has been given for a long period of time and offered poor families

some minimum degree of financial stability through a secure and constant monthly payment. It

offered security, because families could count on this payment and it was received easily (almost no

opportunity cost at all).

“I think it would be best to receive those 50 MDL again, every month. Because now, we might not receive

Ajutor Social for six months in a row and then we might be accepted again but without getting monthly

payments. Sometimes they pay us two months together, then again we don’t receive anything for a while

and we never know when the next payment will be made.” (F, 32 years old, unemployed, beneficiary of

Ajutor Social aid).

“Not being sure whether I will receive the payment of Ajutor Social at the end of the month or if I will be

accepted again in the future is very bad for us. I would rather choose to receive only those 50 MDL

monthly but get them for sure.” (F, 37 years old, housewife, beneficiary of Ajutor Social aid).

“With these 50 MDL, we were certain of receiving the money monthly and could even decide in advance

on what to spend it on and how to distribute our whole income on all expenditures.” (M, 51 years old,

unemployed, non-beneficiary of Ajutor Social).

The bureaucratic process to get access to the previous allowance of 50 MDL was much easier and

faster, in comparison to Ajutor Social where, many documents and much paperwork are required.

“I have 3 children, so it was 150 MDL per month; there are families with six children who receive 300

MDL per month. Ajutor Social pays more money, but we receive this only for a certain period of time, and

then for the next period we need all the documents again. Everything must start again from the very

beginning, and for getting all the documents you need a lot of time and a lot of patience.” (F, 46 years

old, 6 children, beneficiary of Ajutor Social).

“To get Ajutor Social there are many conditions, you have to collect many documents. For those 50 MDL

it wasn’t like that.” (F, 43 years old, janitor, beneficiary of Ajutor Social).

Parents still think that the 50 MDL previously paid were far too small compared with the real cost of

children‘s needs and that its amount should rather have been somewhere between 300 to 500 MDL

per child per month.

Households who benefit from Ajutor Social have no exit strategy at all, in case the payment would be

discontinued. Some focus group participants were certain of finding a way back to their living

conditions prior to Ajutor Social while other participants showed themselves concerned with being

able to handle such a situation.

VI. The transition to Ajutor Social – people’s perception

32

This allowance was 50 MDL per month and its payment was discontinued when Ajutor Social was

established.


28

In spite of the successful expansion of Ajutor Social, there are substantial opportunities to

improve further - many of which have already been independently identified and acted upon by

the Ministry prior to finalization of this study. People seem to be better informed regarding how

to gain access to Ajutor Social, but not regarding eligibility, reapplication or the estimation of

the transfer amounts. As it stands today, the system of Ajutor Social creates unavoidable

inequalities, which are clearly perceived as “unfair” by the population. In fact, even access

conditions are not equal, since opportunity costs for the bureaucratic application process are

relatively higher for worse off households, to the point of being unaffordable for the most

vulnerable ones. Experience in other countries shows that the implementation of Ajutor Social

on the ground could be enhanced through a deeper involvement of LPAs. These, however, are

not currently in conditions for offering this wider support, given their own resource limitations.

While households get very helpful technical support to access Ajutor Social (through social

assistants), it would be important to consider what support they require to design an exit

strategy in case payments are discontinued.

This Chapter reports experiences from the ground in the transition process towards Ajutor Social.

Many of the problems to be mentioned, such as insufficient information or administrative bottlenecks

are already well-known by the government and by donors and are currently being addressed.

However, we choose to report again on problems and bottlenecks, in order to also record that

beneficiary and non-beneficiary populations are well aware of these problems and to give some hints

about the understanding and interpretation the population gives to these problems. Rather than

repeating a list of already well-known issues, this chapter tries to show how, for example,

bureaucratic obstacles convert into real access barriers on the ground.

Ajutor Social for whom?

Interviewed experts recognized that Ajutor Social offers an indisputable benefit for many families

which find themselves in difficult situations, but also highlighted that it stimulates some people to be

dependent on the state. Even disincentives for a productive life, such as giving up a modest labour

income (with wages below levels of Ajutor Social), selling agricultural land or damaging housing

conditions - all these measures occasionally taken in order to increase the chances of being approved

as beneficiary from Ajutor Social - have been reported and observed. Many people in Moldova still

expect the government to help them and solve their problems, rather than making efforts on their own

in order to improve their living conditions.

“I think that it is an issue in general of the Republic of Moldova, because this is the way the mentality

of the Moldovan people works. If there is a real opportunity to get something for free from the state,

why not do everything possible to get it, even if this means resigning from their job.” (Social

assistant)

Even if the need for social transfers for vulnerable people is widely recognized, the transition from

one mechanism of social protection to another de facto has created a parallel system of social

transfers, which produces inequities. These inequities are clearly perceived by the population, but the

justification of their existence is hard for them to understand.

“My disability started about two years ago when all of a sudden I couldn’t walk anymore. I went to the

doctor. He told me it costs this much to make a blood test and thus I then received the disability status. I


29

previously worked in the Russian Federation and also here in Moldova, on construction, for a long time.

It was a good life when I was able to do everything, but my health no longer permits this anymore.

However I’ve earned some money during my life and have some things in my house. I still have small

children. One is at school and the other is still almost a baby. My wife doesn’t work; we don’t have money

even for bread. I cannot receive Ajutor Social because I have a horse, a fridge, a TV, but how can you live

otherwise?” (M, 46 years old, pensioner because of illness, non-beneficiary of Ajutor Social)

According to the perception of focus group participants, groups of people who are often

disadvantaged in receiving social aid include:

- Landowners: Even if there is no categorical exclusion of landowners for receiving Ajutor

Social, they have to present together with their request for Ajutor Social a land tax payment

certification. Frequently taxes are unpaid since their owners have very limited monetary

income.

- Pensioners with low pensions, but with an income close to the guaranteed minimum subsistence

level or a little higher.

“Then when we approve the minimum guaranteed subsistence level every year, and this year it ended

up being a little bit higher than the minimum wage in agriculture. I think that this is a distortion of the

system” (Social Protection Expert)

- Persons with incomplete personal documentation

“There is a young couple with children I am visiting. Recently, we had a serious conversation about

their possibility of applying for Ajutor Social, but the head of the family does not have a passport and

he lost his birth certificate.” (Social assistant)

- Households consisting of several families (particularly young families living together with one of

their parents’ families), in separated households but in the same (housing) space. In these cases

Ajutor Social is calculated for the complete group of people sharing the common space. Such an

extended group of two families frequently does not qualify, since there might be more than one

monetary income. However, several families frequently live together, precisely because they do not

have enough income or resources in order to run two physically separated households.

" I received Ajutor Social for only three months, but since I mentioned that I was living together with my

wife’s parents, I no longer receive Ajutor Social. We do not have small animals or land and have to pay

for all our food. We did not share the Ajutor Social with my mother and father-in-law. He still works, and

receives about 1,000 MDL as his wage. We do not share food with them" (F, 29 years old, 2 children,


For some focus group members, beneficiaries of Ajutor Social, recognizing that they lived together

with other family members led to a reduction of the amount of Ajutor Social being paid.

"I'm a single mother and for the boy I received 368 MDL. I went to the city hall and told them that my

parents are putting the rest of the required money for him. Having learned this, they reduce the amount of

Ajutor Social, because I am staying with them" (F, 22 years old, unemployed, beneficiary of Ajutor Social)


30

- Families who are renting a furnished or equipped house or flat, thus apparently having adequate

living conditions, but in fact not having any property at all.

"You do not deserve this, but I am a lodger and almost nothing of what we have here in the flat is ours. It

is all included in the rent. But I live here with my three kids and we need all these things. And they told me

that according to what they have seen in my small place, they calculated my income and if what they

calculate would be 40 MDL less, I could get Ajutor Social, but since my income is supposedly to high I do

not get anything.” (F, 38 years old, nurse, non-beneficiary of Ajutor Social)

“I had a case where different people lived together in one house with three rooms; in one room lived one

family and in the other one another family, on the same address. I did not know what to do in such a case.

I filled in a separate form for each family, but actually I was later on told that I had to write everybody in

one form. But how am I supposed to write them all together if they are renting different rooms? They are

not even relatives.” (Social assistant)

Even if migration is not a characteristic for categorical exclusion for accessing Ajutor Social, the fact

of migration puts some households in a difficult situation. For example, in a given household the

unemployed father was the one who officially requested Ajutor Social and handed in his

unemployment certificate. The household was approved and received Ajutor Social. After six months

the reapplication was rejected because the father had migrated for work and, even if he did not send

remittances, there was no one else in this household who could present an unemployment certificate.

Information and Application

A successful expansion of the coverage of Ajutor Social is closely linked with the dissemination of its

characteristics, access conditions and others. Focus group participants mentioned that the information

usually flows in a two-stage process; initial and still vague information frequently comes from the

mass media or relatives and friends, and only then is precise information being sought at the social

assistance offices.

Focus group participants perceive that information campaigns have so far only concentrated on

disseminating the fact of the mere existence of Ajutor Social and its application procedures, but there

is no clear understanding regarding the timing required for processing applications, eligibility criteria,

limited periods of benefits, reapplication rules and the basic fact that Ajutor Social, per se, should

benefit the most vulnerable groups in first place.

The application procedure seems simple for people who received a complete explanation regarding

which documents have to be presented and where and how they should apply. Ease of application also

depends on the receptivity of public servants in granting certificates and other required documents. To

prepare a complete file for applying for Ajutor Social, it can take applicants between several days up

to several months, in exceptional cases.

The main factors, which influence the time span for preparing documents, are:

Degree of preparation of the social assistant: his/her ability to communicate and explain to

people which documents they need as well as help to fill in the forms;


31

Living environment of the respondents: families from rural localities who have to go to the

regional centres in order to obtain their documents are disadvantaged. There is a high

opportunity cost since there is little time left for procedures at the regional centre (i.e. arriving

late, and then having to return early), the financial cost of the trip itself, particularly if the

distances are great and they have to repeat the visit several times;

Performance of public administration processes for issuing documents required for applying

for Ajutor Social: queues at the labour office (where unemployment certificates - a document

required for requesting social aid - are issued), which results in many people having to visit

the regional centres several times.

“It took me four days to obtain all the required documents. They asked for family members and if we have

land property. I am Roma, and Roma people do not have land. We are simply nothing more than

unemployed people. At the labour office, this is the most difficult situation. There are many people from

villages and from districts. I had to go there three times in order to get my unemployment certificate.”

(F, 39 years old, unemployed, beneficiary of social aid)

The population does not understand the eligibility criteria and the way the amount of Ajutor Social is

established.

“I never tried to apply for Ajutor Social, because I heard in the village that they ask if you have a car, a

TV set and that those people are refused. That’s why I didn’t go there. I have some chicken, a pig, a TV

set, just as everyone else does, and I thought that I was not going to waste time going there. Maybe I

would have to kill the chicken and take away the fridge and the TV set. I thought about taking away

everything, applying and then bringing everything back afterwards, because this would be the only way

for me to be approved; but in the end, I did not apply.” (M, 35, blacksmith, non-beneficiary of Ajutor

Social)

“Only once have I received Ajutor Social. I was receiving 380 MDL for three months, and after that I

don’t know for what reason our place in Telenesti it was cut. It looks like I should have gone every month

to sign some documents, but I did not, because I did not know. Nobody told me that I should go there

monthly. I didn’t go on the required date and the payment was cut.” (F, 29 years old, unemployed, non-

beneficiary of social aid)

"They have cars, they have shops, they have enough to live on, and yet they additionally receive 1000

MDL; but me and my husband do not work anywhere. We have three children and they gave me 380 MDL

and I waited for three months, and now for three months they did not give me anything." (F, 38 years old,


“A household I am assisting was accepted and received Ajutor Social. According to my knowledge

their situation did not change at all. Maybe some criteria within Ajutor Social changed. The fact is

that the reapplication was rejected and I do not understand why, and neither the girls (from the

rayonal direction) did know why and I cannot explain it to the people.” (Social assistant)

LPAs and Social assistants

Interviewed experts consider that Ajutor Social could perform much better through a deeper

involvement of LPAs.


32

“Sometimes a new beneficiary for Ajutor Social is approved and is informed that they will receive a

certain amount of money from the beginning of next month. They are then asked on what they intend

to spend the money. Usually the vulnerable people do not know from where to take the money and on

what to spend it, so they are advised to spend on certain things. There are some elementary things

that they should do by themselves, but because they do not know how to use Ajutor Social, they should

be advised on how to best spend it. Sometimes it succeeds, other times it does not.” (Social assistant).

“We visited a family in a village (that was benefiting from 1600 MDL) where the barn was cleaner

than the children’s room. This family already had a lot of trouble regarding child protection, but

when they started to receive Ajutor Social, a strict monitoring began. Together with the mayor, the

village secretary and the social assistant, they all went together to the market to buy the necessary

things for the children and when we visited the family again a simple curtain was already there, the

walls were whitewashed, and the blankets were clean. During previous visits the place had always

been a disaster. (Social assistant)

Social assistants and LPAs, however, need adequate conditions to perform this kind of support. Social

assistants mentioned not being motivated for work, mainly because of low wages for heavy workloads

and responsibilities. Some of them mentioned being better off as beneficiaries of Ajutor Social rather

than working in their job.

“I even wanted to resign on New Year’s Eve so that I can get Ajutor Social myself. I have a retired

mother and my husband works only occasionally. Our children are in the 8th and 5th grade. The

younger one has always been sick. His disease transformed into asthma, so we moved over to the

countryside where the air is healthier. He has to go to the hospital two or three times a year, but with

my wage of only 1000 MDL, I frequently do not manage to support my family.” (Social assistant)

Interviewed social protection experts suggested the need for the following interventions for improving

access to Ajutor Social.

- Simplification of the application for Ajutor Social, through the exclusion of some indicators

considered irrelevant for correctly explaining income conditions of a household and the

consideration of other variables which more accurately predict the current welfare conditions

of a household.

- A better cooperation of the labour office, cadastral office, and other institutions with

directions of social work in order to reduce opportunity costs for the application. Some

experts consider that certain public clerks, in order to avoid additional work and to avoid

responsibility, invoke ―the secret of personal data” when colleagues from other institutions

demand certain information on beneficiaries or other kinds of support.

- The creation of a cooperation mechanism between institutions regarding the monitoring of

labour migration of beneficiary households.

- The establishing of clear interventions in the case of an improper use of received payments.

Positive experiences involving social assistants and other LPA representatives deeper in these

tasks already exist.

- The identification of some incentives for beneficiaries to invest the money received from the

social aid in some durable goods and productive goods, such as birds or animals or other

assets.


33

- With the exception of certain categories (such as disabled people, elderly persons, etc.), the

persons registered at the labour work office should periodically fulfil community service

work.

- A permanent training of the social assistants, as well as their improved motivation through

higher salaries.

- The need for training a special inspection.

VII. Poverty impacts of Ajutor Social

In spite of the impressive achievements in improving the targeting of Ajutor Social between

2009 and 2010, ensuring that this programme covers the poorest of the poor households with

children remains a challenge to be addressed in the coming years. This is not a specific

Moldovan problem, but frequently occurs with means-tested transfer programmes, since this

method has technical limitations. After the considerable improvement of the targeting

mechanism in 2010, its technical limits seem to have been already reached, with only small gaps

left for further technical improvement. Poverty impacts of Ajutor Social are already visible but

could even be improved, without involving significant additional financial resources, targeting

the poorest households again on nominative definitions, since the means test fails to identify

them properly. In doing so, many of the perceived inequities identified in this report could be

corrected; the programme would induce more distributive justice and more importantly, reach

its objective of benefiting the poorest.

The Law on Ajutor Social, Nr. 133-XVI from 13/06/2008, was defined ―with the purpose of ensuring

a minimum monthly income to vulnerable families by paying a social aid, established according to the

assessment of the monthly average global income of the family and the necessary social insurance in

order to make more efficient the system of social payments and their targeting of the poorest

persons‖.

We have already seen in Chapter II that Ajutor Social has been successful in reaching households

within the lowest 20% and of the lowest 10% of the income distribution, in a greater proportion than

similar projects in other European and Central Asian countries, and in a larger share in terms of

nominative compensations and allowances. Even if this is a very positive result, it does not tell

whether the program manages to achieve one of its main objectives - ―targeting (…) the poorest

persons‖.

On an annual basis, the National Bureau of Statistics (NBS) officially establishes a national line for

poverty and for extreme poverty. Extreme poverty is defined as the cost, per adult equivalent per

month, of a food basket that yields 2,282 calories per day. The relative poverty line also includes non-

food items of basic needs.33

Fortunately, extreme poverty rates in Moldova are below 1.5% of the

population for 2010, which means that, on average, the extreme poor people have enough food, but

lack income for covering other basic needs. Even if 1.5% seems to be a low proportion, we are still

talking about some 38,000 persons, only considering households with children, and some 55,000

persons considering all households. Recall that Ajutor Social in 2010 covered some 105,000 people

living in any beneficiary households, some 80,000 of them in Q1 (which includes extreme poverty).

Since the overall coverage of Ajutor Social doubles roughly the share of people living in extreme

33

WB, 2009, page 1.


34

poverty in households with children, there are sufficient resources available in the programme to

completely eradicate extreme poverty. However, Figure 3 shows that only 7,000 of the poorest

persons in households with children (22%) are currently covered by Ajutor Social, meanwhile the

remaining 30,000 persons are not covered. Hence, Ajutor Social fails to target the poorest of the poor.

Figure 3: Targeting of Ajutor Social for households with children of Q1 and Extreme Poverty

(2010)


Failing to target the poorest of the poor, however, is not a fault of the Ajutor Social design in

Moldova, but rather a general consequence of technical limitations for means-testing the targeting

mechanisms. The quality of a means-tested targeting mechanism is always defined by the trade-off

between statistically significant variables identified through an econometric exercise, the fact of these

variables standing a reality check of income predictors in real life, the possibility to properly monitor

these variables in beneficiary households, and the capacity to avoid cheating in the implementation of

social transfers.

The performance of income prediction (variable selection and associated coefficients) improved

considerably between the pilot project of Ajutor Social in 2008 and 2010, as prominently proved by

achievements in an improved targeting. In spite of all these improvements, there are still challenges in

reaching the poorest of the poor. This is a common phenomenon for means-tested transfer

programmes, and Moldova is not the only country facing this issue. For example, Gassmann and

Notten (2006)34

compared means-tested and universal transfer schemes, both in terms of targeting

efficiency and their effects regarding general and chronic poverty reduction in Russia. The authors

find that since the reforms of the transfer programmes, more children receive benefits and that there is

an improved targeting of benefits to children living in low-income households. Nevertheless, in 2004,

between one third and a quarter of the poor children did not receive benefits, while about 50% of the

benefits leaked into the three highest expenditure quintiles (which basically tells that the Russian

programme suffers from serious exclusion and inclusion errors). The poverty reduction impact is

modest and improved only marginally since the reforms. The simulations implemented by the authors

34

Franziska Gassmann and Geranda Notten, 2006, Size matters: Targeting efficiency and poverty reduction

effects of means-tested and universal child benefits in Russia, Maastricht University.

80,188

6,942

423,036

30,972

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

400,000

450,000

Q1 Extreme Poor

Population in households with children with Ajutor Social

Population in households with children without Ajutor Social


35

show that universal schemes achieve additional poverty reductions in all indicators because children

who were excluded by error under the means-tested scheme would receive a benefit under universal

schemes.

In order to test how close these findings are to the Moldovan context, this study implemented a series

of exercises to test how much room is left for technical improvement in the existing targeting

mechanism of Ajutor Social. Our findings suggest that the remaining technical space is small (which

means that the existing targeting mechanism is already of a good quality standard) and that even

taking advantage of these spaces would not allow the reach, in a wider context, of extreme poor

households with children. A detailed documentation of the tests can be found in the annex.35

Results

in the annex, in fact, show that the consumption level – as used by NBS - is the best performing

indicator for predicting poverty in households with children, compared to income poverty and a

Multidimensional Poverty Indicator (MDPI) (non-monetary) developed for this study. We find that

MDPI poverty is lower than monetary poverty. This fact suggests that a larger share of households

and households with children in Moldova have at their disposal a network of services and a minimum

level of assets required for a minimum level of wellbeing, compared to the share of households with a

minimum level of monetary income.

However, it is important to understand that a minimum level of monetary income is required,

precisely to make networks and assets perform in a way that provides minimum standards of living.

Welfare is not only comprised of networks of services or assets or monetary income, but also a

combination of all these elements. In this sense, special attention should be paid to the fact that at

least one quarter of households with children at the same time lacks assets, access to services and

monetary income to make perform the existing level of assets and services in a way which ensures a

decent living standard. Hence, this kind of poverty can only be overcome by providing additional

financial resources and access to assets and services for households with children.

Poverty reduction impact of social transfers

World Bank (2011) reports that the estimated impact of poverty reduction that Ajutor Social achieved

in 2010 was of ―0.8 percentage points (…) in poverty headcount rate (from 21.7% with the program,

to 22.5% in the absence of the program)‖ and that ―these impacts are in-line with expectations; as

impacts on national poverty indicators are not expected to be large due to the low coverage of the

program in 2010 (only 14.3% of the poorest quintile)‖.36

At the same time, today‘s schedule for the

on-going reform of social transfers stipulates that ―the untargeted (categorical) Nominative

Compensation benefits will be eliminated starting in 2012‖ and that this elimination ―would result in a

0.7 percentage point increase in poverty rate in 2012‖.37

At the same time, WB projections suggest

that the positive impact of a future increase of Ajutor Social (combined with heating allowances)

35

We are testing different sets of income or expenditure prediction variables, using data from HBS 2009.

Findings show that including variables of family characteristics in a wider range in the means-testing model can

still offer some room for improvement in income prediction, in general. However, all tested models perform

worse when we test their predictive power for low incomes. In practical terms, this means that very low

incomes, which are characteristic for extremely poor households, are overestimated by means-testing models,

meaning that the estimated income is considerably higher than the real income, which makes extremely poor

households ―appear‖ less poor than they really are. Consequently, they might not be selected as beneficiaries of

social transfers or the approved amount of transfers could be relatively low compared to the level of resource

scarcity the household is facing. 36

WB 2011, page 68. 37

Ibid.


36

would fully compensate for the welfare loss caused by the elimination of compensations and result in

an overall poverty reduction of 1.4% by the end of 2012.

The team‘s estimates confirm WB data of the expected poverty impact of the elimination of

nominative compensations;38

hence, WB simulations on future poverty reduction as a result of Ajutor

Social also appear reasonable. Rather than the overall poverty impact, this study is interested in the

distribution of the expected income amongst households with children. In order to provide an

additional element to the discussion, the team implemented for this study a simulation of the poverty

reduction impact of an increase in an additional 16,000 households covered by Ajutor Social, and its

distribution on households with children. The rational of this simulation is the following; between

2009 and 2010, coverage of Ajutor Social increased by roughly 16,000 households, thus Moldovan

institutions have already proved to count on resources and capacities to manage this scope of

increase.39

Figure 4 shows an expected higher impact on households of Q1 (as during the increase of

coverage of Ajutor Social between 2009 and 2010) and a higher impact on households with younger

children, but no additional impact for households with more than three children.

Figure 4: Expected poverty decrease after an increase of Ajutor Social by an additional 16,000

households (percentage points)

Source: compiled by the authors, based on NBS (HBS 2010) data.

These findings are in line with our argument that the means-tested targeting mechanism does not

always successfully target the poorest households (which usually have more children), but also leads

the attention to the discussion of the per capita amount of monthly transfers of Ajutor Social, since a

lower per capita amount of transfers will necessarily lead to a lower relative impact on poverty

reduction. Table 3 shows the mean per capita transfers per household members achieved in 2010. The

data clearly show that real per capita benefits for households with more children are lower. Behind

this phenomenon, we again have the limited capacities of a means-tested targeting mechanism to

38

Our findings suggest a poverty increase of 0.75 percentage points against 0.7 percentage points estimated by

the WB. 39

The simulation is based on HBS 2010, establishing a subsample of eligible households for an increase of

Ajutor Social, combined with probabilities that this event occurs. The distribution of the newly selected

households along the income distribution roughly reproduces changes observed between HBS 2009 and 2010.

-6.00

-5.00

-4.00

-3.00

-2.00

-1.00

0.00All HH

HH with

childrenunder 18

HH with

childrenunder 16

HH with

childrenunder 10

HH with

childrenunder 5

HH with

childrenunder 1

HH with

1 child

HH with

2children

HH with

3children

HH with

4children

HH with

5 andmore

children

All HH Q1


37

correctly predict low incomes.40

This finding also suggests that a non-means-tested targeting of

households in extreme poverty, or with more than three children, and a nominative definition of the

transfer amount (and not paying the difference between the estimated income and a minimum

subsistence level) would help considerably to better target the poorest households with children.

Table 3: Mean per capita transfers of Ajutor Social per number of children for each household

HH TYPE MDL

HH with 1 child 90.6

HH with 2 children 96.6



HH with 5 or more children 52.0


Since households with many children require even more social protection, the per capita transfer for

these households should be higher than mean values. The team implemented a simulation of targeting

households with more than three children with lump sum monthly transfers of 100 MDL per adult and

150 MDL per child. This amount of transfer would increase the poverty reduction effect of Ajutor

Social up to 25% (e.g. achieving up to 1.8% of poverty reduction in 2012 against current WB

estimates of 1.4%), at the same time representing an increase of the overall budget of Ajutor Social by

roughly 15% and at the same time limiting the maximum amount of payments per household to 800

MDL, thus avoiding the disincentive problems described above. Establishing the same limit of a

maximum transfer of 800 MDL per household for all families would totally compensate the cost

increase created by higher transfers for poorer households, thus allowing a redistribution of available

resources to the poorest families.

We are not suggesting that the amount of transfers of a nominative targeting of, for example, all rural

households with more than four children currently not covered by Ajutor Social (5,797 households

according to HBS 2010) should be fixed at the aforementioned amount of 150 MDL per child. The

amount was used as an example. One possible criterion for fixing the amount of payments is to

differentiate payments according to different costs of raising a child in different stages of his/her

childhood. Table 4 below shows an example of indexes of costs of raising a child.41

The chosen

examples refer to the United States, but there are few arguments, which could suggest that the general

trends of i) a higher cost for elder children, ii) a stronger relative increase for low-income households

and iii) a strongest increase for single parent households should not be true for Moldova as well.

40

See footnote 29 and technical annex. 41

The annual cost of the lowest age group equals an index value 1 and the subsequent age groups are expressed

as shares of this index.


38

Table 4: Cost of raising a child throughout the different stages of childhood

Dual Parent Household Single Parent Household

Age Lower third

of income

Middle

third of

income

Upper third

of income

Lower half

of income

Upper half

of income

0 to 2 1.00 1.00 1.00 1.00 1.00

3 to 5 1.02 1.03 1.02 1.13 1.08

6 to 8 1.03 1.03 1.01 1.27 1.14

9 to 11 1.04 1.02 1.00 1.18 1.10

12 to 14 1.16 1.10 0.98 1.27 1.17

15 to 17 1.15 1.12 1.09 1.41 1.21 Source: Lino 2006

Targeting achievements for extreme poor households also draws attention to a different problem: the

usefulness of the category of extreme poverty for targeting social policies. Even if the fact of having

very low and, at the beginning of the 2000s, fast decreasing rates of extreme poverty in Moldova is a

very positive development achievement, the statistical share of extremely poor households becomes

very small, which increases the difficulty to target and describe their characteristics accordingly via

statistics. Consequently, the design of adequate intervention mechanisms of social policies also

becomes difficult and inefficient. It could therefore be more efficient and effective to consider a

different categorization - for example, the lowest income decile against the lowest quintile as standard

categories for targeting social policies, rather than relative and extreme poverty according to national

poverty lines.

VIII. Conclusions and Policy Options

The new system of social transfers currently being rolled out in Moldova has been impressively

effective in targeting the poorest households, compared to similar projects in other countries. Ajutor

Social reaches the poor but still faces challenges to reach the poorest of the poor due to the fact that a

means-tested selection mechanism has its technical limitations for identifying the poorest households,

many of them households with children.

Ajutor Social currently reaches approximately half of the overall eligible households. It can already

be foreseen that even when it reaches a complete coverage of the targeted 70,000 households

(according to current selection criteria), reaching extremely poor households or a high share of

households with four and more children will remain an issue.

These lessons learned show that while Ajutor Social succeeds in making social transfers more cost-

and more result-efficient, efficiency gains remain limited to beneficiary households. Meanwhile, poor

non-beneficiary households are at risk of falling back deeper into poverty. These gaps can however be

addressed through further adjustments in social aid, as well as complementary social policies.

This study shows that specific nominal targeting mechanisms are still the most powerful mechanism

for reaching the poorest population groups and proposes these as an option for the GoM to consider

for the poorest households with children. The study shows that certain levels of assets in households

and access to social services exist, but that monetary resources are also required in order to make


39

these assets and services work as a source of welfare for children. At the same time, at least one

quarter of households with children lacks a minimum level of access to services and assets and a

minimum level of available financial resources. These households could be targeted for monetary

transfers via a nominative scheme, and would need additional support to overcome their simultaneous

structural, social and monetary exclusion. One of the interviewed Social Protection Experts for this

study found a way to explain this challenge in easy words.

„Ajutor Social fulfils its purpose when 80% of beneficiaries are families with children and they at

least have something to eat and some clothes. This would already be an important achievement, but

the following stage is that these children have to be monitored and supported starting from

kindergarten throughout their entire childhood and youth. It is a must that children from these

families get places in the kindergarten and thus future perspectives exist.” (Social Protection Expert)

By monitoring and supporting children from vulnerable households, Government can help ensure the

development of the potential capabilities in these children, which help them in the future to escape

poverty through their own efforts.

The past two years of transition to Ajutor Social have shown some lessons learned: a need for a better

understanding of eligibility criteria, of determination of transfer amounts, of continuity of payments

and of social support for exiting Ajutor Social. Payments need to be fast (approval), secure and

regular. A second information (mass media) campaign would be very helpful in addressing these

issues and explaining the nature and the unavoidability of mistakes generated through the transition

period.

Policy Options

Continue the expansion of Ajutor Social and target the poorest households (e.g. rural

households with four or more children) through allowances, since they remain widely

excluded from means-tested targeting and are disadvantaged when it comes to the definition

of transfer amounts through the current mechanism.

If these nominative allowances for the poorest households are adopted, it would be important

to take into account differentiated amounts of payment according to the age-specific cost of

raising a child. While the study did not specifically explore the issue, the differential costs of

raising a child with disability could also be taken into account, in-line with existing good

practice in Moldova.

Preferably hand over these transfers to mothers or female care-givers (where possible),

according to worldwide evidence showing that women use additionally available financial

resources for children more than men. Additionally, managing such kinds of funds has proved

to be an important mechanism of women‘s empowerment.

For defining the total amount of transfers per household, consider establishing ceiling per

program and per capita (for example 100 MDL per adult and 150 MDL per child) and

maximum amounts per household (e.g. maximum 900 MDL, even if the sum of per capita

payments would be higher). Doing so would allow government to have clearly predictable

expenditure amounts and beneficiaries would know for certain how much money they could


40

count on. In order to maintain the purchase power of such a kind of payments, they can be

adjustable according to the consumer price index on a periodical basis.

Apply means-testing only for households with less than four children and for households

without children.

For very poor households, complement Ajutor Social with other social and public services

such as ensuring pre-school education, support to ensure school attendance and periodic

health checks during an extended period of childhood (for example, up to 15 years).

During the transition period from one model of social transfers to another, accompany any

withdrawal transfers with up-scaling of Ajutor Social, especially considering the expected

impact of the withdrawal on poor households with children.

One way to ensure a smooth transition would be evaluating households regarding their

eligibility for Ajutor Social aid before suspending other kinds of payments from them (in

order to have a smooth transition and better control the cash flow in poor households).

Implement a new information campaign regarding eligibility criteria, the determination of

transfer amounts and the unavoidability of inequities generated during the transition period.

Provide additional assistance to the most vulnerable potential beneficiary groups of Ajutor

Social who are not in conditions of carrying out the bureaucratic conditions to access the

transfers- e.g. Roma families without personal identification documentation (this could be a

role for the upcoming Roma mediators planned by the MLSPF). This could happen inside

Ajutor Social with special (exceptional) administrative rules or even more efficiently with

nominative allowances as in the proposed case of rural households with four or more

children.

Reduce the application opportunity costs of Ajutor Social for the poorest households (e.g.

through a one-off first payment subsidy). Current reform proposals from the MLSPF and the

WB already address the speed of application approval. This will in fact reduce the time aspect

of opportunity costs; however, it does not solve the financial part of opportunity costs.

Knowing, for example, that some travel costs would be covered, independently from the

approval or disapproval of the application, could encourage a higher number of poor families

to initiate the application process.

Involve LPAs and Social Assistants deeper in the monitoring and orientation for the use of

Ajutor Social. Social inspection has already been created. Social Inspection staff could

receive training on how to monitor Ajutor Social beneficiary households and how to best

convince and motivate them to spend additional resources in favour of children in the

households.

Develop clear procedures for intervention and additional social support in cases of improper

use of funds (e.g. drug or alcohol abuse) received via Ajutor Social. Therefore, a kind of

referral system for social emergency cases could be considered. Moldova already has


41

successful experiences in referral systems, for example, in domestic violence and human

trafficking.

Consider the possibility of supporting the poorest families independently of their employment

condition. This proposal does not suggest creating a kind of non-contributable unemployment

insurance, but rather refers to the fact that the poorest of the poor children have the right to

grow up in dignity and develop their potential, independently of complex procedures that are

difficult to be accomplished by the household head.

Consider expanding the database for registering and monitoring all social cash transfers

(supported by the World Bank) to also include potential and existing beneficiaries of social

services. Such a database would help to increase cost-efficiency, by identifying overlapping

payments and exclusion errors regarding other households who should benefit but do not, as

well as facilitate the joint provision of cash transfers and social services.


42

Annex

Revised Documentation

Bradshaw, Jonathan, Emese Mayhew and Gordon Alexander, 2010, Minimum Social Protection for

Families with Children in the CEE/CIS Countries in 2009 – A Report for UNICEF, June 2010

Bradshaw, Jonathan and Dominic Richardson, An Index of Child Well-Being in Europe, Child Ind

Res (2009) 2:319–351, DOI 10.1007/s12187-009-9037-7

Conseil Sante, Sofreco & Adecri, Addressing needs of vulnerable layers of the population in the

Republic of Moldova, Chisinau, November 2010.

Dzyrgur, Yuriy and Kateryna Maynzyuk: Budgeting in the context of child care reform in CEE/CIS;

conference presentation at ―Child Care System Reform: Commitment, Partnership and

Action" covering Armenia, Belarus, Georgia, Moldova and Ukraine, Chisinau, Moldova,

November 2009

European Commission; Study on Social Protection and Social Inclusion in Moldova. Institute for

Development and Social Initiatives ―Viitorul‖, Chisinau, Moldova, 2009.

European Commission Moldova; Assessment of the Law Implementation Law no. 133/2008 on the

Social Support in the Republic of Moldova - First semester 2009, Chisinau 2009

Gassmann, Franziska and Geranda Notten, 2006, Size matters: Targeting efficiency and poverty

reduction effects of means-tested and universal child benefits in Russia, Maastricht University

Lino, Mark, Expenditures on Children by Families, 2006 (Miscellaneous Publication Number 1528-

2006), United States Department of Agriculture, Center for Nutrition Policy

Ministerul Economiei Moldova, 2010, Raport Privind Saracia si impactul policicilor 2009, Chisinau

MLSPF Annual Social Report 2010, Chisinau



NBS/UNICEF Moldova, Children of Moldova, Chisinau 2008

OPM/Every Child, Monitoring the Performance of the Social Support Programme: Ensuring the

Transition Towards a Targeted Social Assistance; Authors: Valentina Barca, Ludovico

Carraro, Alexandru Sinchetru, October 2010

OPM/Every Child, Monitoring the Performance of the Social Support Programme: new trends in 2010

and overall performance; Authors: Valentina Barca, Ludovico Carraro, Alexandru Sinchetru,

April 2010

OPM/Every Child, Study into Reasons for Low Take-Up and Retention; Authors: Valentina Barca,

Ludovico Carraro, Alexandru Sinchetru, February 2010


43

OPM/Every Child, Monitoring the Performance of the Social Support Programme: Analysis of MSAS

and HBS Data - The first 9 months of implementation; Authors: Ludovico Carraro, Alex

Hurrell, Valentina Barca and Luca Pellerano, December 2009

OPM/Every Child, Assessment of child care reform and technical assistance to subregional

consultation — Moldova, September 2009

OPM/Every Child, Moldova – Support to the Delivery of Effective and Sustainable Social Assistance

Services (DFID), Organizational Assessment of Raions, September 2008

OPM/Every Child, Informing the reform of cash benefits in Moldova - Insights from the analysis of

the 2006 Household Budget Survey, Ludovico Carraro, January 2008

OPM/Every Child, Moldova Social Assistance Reform - Beneficiary Assessment, October 2007

Oxford Poverty and Human Development Initiative (OPHI), 2010, Multidimensional Poverty Index

(MPI) At a Glance - Country Briefing Moldova, July 2010

Redmond, Gerry, Leonardo Menchini and Francesca Francavilla, The impact of cash transfers on

child well-being in Central Asia and Eastern Europe, Working Paper September 2007

Republic of Moldova, Government Decision No. 1167, 16/10/2008; Regulations on the establishment

and payment of social support; published: 21/10/2008 in the Official Monitor No. 189 art No.

1173

Republic of Moldova, Parliament Law No. 133, 13/06/2008 on Social Support; published: 30/09/2008

in the Official Monitor No. 179 art No. 625

Republic of Moldova, Parliament Law No. 123 18/06/2010 on Social Services; published: 03/09/2010

in the Official Monitor No. 155-158 art. No. 541, in force since 03/03/2011

Sumner, Andy, 2010: Child poverty, well-being and agency: What does a ‖3D Well-being‖ approach

contribute? Journal of International Development 22, 1064–1075 (2010), Published online in

Wiley Online Library, (wileyonlinelibrary.com) DOI: 10.1002/jid.1746

UNICEF Moldova, The Situation of Roma Children in Moldova, Chisinau 2010

UNIFEM Moldova, Approaches to social exclusion in Moldova. Methodological and analytical

aspects; Authors: Maria Vremiş, Viorica Toartă, Anatolii Rojco, Diana Cheianu-Andrei;

Chisinau 2010

World Bank, Project Appraisal Document: Strengthening the Effectiveness of the Social Safety Net

Project, May 4, 2011, Report No: 59913-MD

World Bank, Moldova: Consequences of Shifts of Remittances and Energy Prices for Consumption

and for Poverty Rates, June 2009, Report No. 49019-MD


44

Annex:

A. Interview guide for focus groups

I. Introduction

1.1. Presentation of the research theme

1.2. Discussion rules

- Honest answers,

- Wrong or right opinions do not exist, but rather only different points of view,

- Assuring the confidentiality,

- Necessity of audio recording.

1.3. Presentation of the participants (age, occupation, number of children)

- Think of your life; which are the things that worry you the most at the moment?

II. Social support

1. Please tell me what the main sources of income are for your household? What payments do

you still receive or have you received in the last 12 months from the state? Which of them,

do you think, is the most important? Why do you think so? And which is the less important

payment? For what reasons?

- Conditions for the reception of Ajutor Social (only for participants who receive Ajutor

Social; questions 1 to 5 were also posted to families who do not receive Ajutor Social but had

experiences with the reception of the previously existing monthly allowance of 50 MDL)

1. From what moment on did you receive social support payments? Where and how did you

find about the possibility of applying for social support?

2. How difficult was it for you to prepare the documents in order to receive this help? Please

describe what steps you took.

- Who helped you?

- How much time did it take? How many trips?

- What was the most difficult thing to do?

3. Tell me please, how do you spend the money received from Ajutor Social? To what extent

does the Ajutor Social help your family to cover its expenses?

4. Do you make specific expenditure for your children where you spend the Ajutor Social

money (on educational costs, health costs, food [more, better], other)? If so, is it now easier

for you than before to make these expenditures?

5. In general, from the moment on you started receiving Ajutor Social, what did you manage to

buy or do thanks to this payment?

6. After having received Ajutor Social payments for 6 months, did you reapply in order to keep

getting the payments? What are the reasons for reapplying or not reapplying?

7. Imagine that for whatever reasons you were to no longer receive any Ajutor Social, how

deeply would your family be affected? Please describe the concrete situation? What would

you do in this situation?

8. Did you hear about anybody who requested Ajutor Social and did not receive it? For what

reasons?


45

9. What categories of persons from your locality receive Ajutor Social? To what extent do you

think that Ajutor Social actually goes to the households in your locality who really require

this support? Why do you think so?

10. What category of persons from your locality receive nominative compensations (specify

which ones). To what extent you think these households are really poor and rely on these

payments?

11. Do you have knowledge of what these payments are usually spent on? Do you have concrete

examples?

12. What do you think about the involvement of the social assistant in the management of the

resources received? How do you see this involvement?

13. In your opinion, how should state institutions monitor and check how the money given for

help is spent?

14. Would you agree that there is a need for monitoring the adequate expenditure of these funds

for welfare purposes? If so, would you collaborate actively with such a monitoring

mechanism?

15. Would you agree that households, which do not adequately spend the resources received,

should suffer a certain kind of sanction (e.g. temporary interruption of payments)?

III. Health

1. Throughout the last year, did your child/your children get sick and you were not able to visit a

doctor? For what reasons you did not go to the doctor? What did you do instead?

2. To what extent and how could children benefit from medical services and treatment even in

the case of their parents not having enough money?


46

TECHNICAL ANNEX

Quantitative analysis – evidence from HBS 2009

Poverty indicators

UNICEF‘s main concern is on child welfare. Consequently, the first concern of the statistical analysis

addresses the question how well different poverty indicators are able to reflect the living conditions of

households with children.

Traditionally, the National Bureau of Statistics (NBS) is quantifying the poverty incidence

(percentage of poor population) in the country according to consumption levels. This is a generally

accepted methodology in many countries of the world. However, information included in the

Household Budget Survey (HBS) also allows the construction of alternative poverty indicators, such

as income poverty or other multi-dimensional poverty indicators.

Figure A1: Poverty levels for different indicators and household types


Figure A1 shows poverty levels for consumption poverty, income poverty and a Multi-Dimensional

Poverty Index (MDPI) estimated for this assessment. Consumption poverty is based on the official

indicator of per equal adult monthly consumption level and reflects the exact official poverty rate of

26.3% in 2009. Income poverty is based on the official per capita income indicator in HBS (per equal

adult monthly income level; assessed against an income poverty line of 875.15 MDL, meanwhile the

consumption poverty line is 945.90 MDL).

The MDPI is based on five different categories, taking into account information gathered during the

focus groups, regarding the use of received public transfers and based on general criteria of which

household conditions should be considered as favourable for a child‘s wellbeing. Leaving technical

details aside, an MDPI reveals the level of combination of different types of deprivations that batter a

household at the same time.

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

All households Hosueholds with

children under 18

Households with

children under 16

Households with

children under 5

Consumption poverty Income poverty Multidimensional poverty


47

Our MDPI consists of five categories. The first category is based on a traditional basic needs

approach of housing infrastructure and access to public services and considers the access to piped

water, sanitation, WC and heating. A household, which fails adequate access to any three of these four

categories, is considered as being poor in the first component. The second component refers to

overcrowding, applying the commonly used threshold of 2.5 or more persons per dormitory. The third

dimension refers to school enrolment of the population under 18 years of age. Dimensions four and

five are monetary dimensions assessing the share of household consumption that goes to health

expenditure and education expenditure. In both cases, a household is considered as being poor, once

its health or education expenditure exceeds the national mean household expenditure for these two

dimensions in more than 1/3.42

Going back to Figure A1, it can be noted that income poverty levels are higher than consumption

poverty levels and MDPI levels are lower. Observed differences could be consequences of the way

NBS measures income or the level (amount) of the poverty line. MDPI levels being lower than

monetary poverty basically means that a network of services and a minimum level of assets exist,

which have a wider range than the available minimum monetary income in order to make perform

these assets in a way to ensure a decent standard of living. Figure A1 shows a trend of decreasing

levels of income and MDPI poverty for households with younger children. We cannot appreciate the

same trend for income poverty. This fact suggests that income indicators are able to identify a larger

group of poor households with children, which are not being identified by the other indicators.

The important point regarding the differences of poverty levels associated to different poverty

indicators is not the difference of the level as such, but rather the assessment of which are the

identified households. We have to take a look at this fact since we can certainly not suppose that all

consumption poor households are a share of the group of income poor households (since the income

poverty level is higher) and that all MDPI poor households are a share of the group of income- or

consumption-poor households.

42

The estimation was implemented according to the Alkire and Foster 2007 approach available in the Stata

DASP module. The following weights are applied to the five different dimensions. Dimension 1 (services)

0.5467; dimension 2 (overcrowding) 0.1196; dimension 3 (enrolment) 0.0416, dimension 4 (health expenditure)

0.2614 and dimension 5 (education expenditure) 0.0307.


48

Figure A2: Share of exclusivity in identification of poor households using different indicators


Figure A2 shows that for income and MDPI, a level of roughly 20% and 40%, respectively, of

households were identified as poor, and are exclusively poor by these concepts (this is that a given

household ―A‖ is identified as poor regarding its income level but not regarding its consumption level

or regarding the dimensions of the MDPI). For both indicators, we can observe a slightly decreasing

trend when starting to consider households with children or households with younger children. In

contrast, the exclusive capacity of households to be identified as poor based on their consumption

level starts at similar levels to the MDPI, but shows an increasing trend when starting to consider

households with children. This result suggests that even if income poverty looks like the more

powerful indicator for identifying poverty in household with children, relying only on income poverty

would not allow the identification of a wide range of consumption poverty in households with

children. At a national level, more than 50,000 households with children, poor according to their

consumption level, are not identified as poor according to their income level. The number households

left out based on MDPI are roughly 30,000. Hence, the best choice to predict monetary poverty in

households with children is in fact consumption poverty, as usually used by NBS.

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0


children under 18

Households with

children under 16

Households with

children under 5

only income oncly consunsumption only MDPI


49

Figure A3: Share of coincidence in household identification as being poor


A different view of the same phenomenon is presented in Figure A3, which shows the level of

coincidence between households, which are identified as poor simultaneously by income and

consumption, by income and MDPI or by consumption and MDPI. As expected, income and

consumption show higher levels of overlapping than the combination with MDPI (almost at the same

level for both, MDPI & income and MDPI & consumption). Nevertheless, the levels of overlapping

are surprisingly low and all of these show a downward trend for households with children or younger

children. The level of coincidence of MDPI and monetary (income or expenditure) poverty predictors

of at least 25% illustrates that at least one quarter of households with children lack at the same time

assets, access to services and monetary income to make perform the existing level of assets and

services in a way, which ensures a decent living standard. Hence, this kind of poverty cannot be

overcome by only providing additional financial resources but by also providing access to assets and

services for households with children.

Welfare prediction

Results from the previous section strongly suggest that there are important characteristics of poverty

in households with children, which go beyond monetary conditions, and that poverty levels of

households with children are not identical to poverty levels at the national average. Since

beneficiaries of Ajutor Social aid are selected according variables which are built a set of best

predictors of different conditions of poverty, the next question that arises for our assessment is if it

would make sense to implement different prediction models for households with children and for

households without children (according to the same rational that currently justifies the existence of

different prediction models for urban and rural households). In order to obtain some empiric evidence

for responding these questions, we ran a couple of regression models, the results of which are shown

below. It is important to be aware that this assessment does not intend to suggest that there have to be

different models for households with children, or which should be the set of variables to be taken into

account. Instead, we want to present some evidence that can help to understand if it would make sense

to have different prediction models.

0.0

10.0

20.0

30.0

40.0

50.0

60.0


children under 18

Households with

children under 16

Households with

children under 5

income & consumption Income & MDPI consumption & MDPI


50

Table A1: Statistically significant variable sets for prediction of wefare indicators (HBS 2009)

A1.1 Prediction models with significant variables for consumption

All

HH

Poor

HH

HH with

children

under 16

HH with

children

under 5

Poor HH

with

children

under 16

Poor HH

with

children

under 5

Total persons in HH X X X X X

HH with children under 16 X

HH head female X X

Age of HH head X X X

Couple married or united X X X

HH with migrant abroad X X

Agriculture as main income source X X X X X X

Remittances as main income source X X X X

HH head with secondary education X X X X X

HH head with tertiary education X X X X X X

Rural household X X X X

HH central region X X

HH in Chisinau X X X X

N 5,532 2,056 1,425 581 539 218

R2 0.2618 0.1134 0.2604 0.2861 0.0629 0.1087


A1.2 Prediction models with significant variables for income

All

HH

Poor

HH

HH with

children

under 16

HH with

children

under 5

Poor HH

with

children

under 16

Poor HH

with

children

under 5

Total persons in HH X X X X

HH with children under 16 X X

HH head female X X

Age of HH head X X

Couple married or united X X

HH with migrant abroad X X X X

Agriculture as main income source X X X X


HH head with secondary education X X X X X X

HH head with tertiary education X X X X

Rural household X X X X X

HH central region X X X X X X

HH in Chisinau X X X X

N 5,532 2,056 1,425 581 539 218

R2 0.2649 0.2514 0.2503 0.2745 0.0814 0.1143



51

A.1.3 Prediction models with significant variables for MDPI

All

HH

Poor

HH

HH with

children

under 16

HH with

children

under 5

Poor HH

with

children

under 16

Poor HH

with

children

under 5

Total persons in HH X X

HH with children under 16 X

HH head female X

Age of HH head X X X

Couple married or united X X X

HH with migrant abroad X X X

Agriculture as main income source X X


HH head with secondary education X X

HH head with tertiary education X X X X X X

Rural household X X X X X

HH central region X X X X X X

HH in Chisinau X X

N 5,532 2,056 1,425 581 539 218

R2 0.2831 0.1927 0.2184 0.2037 0.0414 0.0831


The results of Table A.1 suggest three important findings: i) it is possible to identify a common set of

variables which performs similarly well in predicting levels of income, consumption or MDPI; ii) any

of these models loses prediction power trying to predict low levels of monetary resources or MDPI as

well as trying to predict welfare levels specifically only for households with children; and iii) the

estimated models consider variables of the family group and not (as current means-testing does)

mainly assets and durable goods.

impact of cash transfers on poverty and well-being of … · v. child welfare and social transfers...

Documents