Report on the Provision of Technical
Training on Multidimensional Poverty
Measurement in the Republic of Kazakhstan
Juliana Yael Milovich
Oxford Poverty and Human Development Initiative (OPHI)
July 2022
2
Table of Contents
Introduction ....................................................................................................................................................... 3
Goals and activities according to the Terms of Reference (ToRs) ....................................................... 4
Section 1: Description of the MPI technical workshop and the results achieved .................................. 5
Description of the daily activities ............................................................................................................... 5
First meeting with the statisticians and policymakers: Friday 8th July 2022 ........................................ 8
Challenges ...................................................................................................................................................... 9
Section 2: The AF method and its steps ........................................................................................................ 9
Example of the AF method ........................................................................................................................ 9
Practical steps to calculate the deprivation matrix ................................................................................. 11
Practical steps to aggregation, dimensional breakdown, and disaggregation by groups.................. 13
Section 3: Calculating the pilot MPI using the Household Budget Survey 2021 .................................. 16
Preliminary structure of the pilot MPI for Kazakhstan ........................................................................ 16
Weights ......................................................................................................................................................... 16
Deprivation cut-off and poverty cut-off ................................................................................................. 16
Preliminary results ....................................................................................................................................... 16
Section 4. Recommendations and Next Steps ............................................................................................ 20
Recommendations on the indicators ....................................................................................................... 20
Recommendations on the follow-up process ......................................................................................... 22
Technical next steps ................................................................................................................................... 23
Policy next steps .......................................................................................................................................... 24
Conclusions ...................................................................................................................................................... 24
Appendix ........................................................................................................................................................... 26
Agenda of the Technical Workshop ........................................................................................................ 26
List of institutions present during the policy meeting of Friday 8th July 2022 .................................. 30
Agenda of the meeting with policymakers on Friday 8th July 2022 (in Russian) .............................. 30
3
Introduction
While poverty measures have historically defined poverty as a lack of income, the lived experiences
of poor people suggest that poverty encompasses many more dimensions. A person who is poor can
suffer from multiple overlapping deprivations simultaneously. For example, they may have poor
health, lack access to clean water or electricity, or have insufficient schooling. A multidimensional
poverty measure can complement monetary poverty measures by presenting a more comprehensive
picture of the many deprivations experienced by the poor.
Multidimensional Poverty Indices (MPIs) based on the Alkire-Foster method allow the analysis of
both the incidence and breadth of poverty, as well as comparisons of the levels and composition of
poverty for different groups of populations, such as rural and urban areas, sub-national regions and
age groups. This reflects the commitment in the Sustainable Development Goals (SDGs) to “leave no
one behind” as well as the first area of priority of the National Development Plan of Kazakhstan until
2025, that is to ensure the “well-being of the citizens” through three priorities: “fair social policy”,
“affordable and effective health care system”, and “quality education”.1
The preamble for the 2030 Agenda for Sustainable Development states: “We recognize that
eradicating poverty in all its forms and dimensions… is the greatest global challenge and an
indispensable requirement for sustainable development.”. In this context, the development of a
national MPI would enable Kazakhstan to monitor and track poverty in all its forms and dimensions,
based on the national context and development priorities. This can be used for reporting towards
SDG target 1.2 and SDG indicator 1.2.2, which specifically urges countries to “by 2030, reduce at
least by half the proportion of men, women and children of all ages living in poverty in all its
dimensions according to national definitions”. The MPI could then report progress on this target, as
well as guide Kazakhstan’s own poverty reduction strategy.
Different countries around the world have designed and computed national MPIs with different goals
or purposes, some of which would include monitoring multidimensional poverty and helping to
coordinate social policies aiming to reduce poverty and deprivations. Depending on the context (and
the data available), countries have included a specific list of dimensions and indicators, which reflect
the needs of people living in the country. The final measure always reflects the situation of the country
and provides important information to monitor poverty and other development goals. In the case of
Kazakhstan, the National Development Plan until 2025, for instance, calls for improvements in the
budget allocation system, public administration plans, and monitoring and evaluation of the
development strategy, preserving the basic principle of state of planning “human-centeredness” to build
the national projects “around the needs and requirements of citizens”. In this context, a national MPI would
be a useful tool to guide these reforms. In addition, given the flexibility of the Alkire-Foster method,
it is possible to design and compute a national MPI that identifies the poorest population of the
country, provides information for social policies to target them, and identifies potential priorities to
reduce their levels of poverty.
The National Development Plan 2025 has 10 nationwide priorities grouped into three key areas: well-
being of citizens, quality of institutions and strong economy. The well-being of citizens direction
includes the following three priorities:
1 The National Development Plan of Kazakhstan until 2025 is a document of the first level of the State
Planning System and is developed to implement the long-term Development Strategy of Kazakhstan until
2050 and National Priorities: https://primeminister.kz/en/news/kazakstan-damuynyn-ulttyk-zhospary-
aleumettik-al-aukat-mykty-ekonomika-zhane-kolzhetimdi-densaulyk-saktau-1725726
4
Fair Social Policy: “provides for the implementation of systemic measures aimed at promoting productive
employment and ensuring social well-being”.
Affordable and Effective Health Care System: “envisions the development of the concept of a
sustainable health care system that contributes to the improvement, maintenance and restoration of people's
health, as well as the well-being of present and future generations”.
Quality of Education: “provides for the development of competitive human capital for the implementation
of the new economic course [where] the modernization of the education system will be aimed at improving its
quality and accessibility”.
These priorities are related to the provision of services and opportunities to the entire population in
Kazakhstan. However, to guarantee the right distribution of resources, it is necessary to identify who
the poor and most disadvantaged in Kazakhstan are. A national MPI will provide important evidence
on this topic, not only to identify the poor, but also to inform about the severity of their poverty and
the dimensions and indicators that contribute to create this situation. These priorities and the
additional strategies that are listed in the National Development Plan 2025, but also in other national
documents such as the “Strategy “Kazakhstan-2050””2, propose a list of targets that can be
incorporated as dimensions and indicators of the national MPI, so that it can serve as an official and
permanent statistic to monitor progress towards the achievement of the specific national goals.
This report is a summary of the mission undertaken by the consultant, Juliana Milovich, in Nur-
Sultan as part of the process to develop a National Multidimensional Poverty Index (MPI) for
Kazakhstan. The first section provides a description of the activities performed and the results
achieved during the technical workshop in Nur-Sultan; the second section discusses the preliminary
structure of the pilot MPI and the steps to build it using the Household Budget Survey 2021; and the
final section provides conclusions and recommendations for the next policy and technical steps
towards the construction and adoption of a MPI for Kazakhstan as an official and permanent statistic.
Goals and activities according to the Terms of Reference (ToRs)
The goal and main activities of the project were to:
Provide methodological training to the staff of the Bureau of National Statistics (BNS) of
Kazakhstan for developing a national MPI in Kazakhstan.
Review of the methodological materials available at BNS and UN Economic Commission
for Europe (UNECE), such as overview of data sources, the 2021 census questionnaire and
household budget survey (HBS) questionnaires, to identify the considerations and reasoning
behind the possible choices for the dimensions to be included in the MPI.
Prepare a package of training material and deliver training to BNS staff on the development
and use of a national MPI during a technical assistance mission to Nur-Sultan. The training
should cover the following:
o Overall counting based on the dual cut-off method (Alkire-Foster),
including an exercise;
o Indicator computation;
2 https://strategy2050.kz/en/
5
o Generation of deprivation matrix and weighted deprivation score at
household level;
o Aggregation to create the information platform comprising MPI,
Headcount Ratio, Intensity, and Indicator composition of multidimensional
poverty;
o Disaggregation by age, disability status, region etc;
o Robustness analysis;
o Trends over time.
Provide and explain the programming code for each learning area and share relevant online
resources.
Review the experimental calculations of MPI carried out by BNS and provide comments and
recommendations
Make a presentation at the meeting of statisticians and policymakers to show the possibilities
of using MPI in different policy contexts and the process of finalization and authorisation
of the index for use as an official statistic. Examples of various models in other countries
will be shared. Provide guidance on the practical steps needed for presenting to policymakers
how the MPI should be used and in what context.
Draft a report summarising the results of the work. The report should include a description
of the activities conducted, an analysis of the practical steps to be undertaken in presenting
and sharing with policy leaders how MPI should be used and in what contexts. Finally, it
should also provide conclusions and recommendations to BNS concerning further work on
the use of the MPI, including a consultation process with stakeholders on determining the
dimensions for the MPI, and on the interpretation of results.
Section 1: Description of the MPI technical workshop and the results achieved
This section describes in detail the activities during each day of the technical workshop, as well as
some notes on the first meeting with policymakers that took place on Friday 8th July 2022 and the
main challenges that were discussed. The Agenda of the technical workshop is found in the Appendix.
Description of the daily activities
Day 1
Due to travel disruptions, the consultant, Juliana Milovich, was unable to join the first day of the
workshop. Rafkat Fagamzianovich, an independent consultant of UNECE, facilitated the
presentations and together with the team from BNS, spent the first session of the workshop watching
the training videos from James Foster and Sabina Alkire on the AF method and the interpretation of
its results. They had a lively discussion on the methodology and noted some questions on how the
poverty cut-off can be selected and how to interpret the censored headcount ratios. In the afternoon,
they discussed the indicators and dimensions that could be considered in the structure of the pilot
MPI for Kazakhstan, based on the review of the variables available in the 2021 Household Budget
Survey (HBS) and how they could be improved. They decided to take out electricity because
deprivation is 0%, as well as income poverty in order to keep the latter for a complementary analysis
and exercise. They also identified four additional indicators: satisfaction with cleanliness of air,
cleanliness of territory and satisfaction with the quality of water. After extensive exchange with
Rafkat, the team was made aware that satisfaction variables are not ideal to accurately measure the
6
deprivations that people might face, but they decided to keep them for now, due to lack of other
relevant questions in the survey. In future surveys they will try to add more objective questions to
measure the indicators.
Day 2
Rafkat and I went to the Bureau of National Statistics (BNS) where I met Natalia Belonosova,
Gulzhan Daurenbekova, Marzhan Amerzhanova, Samal Kereibayeva, Rymzhan Kassenova, Dana
Malikova, Nagima Zhumanova, and Aizhan Makshayeva, from the Department of Labour Statistics
and Living Standards of the Bureau of National Statistics; Laura Kyndybai and Gulnar
Dilmagambetova, from the Department of Information Support of Household Statistics (Computer
Centre); Nauryz Baizakov, an econometrician-analyst from the Analytical Centre of the Information
and Computing Centre of BNS; Bakbergen Toktasyn, analyst of the Centre for Macroanalytics and
Forecasting of the Institute of Economic Research of the Ministry of National Economy of the
Republic of Kazakhstan; and the two interpreters, Azhar Suleimenova and Raushan Nukeshanova.3
We started the morning by addressing some doubts regarding the AF method, particularly on the
interpretation of the censored headcount ratios, the definition of the deprivation cut-offs and how
to set the poverty cut-off. We analysed these questions through an example using a matrix with four
indicators and four people, as well as the example of the global MPI on how we build and interpret
a deprivation profile. We reviewed the steps of the method, the interpretation of main results, and
how they can easily be communicated.
After the tea break, I presented the codes of the deprivation matrix, using the example of access to a
safe source of drinking water and the number of years of schooling. We went step by step first on a
white board and then on the Stata browser and the do.file of Stata, to show how we build the variables
for the applicable population at the individual and household level, how we build the indicator at the
individual and household level and how to deal with missing values in particular. We had a very
interactive session and all technical and non-technical participants familiarised themselves with the
methodology steps and the reasoning behind them and gained a good understanding on how to
interpret the results.
In the afternoon, we divided the participants in two: one technical working group comprising Laura
and Gulnar, and one non-technical, more policy-focused working group with Natalia, Gulzhan,
Marzhan, Samal, Dana, Rhymzan, Nagima and Aizhan. The policy team started working on a policy
table identifying the deprivation cut-offs of the indicators, linking the indicators to national and
international documents and the SDGs, identifying the policy/programmes in place that could track
the evolution of each indicator over time and the institutions that could be responsible for each
indicator. This was really good because they got involved in the work and felt very much motivated,
acknowledging that they play an important role in this process to provide the policymakers with the
accurate information that justifies the structure of the pilot MPI and how important this process is
to track poverty over time. On the technical side, Laura and Gulnar started cleaning the HBS data to
retain the variables that would be needed to build the MPI.
Rafkat did a tremendous work in complementing clarifications, organising the working groups, and
supporting Gulzhan and Marzhan with the more policy activities. And the whole day went by with
the wonderful support of the two interpreters who were translating continuously as we spoke, even
during lunch!
3 Refer to the List of participants in the Appendix for more detail.
7
Day 3
On Wednesday, we had a very intensive day, where everyone - divided in two working groups (policy
and technical) – was engaged and worked very hard. The policy group, supported by Rafkat, worked
on justifying the indicators with policy documentation, linking them to the SDGs, searching for
policies and programmes that could impact each indicator and the government institutions that
needed to be engaged. The technical group, supported by myself, worked together with the excellent
support of Raushan and Azhar (the interpreters), on cleaning, translating and merging the datasets,
and on specifying the deprivation cut-offs and applicable population of each indicator in the Pilot
structure. We ended the day with a final single database of the HBS 2021 survey, ready to be used the
next day to start building the indicators. At the end of the day, Gulzhan thanked everyone for the
hard work and their engagement. She felt the environment was lively, interactive and that they had
made significant progress.
Day 4
On Thursday morning, 7th July, I went through the codes in Stata that enable to compute the
aggregate measures of the incidence of multidimensional poverty (H), the intensity (A) and the
Multidimensional Poverty Index (MPI), recalling the detailed steps. I also presented the codes to
perform the decomposition analysis of multidimensional poverty by each of the indicators
considered, and the codes required to do the disaggregation analysis by subgroup of population,
using as example the regions of the country.
It was essential that the technical team understood how each line of code is translated to a specific
step of the process of measuring multidimensional poverty, so they can easily reproduce the codes in
SPSS using the syntax that OPHI shared. The interpretation of the results produced in each of the
steps of the process was also analysed, which is key for the policy-focused team to be able to explain
it to a wider audience.
In the afternoon, the working groups continued their respective work - policy and technical - and we
succeeded in finalising both tasks assigned. The policy-focused team searched to complete the policy
table, linking each indicator with policy documents, SDGs, government policies/programmes and
institutions in charge. The policy team also worked on preparing the presentation for the meeting
with the policymakers that was scheduled for the next day, Friday, 8th July. On their side, Gulnar and
Laura in the technical team computed all the indicators, except from the indicator on unemployment,
which required additional steps to consider. In this case, I guided them through its construction,
accounting for the applicable population (individuals aged 15 years old or more), and building first
the indicator at the individual level, to then build the final indicator at the household level.4 Most of
the indicators they had chosen were mainly identified with “level of satisfaction” questions and all
questions were answered by one person in the household, representing the whole household.
According to Laura and Gulnar, there was no missing information in most variables, everyone
answered and there was no option “don't know”. However, in two or three questions, the option
“don't know” was present, and a high percent of missing values (up to 7%) was found. This will
need to be reviewed to check for any possible bias in the data, and in the case this bias exists, the
indicator might need to be re-considered. Since the MPI is built using complete information for all
4 See the sub-section of the “Practical steps to calculate the deprivation matrix” for detailed related
information.
8
the indicators, the data observations that had missing information for at least one indicator were
dropped. This was performed for the time being, to be able to make progress in producing the
preliminary results. However, the team still needs to do a proper analysis on this.
For the computations of the indicators, Gulnar and Laura worked on SPSS, directly using the toolbox,
and I worked on Stata, building the syntax in the do.file. Each time an indicator was computed, we
compared results to check that the number of observations identified as deprived and non-deprived
were the same. All results matched. Now, with a bit more time and the SPSS syntax available to them,
it is recommended that they write the syntax in SPSS, so it will be easier to follow the computations
and also to transfer their knowledge to other members of their team in the future.
By the end of the working day all the indicators of the pilot MPI structure were built together with
the corresponding policy justifications and the presentation of the policy team for the meeting with
the policymakers was ready. I worked on the aggregation, dimensional breakdown, and regional
disaggregation, and prepared some figures for their presentation. Before leaving the office, we
received the visit from the Director General of the BNS, who was debriefed about all the work that
had been done by the team and was very keen in learning from the final preliminary results and the
continuity of the work towards the finalisation of the MPI.
Day 5
On Friday morning, the policy team worked on finalising their presentation for the meeting, including
the preliminary results that were produced for the preliminary structure of the pilot MPI. Between
11:00 and 1:30pm we had the meeting with the policymakers, and in the afternoon we had a wrap-
up meeting to discuss and agree on the next steps. We decided to continue working with recurring
meetings and clear to-do tasks, both on the policy and technical fronts.
All in all, it was a very intensive and memorable week, where the BNS team and the computer centre
team made herculean efforts to learn and work and prepare themselves for the important presentation
to the policymakers, which seems to have been a great success.
First meeting with the statisticians and policymakers: Friday 8th July 20225
Representatives from the Ministry of Labour and Social Protection of the Population, the Ministry
of National Economy, the Agency for Strategic Planning and Reforms and the Analytical Centre of
the Information and Computing Centre of BNS, were present during the meeting, as well as
colleagues from UNICEF. Natalia made the welcoming remarks, and I did a short presentation on
the importance of measuring multidimensional poverty and the policy use of the MPI. Afterwards,
Samal and Marzhan presented the work the BNS team had done on developing a national MPI for
Kazakhstan, including the need for an MPI in the country, the review of the questionnaire and policy
documents and the preliminary selection of the indicators, and finally the preliminary results. The
presentation was followed by a lively discussion and the BNS team received much constructive and
positive feedback from all the participants. Afterwards, the UNICEF team presented some
observations from the additional questions they collected on children. The session ended with a
separate discussion between the ministerial representatives and the BNS team on how they will
proceed for further exchange. A short exchange took place between UNICEF, Natalia, and myself
on the possibility of joining forces between the process of building a national MPI and that of
5 See the Appendix for the Agenda (in Russian) of this meeting.
9
measuring child poverty, so that the measures can effectively be used to guide policy action in
Kazakhstan.
Challenges
The main challenges we faced during the workshop had to do with language, type of software and
time. Myself not being a Russian speaker, I depended on the support of the interpreters to be able to
communicate with the team and the policymakers. This was nonetheless easily overcome thanks to
the fantastic work of the interpreters and the great synergies that the whole team built in such a short
time, which made the communication easy and fluent. The statistical software used for the
computation of the MPI posed a challenge too, since the technical team in Kazakhstan is expert in
the use of SPSS and FoxPro, and myself in the use of Stata. However, this challenge was overcome
through the detailed analysis of the codes that enable to compute each of the steps to build the MPI
and perform its corresponding analysis Moreover, as previously mentioned, when building the
deprivation matrix, the work was simultaneously done in SPSS by the BNS team and in Stata by
myself. The BNS team performed the main calculations and I reviewed them afterwards, sharing
comments and recommendations. This procedure also made it possible to compare the results
obtained through both software and ensure that they matched. In addition, the entire syntaxis to
build the MPI and perform the composition and disaggregation analysis in SPSS, has also been made
available to the entire team, so that the technical team can work in adapting it according to the
structure of the MPI for Kazakhstan and its analysis. Finally, time was very short, and it was essential
to build capacity on the AF method, the computation of the pilot MPI for Kazakhstan and the
interpretation of its preliminary results. However, the BNS team worked steadily and intensively
during the entire week, entirely committed and engaged with the work they were developing. This
was key for the smooth running of all activities in the time that was available. Therefore, as detailed
above, we managed to overcome these challenges and make steady progress on all fronts.
Section 2: The AF method and its steps
This section introduces the AF method through a very simple example using four indicators and four
people in an example society. It also presents the detailed steps of the coding process to build the
deprivation matrix and the aggregation, dimensional breakdown, and disaggregation steps for the
production of results and analysis.
Example of the AF method
Suppose there is a hypothetical society containing four people and multidimensional poverty is
analysed using four indicators: hectares of land, years of schooling, malnourishment, and access to
improved sanitation. The 4x4 matrix X contains the achievements of the four people in the four
indicators.
Hectares of
Land
Years of
Schooling
Body Mass
Index
Access to
Improved
Sanitation
7 14 19 Yes Person 1
3 13 19.5 No Person 2
4 3 17 No Person 3
8 1 22 Yes Person 4
z = 5 5 18.5 Yes
X =
10
For example, Person 3 owns 4 hectares of land, whereas Person 4 owns 8 hectares. Person 1 has
completed 14 years of schooling, whereas Person 2 has completed 13 years of schooling. Person 3 is
the only one who is malnourished of all four persons. Two persons in our example have access to
improved sanitation.
Thus, each row of matrix X contains the achievements of each person in each of the four indicators.
The deprivation cut-off vector is denoted by z = (5, 5, Not malnourished, Has access to improved
sanitation), which is used to identify who is deprived in each indicator. The achievement matrix X
has three people who are deprived (see the underlined entries) in one or more indicators. Person 1
has no deprivation at all.
Based on the deprivation status, we construct the deprivation matrix g0, where a deprivation status
score of 1 is assigned if a person is deprived in an indicator and a status score of 0 is given otherwise.
All indicators are equally weighted and thus the weight vector is w = (0.25, 0.25, 0.25, 0.25). We then
apply these weights to the deprivation matrix to obtain the weighted deprivation matrix. The weighted
sum of these status scores is the deprivation score (ci) of each person. For example, the first person
has no deprivation and so the deprivation score is 0, whereas the third person is deprived in all
indicators and thus has the highest deprivation score of 1. Similarly, the deprivation score of the
second person is 0.5 (0.25 + 0.25).
The union identification approach identifies a person as poor if she is deprived in any of the four
indicators. In that case, three of the four people in this example are identified as poor (i.e. persons 2,
3 and 4). On the other hand, the intersection approach requires that a person is identified as poor if
she is deprived in all indicators simultaneously. In that case, only one of the four people is identified
as poor in this example (i.e. person 3). An intermediate approach sets a cut-off between the union
and intersection approaches, say, k = 0.5, which is equivalent to being deprived in two out of four
equally weighted dimensions. This strategy identifies a person as poor if she is deprived in half or
more of the weighted indicators, which in this case means that two of the four people are identified
as poor (i.e. persons 2 and 3).
Once the poor have been identified, the weighted deprivation matrix is censored so that the measure
can focus only on the deprivations of the poor – that is, deprivations of those identified as non-poor
Hectares of
Land
Years of
Schooling
Body Mass
Index
Access to
Improved
Sanitation
0 0 0 0 Person 1
1 0 0 1 Person 2
1 1 1 1 Person 3
0 1 0 0 Person 4
w = 0.25 0.25 0.25 0.25
g 0 =
Hectares of
Land
Years of
Schooling
Body Mass
Index
Access to
Improved
Sanitation
Deprivation
score, c i
0 0 0 0 0
0.25 0 0 0.25 0.5
0.25 0.25 0.25 0.25 1
0 0.25 0 0 0.25
=
11
are replaced with a zero. This leads to the censored deprivation matrix and the censored deprivation
score, as shown below for k = 0.5.
Note that there is one case where the censoring is not relevant: when the poverty cut-off corresponds
to the union approach, then any person who is deprived in any dimension is considered poor and the
censored and original matrices are identical.
As discussed above, the headcount ratio H is the proportion of people who are poor, which is two
out of four persons in the above matrix. That is, H = 2/4 = 1/2 or 50%.
The intensity A is the average deprivation share among the poor, which in this example is the average
of 0.5 and 1 (i.e. the deprivation scores of the two people that are poor, persons 2 and 3). That is, A
= 0.75 or 75%.
It is easy to see that the multidimensional headcount ratio or MPI is M0 = H x A = 0.5 x 0.75 =
0.375. It is also straightforward to verify that M0 is the average of all elements in the censored
deprivation score vector c(k), i.e. M0 = (0 + 0.5 + 1 + 0)/4 = 0.375. Analogously, it is equivalent to
compute as the weighted sum of deprivation status values divided by the total number of people: M0
= (0.25*2 + 0.25*1 + 0.25*1 + 0.25*2)/4 = 0.375.
Following the explanations above, the analysis can be completed by computing decompositions by
populations subgroups and dimensional breakdowns.
Practical steps to calculate the deprivation matrix
The deprivation matrix is composed of various vectors, where each one provides information on the
deprivation conditions of each individual in a specific indicator. To build the deprivation matrix, one
needs to create each of the indicators that are considered in the structure of the MPI, by identifying
the deprivation or non-deprivation of each person and household in the dataset.
To create each of the indicators, there are five main considerations:
1- Unit of identification: it is essential to clearly specify the unit of identification, that is, who
is identified as poor or non-poor, which is the same as who is identified as deprived and non-
deprived (a person, a household, an institution, a geographic region, for instance). This
differs from the unit of analysis, which refers to how data are reported (often at the individual
level in percentage of people).
2- Applicable population: it is also key to precisely define the applicable population, which is
the group of people for which the indicator is relevant.
3- Deprivation cut-off: it is also essential to clearly specify the deprivation cut-off, which is
the minimum realisation that a person needs to satisfy in order not to be identified as
deprived.
4- Coding/Labels/Answers of the raw variable(s): it is important to check the answers of
each of the raw variables that will be used to build the indicator of the deprivation matrix.
Hectares of
Land
Years of
Schooling
Body Mass
Index
Access to
Improved
Sanitation
Censored
deprivation score,
c i (k )
0 0 0 0 0
0.25 0 0 0.25 0.5
0.25 0.25 0.25 0.25 1
0 0 0 0 0
g 0 (k) =
12
5- Filter(s) in the questionnaire: it is also necessary to check whether there are any jumps or
filters between questions within the survey questionnaire, that need to be taken into
consideration when building a specific indicator of the deprivation matrix.
We can see this with the following example. Suppose we would like to consider in the MPI an
indicator of School Attendance, to measure whether children are attending school or not. The definition
of the indicator is the following: A household is deprived if any school-aged child (6-14 years old) is not currently
attending school.
1- Unit of identification: the unit of identification is the household. That is, all the members
in the household are going to be deprived if at least one school-aged child, between 6 and
14 years old, is not currently attending school.
2- Applicable population: in this example, the applicable population is school-aged children,
between 6 and 14 years old. First, we are going to identify the individuals between 6 and 14
years old who are deprived or non-deprived in this indicator
3- Deprivation cut-off: we identify whether a child between 6 and 14 years old is deprived or
not in this indicator, using the deprivation cut-off. In this example, this is “not currently
attending school”. If a school-aged child is not currently attending school, she or he are going
to be identified as deprived. On the other hand, if a child between 6 and 14 years old is
currently attending school, then this child is going to be identified as non-deprived.
4- Coding/Labels/Answers of the raw variable(s): the answers to the question used in the
data differs between surveys and countries. It could be for example: 0 for non-attending; 1
for attending. In some surveys, the non-response is coded with a number 8 or a number 9.
It’s essential to check these answers and identify which of them are identified as a
deprivation, which as a non-deprivation, which as a missing value (non-response).
5- Filter(s) in the questionnaire: there could be some filters in the questionnaire. For
instance, in some surveys children who never ever attended school may not answer the
questions related to current school attendance. In this case, these children might have a non-
response or missing value in the indicator, and it is important to decide whether they would
be identified as deprived or non-deprived.
Once these considerations are taken into account, the main steps in practice to code the indicator
using the statistical software are the following:
a. Create a variable for the applicable population: in the example of the School Attendance
indicator, we would create a variable equal to 1 if the individual is between 6 and 14 years
old, and 0 otherwise. If a person doesn’t have information on the age, then this variable will
have a missing value for this person.
b. Create a variable for the applicable population at the household level: we identify the
households who have children between 6 and 14 years old and those who doesn’t. We would
create a variable equal to 1 for all the members of a household where there is a child between
6 and 14 years, and equal to 0 for all the members of a household where there is not a child
between 6 and 14 years old.
c. Create a variable for the deprivation at the individual level: we create a variable that
identifies whether a school-aged child is currently attending school or not. This variable will
take the value of 1 if a child between 6-14 years old is not attending school, thus she or he is
deprived, and equal to 0 otherwise, thus non-deprived. Note that this variable has missing
values for all the members of the household who are younger than 6 years or older than 14
years. It is only created for the applicable population.
13
d. Create a variable for the deprivation at the household level: we identify the households
who have at least one school-aged child6 who is deprived and those where all school-aged
children are non-deprived. We create then a variable equal to 1 for all members of a
household where there is at least a child 6-14 years old who is deprived, and equal to 0 for
all the members of a household where all the children 6-14 years old are non-deprived. Note
that this variable will have missing values for all the members of a household where there
are no children aged 6-14 years old. This variable is the one that corresponds to the indicator
of the deprivation matrix.
e. Replace with a non-deprivation the individuals living in a household where there is
no applicable population: if in the household there is no member aged 6-14 years old, the
previous variable will have missing values. A normative decision would need to be taken on
whether these individuals should be identified as deprived or non-deprived or be left as
missing. Note that if they are left as missing, they will not be considered in the final
calculations of the MPI. Since we would like to keep as many observations of the dataset as
possible for the final calculations, we may want to identify these individuals with a non-
deprivation. In this case, we replace the indicator with a 0 for all members of a household
where there is no child 6-14 years old. We have then the final indicator.
f. Analyse and properly identify the missing values: once we have computed the final
indicator, we want to analyse the number of observations who are deprived, those who are
non-deprived and those who have missing information. This enables to check whether the
numbers are correct according to the country context and study which are the individuals
for which there is no information on the indicator. If the percentage of observations who
doesn’t have information on the indicator is sufficiently large7, then it is recommended to
study whether there is bias in the data -for instance, the missing information corresponds to
individuals living in rural areas-, or whether this missing information is randomly assigned.
If there is a doubt about possible bias in the data, it’s recommended to reconsider the
definition of this indicator, in order to account for a more complete source of information.
Note that some of these steps are not required when the indicator that is built uses information that
is the same for all the members of the household, such as an indicator to measure a deprivation on
the type of source of drinking water, for instance. In this case, there is no need to build the variables
for the applicable population and the indicator at the individual level. The indicator at the household
level can be directly created by identifying the answers to the questions used that correspond to a
deprivation, those that correspond to a non-deprivation, and those that don’t have an answer and
thus are missing values.
The deprivation matrix is then made up of different vectors, each corresponding to an indicator of
the structure of the MPI, and with values equal to 1 for a deprivation and equal to 0 for a non-
deprivation.
Practical steps to aggregation, dimensional breakdown, and disaggregation by groups
Once the deprivation matrix is built, there are several steps to implement in practice in order to
compute the incidence of multidimensional poverty (H), the intensity of multidimensional poverty
(A) and the MPI:
1- Keep relevant sample: we keep only the observations for which we have information in all
indicators. The observations that don’t have information (have missing values) in at least one
6 In this example all household members are identified with a deprivation if at least one school-aged child is
deprived. But it is important to note that this can be different, for instance, it can be all children, or half the children
or other definition that corresponds best to the country context.
7 There is no golden rule to identify how much is sufficiently large, but in OPHI we usually consider 2% as a
maximum percent of missing values to consider the indicator.
14
indicator will be dropped from the final sample. It is also important to check in this step
whether there are members of the household who are non-permanent and might not need
to be considered in the final calculations.
2- Declare the sampling design of the survey: if the data source used to build the MPI is a
survey and not a census, then the sample of the survey is considered to be representative of
the entire country population by using three variables of the sampling design: the sampling
weight, the strata, and the primary sampling unit (psu). It is key to identify these three
variables that guide the sampling design of the survey, so that all the results will be estimated
accurately and will be representative of the national population.
3- Define the weights of each indicator: for each indicator, a variable will be created with
the relative value of the weight8 that is given to the specific indicator. For instance, if the
School Attendance indicator is given a value of 1/6, then the variable of the weight will be
equal to 1/6 for all the observations in the dataset. Note that if there are, for instance, 10
indicators in the structure of the MPI, there has to be 10 variables for the indicators, each
one measuring the deprivations and non-deprivations of each person in the sample; and 10
variables for the weights, each one capturing the relative value of each indicator within the
structure of the MPI.
4- Build the weighted deprivation matrix: once the deprivation matrix is built and the
relative weights of the indicators are set, we create the weighted deprivation matrix by
multiplying each vector of the deprivation matrix (each indicator) by its corresponding
weight. We will obtain a matrix equal to the deprivation matrix, but instead of having a
number equal to 1 for a deprivation, we will have a number equal to the value of the relative
weight of that specific indicator (1/6 for the current example of School Attendance), and a
number equal to 0 for those individuals who are non-deprived.
5- Create the counting vector: once the weighted deprivation matrix is built, we create a
variable called the counting vector, by adding up all the weighted deprivation of each person
in the sample. This variable provides the deprivation profile of each person in the sample.
This deprivation profile can take a value between 0 (not deprived in anything) and 1
(deprived in everything).
6- Identify the poor persons: once the deprivation profile of each person is built, we identify
who is poor and who is not by comparing the value of the counting vector of each person
to the value of the poverty cut-off. A person is identified as multidimensionally poor if
her/his deprivation profile is equal or higher than the poverty cut-off. For instance, if the
deprivation profile of a person is equal to 1/2 (or 50%) and the poverty cut-off is equal to
1/3 (or 33%), this person is identified as multidimensionally poor because her/his
deprivation profile is higher than the poverty cut-off. We then create a variable equal to 1 if
the person is identified as multidimensionally poor, and equal to 0 if the person is not
multidimensionally poor.
7- Create the censored counting vector and the censored deprivation matrix: once the
multidimensionally poor people are identified, we build a variable equal to the counting
vector, but we replace with a 0 (a non-deprivation) the deprivation profile of the individuals
who are not identified as multidimensionally poor. We do this because, when building the
MPI, we only consider the deprivation profile of the persons who are identified as
multidimensionally poor. The minimum value of this censored counting vector can be at
least equal to the poverty cut-off (in this example 1/3) and at most equal to 1 if a poor person
is deprived in every indicator. We do the same with the deprivation matrix and we create the
same indicators that we have in the deprivation matrix, but we censor or replace with a 0 the
deprivations of non-poor people.
8 Note that this is the weight of the indicator within the MPI structure. That is, the relative value that is given
to a specific indicator. It does not correspond to the sampling weight of the survey.
15
8- Calculate the incidence and the intensity of multidimensional poverty, and the MPI:
we can then calculate the incidence of multidimensional poverty by taking the average of the
identification vector built in step 6, using the sampling design of the survey. Recall that this
vector of identification takes the value equal to 1 if the person is identified as
multidimensionally poor, and equal to 0 otherwise. Therefore, by taking the average of this
vector and using the sampling design of the survey, we count the number of people who are
poor in the society, and we divide it by the total population, obtaining the percentage of
people who are multidimensionally poor – the incidence (H). To calculate the intensity of
multidimensional poverty (A), we take the average of the censored counting vector only for
the poor people in the society. Recall that this vector contains the sum of the weighted
deprivations experienced by each poor person in the sample and is equal to 0 for non-poor
people. To calculate the average of this vector only for poor people, we sum the total
weighted deprivations of all poor people, and we divide it by the total number of poor
people. Using the sampling design of the survey, we identify the average share of deprivations
that poor people experience in the society. The MPI, which is the multiplication of H and
A, is calculated with the statistical software as the average of the censored counting vector
for the entire population (poor and non-poor). To calculate the average of this vector for
the entire population, we sum the total weighted deprivations of all persons in the sample
(poor and non-poor9), and we divide it by the total number of people in the sample. Using
the sampling design of the survey, we obtain the MPI, which is interpreted as the share of
possible deprivations experienced by poor people in the society.
Once the incidence (H) and the intensity of multidimensionally poverty (A) and the MPI are
estimated, it is possible to decompose the level of poverty by indicators and analyse:
9- The censored headcount ratios: this is the percentage of the population who is poor and
simultaneously deprived in each indicator. It is calculated by taking the average of each vector
of the censored deprivation matrix. It enables to identify the highest deprivations faced by
poor people in the society.
10- The contributions of each indicator to overall poverty, in absolute value and in
percentage: the absolute value of the contributions of each indicator is calculated by
multiplying the value of censored headcount ratio of each indicator with the relative weight
of the indicator. The percentage contribution is calculated as the absolute contribution
divided by the MPI. It enables to identify which are the deprivations that contribute more to
overall poverty and inform policy action accordingly.
Once the results are obtained at the national level, they can also be disaggregated by subgroup of
population that is relevant to analyse within the country context. For instance, it could be relevant to
analyse the incidence, the intensity, the MPI and its decomposition by age groups. In this analysis,
steps 8, 9 and 10 are replicated for each subgroup of population. In the example of the age groups,
for instance, this would enable to calculate the incidence and the intensity of multidimensional
poverty and the MPI for children, for adults and for the elderly. Then, we could calculate the censored
headcount ratios and the contributions of each indicator to the MPI of children, of adults and of the
elderly, providing a very detailed and comprehensive picture of poverty in the country.
9 Recall that non-poor people have a value of 0 in the censored counting vector because they are either not
deprived in anything or their deprivation profile is lower than the poverty cut-off.
16
Section 3: Calculating the pilot MPI using the Household Budget Survey 2021
This section presents the preliminary structure of the pilot MPI for Kazakhstan, as well as the
preliminary results of multidimensional poverty at the national level, its decomposition by indicator
and the contribution of each indicator to overall poverty by region.
Preliminary structure of the pilot MPI for Kazakhstan
The preliminary structure of the national MPI for Kazakhstan has four dimensions and 16 indicators
(Table 1), carefully discussed, and justified by national and international policy documents.10
The MPI is built by using the household deprivation profiles in these indicators.
Weights
For the purpose of the current exercise, all dimensions are weighted equally and each indicator
within each dimension is also weighted equally.
Deprivation cut-off and poverty cut-off
The AF method uses a dual cut-off. First, it is determined whether a person is deprived or not in
each indicator using an indicator cut-off. If an individual’s achievement falls below the indicator cut-
off, then he/she is considered as deprived in that indicator. The deprivation cut-offs, which are
specified in Table 1, have been decided in a normative way and their justification is detailed in the
Policy Table in the Appendix. Secondly, the AF method uses a poverty cut-off (k) to identify whether
a person is multidimensionally poor or not. If the deprivation profile of a person – calculated as the
sum of the weighted deprivations that the person experiences – is equal or higher than the poverty
cut-off, the person is identified as multidimensionally poor.
In the current exercise, the poverty cut-off has been set to 1/4 or 25 percent. This implies that, given
that the preliminary structure has four dimensions, and each dimension is weighted equally (25
percent for each one), a person is identified as multidimensionally poor if she/he is deprived in one
or more dimensions.
Preliminary results
This sub-section presents the preliminary results for the national MPI of Kazakhstan, using the
Household Budget Survey 2021. We first present the national MPI as well as the incidence and
intensity of poverty among the poor. We then show how people are poor according to each indicator,
who is poor among different regions in the country and how each indicator contributes to overall
poverty in each region.
Table 2 shows that the incidence of multidimensional poverty or the poverty rate in Kazakhstan is
23.6%, meaning that nearly one out of four people in Kazakhstan is multidimensionally poor by the
national MPI11. The intensity of poverty, which reflects the share of weighted deprivations each poor
10 See Policy table in Appendix for detailed information.
11 Since all survey-based estimates are based on a sample, each has a margin of error. Thus, the 95% confidence
interval is also presented in the table. In the case of the incidence of the national MPI, we can say with 95%
confidence that the true headcount ratio of multidimensional poverty of the entire national population is
between 21.0% and 26.2%.
17
person experiences on average, is 32.2%. This indicates that each poor person in Kazakhstan is, on
average, deprived in 32.2% of the weighted indicators. The national MPI has a value of 0.076.
Table 1: Pilot MPI Kazakhstan - Dimensions, indicators, weights, and percent population
deprived in each indicator (%)
Recall that the MPI is calculated by multiplying the percentage of population who is
multidimensionally poor (the incidence, H) by the share of weighted deprivations that the poor
people face on average (the intensity, A). The value of 0.076 shows that poor people experience 7.6%
of the total possible deprivations that could be experienced if everyone was deprived in all
Deprivation cutoff
The household is deprived if…
Quality of
education
the level of satisfaction is 1-3 over 10 1/12 5.0%
Accessibility of
education
the level of satisfaction is 1-3 over 10 1/12 5.0%
Attendance at
preschools
at least one child 1-6 years of age is not attending
preschool due to the following reasons: preschool
is expensive or preschool is far way or relatives
look after the children or the child doesn't have
residential registration
1/12 7.4%
Quality of
health services
the level of satisfaction is 1-3 over 10 1/28 7.7%
Accessibility of
health services
the level of satisfaction is 1-3 over 10 1/28 5.6%
Inability to
access health
services
at least one member 15+ who was sick during the
year couldn't access the health services due to:
services being very expensive or the medicine is
too expensive or long queues or absence of
specialist or health care facility is too remote/no
opportunity to access or absence of medicines or
poor quality of services/don't trust
1/28 7.5%
Clean air
the level of satisfaction is 1-3 over 10 (absent of
pollution, smoke, dust, muds)
1/28 5.8%
Cleanliness of
the surrounding
area
the level of satisfaction is 1-3 over 10 (Absence of
waste or garbage )
1/28 2.5%
Source of
drinking water
the households gets the water from tank truckers
or river/ponds/lake
1/28 2.6%
Quality of
drinking water
the level of satisfaction is 1-3 over 10 1/28 7.6%
The standard of
accommodation
(sqm)
a person lives in less than 15 squared meters 1/16 47.4%
Fuel for heating the household uses solid or liquid fuel for heating 1/16 24.3%
Sewerage
(sanitation)
the household has a toilet with pit latrine without
slab or no toilet or septic tank
1/16 40.8%
Access to the
internet
the household doesn't have personal access to
internet
1/16 33.0%
Household debt
a person 15+ failed to pay rent/mortgage, loan or
utility services twice or more
1/8 19.1%
Unemployment
at least one person 15+ if it's not working
(unemployed or not searching for job)
1/8 11.6%
Pilot MPI for Kazakhstan
Percent
population
deprived
(%)
1/4
1/4
Weight of
the
indicator
Weight of
the
dimension
1/4Education
Health &
Environment
Housing and Living
Conditions
Living
Standards/financial
inclusion
Dimension Indicators
1/4
18
indicators.12 The national MPI is the official statistic because it is most precise and most sensitive to
change - if any deprivation of any poor person goes down, the MPI will go down - but for non-
technical users, incidence may be more intuitive, so it is usual to always discuss both.
Table 2: Multidimensional poverty in Kazakhstan
Poverty cut-off (k) Value Value Confidence Interval (95%)
k-value = 25%
(deprived in 1
dimension or
more)
MPI 0.076 0.067 0.085
Incidence of multidimensional
poverty (H, %)
23.6% 21.0% 26.2%
Intensity of multidimensional
poverty (A, %)
32.2% 31.4% 32.9%
Source: Author’s calculation based on data from the Household Budget Survey, 2021.
At the regional level, multidimensional poverty varies substantially (see Figure 1). The region of
Shymkent City and Pavlodar Province are the least poor regions in Kazakhstan, whereas Kostanay
and Turkestan Provinces are the poorest regions in the country.
Figure 1: Multidimensional Poverty Index by region
Source: Author’s calculation based on data from the Household Budget Survey, 2021.
Figure 2 presents the percentage of the population who is multidimensionally poor and deprived in
each of the indicators. These are called “censored headcount ratios”. The analysis of the censored
headcount ratios shows those indicators in which the national MPI poor people face the highest
levels of deprivation. A reduction in any deprivation of any poor person (that is, a reduction of any
censored headcount ratio) will reduce the national MPI and improve the lives of poor people in
Kazakhstan.
Figure 2 shows that a large percentage of people who are multidimensionally poor are also deprived
in sanitation (15.6%). Providing an improved sewerage systems will reduce this deprivation, which
affects around 2.6 million people in Kazakhstan. Moreover, 14.8% are multidimensionally poor and
12 With 95% confidence, the true value of the MPI is between 0.067 and 0.085.
0,000
0,020
0,040
0,060
0,080
0,100
0,120
0,140
0,160
0,180
0,200
19
live in a household were each member lives in less than 15 square meters, while 13.2% are
multidimensionally poor and don’t have access to internet. A similar percentage of the population in
Kazakhstan (13%) is poor and lives in a household where at least one person aged 15 years old or
more, has failed twice or more times to pay his/her rent or mortgage, or loans, or utility services.
Confronting these deprivations are top priorities for poverty reduction in Kazakhstan.
Figure 2: Proportion of population who is poor and deprived in each indicator (%)
Source: Author’s calculation based on data from the Household Budget Survey, 2021.
To chart policy priorities and design high-impact policies in Kazakhstan, Figure 3 shows the
percentage contributions of each of the weighted indicators to the national MPI for each region.
Regions are ranked from the poorest to the least poor, according to the national MPI numbers
presented in Figure 1. In Shymkent City, the region with the lowest level of multidimensional poverty,
unemployment has the largest contribution to the MPI. Debt also contributes most significantly to
Turkestan and Kostanay poverty, the poorest provinces in Kazakhstan. Inadequate standard of
accommodation (<15 square meters per person) contributes equally to poverty in all regions.
To use the percentage contributions for policy, consider the following example. The province of
Jambyl and East Kazakhstan (EKR) have nearly the same MPI value, and so one might think that
anti-poverty policies would be the same. But unemployment and debt contribute more to poverty in
Jambyl Province than in EKR Province, whereas deprivations in the level of satisfaction with the
quality of education, attendance of preschool and the level of satisfaction with the quality of health
services contribute more to poverty in EKR Province. In terms of policy this means that a uniform
approach is not cost-effective, because the different composition of indicators in each province
requires different policy and budgetary responses.
3,5% 3,5% 4,0% 4,0% 3,2% 2,8% 2,6% 1,6% 1,4%
3,3%
14,8%
11,1%
15,6%
13,2%13,0%
7,6%
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
20%
Q
u
al
it
y
o
f
ed
u
ca
ti
o
n
A
cc
es
si
b
ili
ty
o
f
ed
u
ca
ti
o
n
A
tt
en
d
an
ce
a
t
p
re
sc
h
o
o
ls
Q
u
al
it
y
o
f
h
ea
lt
h
s
er
v
ic
es
A
cc
es
si
b
ili
ty
o
f
h
ea
lt
h
se
rv
ic
es
In
ab
il
it
y
to
a
cc
es
s
h
ea
lt
h
se
rv
ic
es C
le
an
a
ir
C
le
an
lin
es
s
o
f
th
e
su
rr
o
u
n
d
in
g
ar
ea
S
o
u
rc
e
o
f
d
ri
n
k
in
g
w
at
er
Q
u
al
it
y
o
f
d
ri
n
k
in
g
w
at
er
T
h
e
st
an
d
ar
d
o
f
ac
co
m
m
o
d
at
io
n
(
sq
m
)
F
u
el
f
o
r
h
ea
ti
n
g
S
ew
er
ag
e
(s
an
it
at
io
n
)
A
cc
es
s
to
t
h
e
in
te
rn
et
H
o
u
se
h
o
ld
d
eb
t
U
n
em
p
lo
ym
en
t
20
Figure 3: Percentage contribution of each indicator to the MPI by region
Source: Author’s calculation based on data from the Household Budget Survey, 2021.
Section 4. Recommendations and Next Steps
This section provides some recommendations to consider in the process of development of the
national MPI for Kazakhstan. It first adds some recommendations on the indicators and, in a second
stage, some recommendations on the follow-up process. The end of this section covers the next
technical and policy-focused steps towards the development and finalisation of the MPI for
Kazakhstan.
Recommendations on the indicators
As an overall note, it can be mentioned that the subjective ‘level of satisfaction’ questions are
problematic, and this measure uses a lot of these. This is technically a worrying situation, as educated
elite have different ‘frames of reference’ from the poor. We are aware that Kazakhstan is data
constrained, which introduces some difficulties on the type of space that can be measured. Usually,
the poverty cut-off is used to distinguish false positives, but because subjective satisfaction questions
add up to at least 25% (the current poverty cut-off), non-poor people who often answer that they are
highly dissatisfied either due to personality (pessimistic or introverted) or to their frame of reference,
rather than to an objective situation, will be designated poor. In addition, there is a high likelihood
that trends will be distorted by changes in frames of reference. In this sense, poverty would go up if
objective deprivations went down but frames of reference changed. For all these reasons, a key
recommendation is to replace these subjective indicators with objective indicators.
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Shymkent city
Pavlodar Province
Akmola Province
Atyrau Province
Almaty city
city of Nursultan
Mangistau Province
NKR (North Kazakhstan Province)
WKR (West Kazakhstan Province)
Jambyl Province
EKR (East Kazakhstan Province)
Alma-Ata's Province
Karaganda Province
Kyzylorda Province
Aktobe Province
Kostanay Province
Turkestan Province
Quality of education Accessibility of education
Attendance at preschools Quality of health services
Accessibility of health services Inability to access health services
Clean air Cleanliness of the surrounding area
Source of drinking water Quality of drinking water
The standard of accommodation (sqm) Fuel for heating
Sewerage (sanitation) Access to the internet
Household debt Unemployment
21
Regarding each of the indicators, some comments and specific recommendations are listed here
below:
1. EDUCATION DIMENSION
Quality and accessibility of education: Whose satisfaction is it? Who answers the
question? Do poorer or less educated or rural people have different ‘satisfaction’ levels from
urban elite? Is the difference between the percentage of population who is deprived in these
indicators (uncensored headcount ratios – last column of Table 1 –) and the percentage of
population who is poor and deprived in these indicators (censored headcount ratios – Figure
2) high? It is important to understand this and to analyse to what extent the captured
information is affected by ‘adaptive preferences’, especially as these indicators have the same
weights as the ‘objective’ indicators (1/12).
Note: is there 100% attendance of school-aged children? Are all adults educated or there’s
no data to measure school attendance?
Attendance at preschool – are preschools safe and of good quality, so that they are always
better than family relatives? In occasions, if the preschool is not of good quality or safe for
children, being looked after a relative could be a better option.
General comment: all questions pertain to children – none of them capture information on
adults’ education. What percentage of households lack children, so could not answer any
questions in this dimension? Note that the households that only have adults and no children
are automatically non-deprived in education. Is this something to which the BNS team agrees
on? A possible suggestion could be to name the dimension ‘child education’ instead.
2. HEALTH & ENVIRONMENT DIMENSION
Quality and accessibility of health: same comments as in the same indicators of the
Education dimension above.
Inability to access health services: it is recommended to do an analysis on the different
reasons across regions and groups of population. This would enable to provide a more
detailed information for policy to address the one(s) that are most prevalent in each area.
All satisfaction indicators: same comments as above.
Source of drinking water: the Sustainable Development Goal (SDG) 6 also requires the
water source to be ‘on site’. Does the information in the indicator capture the ‘on site’ source
of drinking water?
3. HOUSING & LIVING CONDITIONS DIMENSION
Standard of accommodation (overcrowding): it has a high weight and a very high
uncensored headcount ratio (47.4% – last column of Table 1), making this deprivation very
visible. Is the deprivation cut-off (the household is deprived if a person lives in less than 15 squared
metres) a national standard? It will be scrutinised.
22
Fuel for heating: the deprivation cut-off identifies a deprivation if the household uses solid
or liquid fuel for heating. It will be important to analyse which is the composition of the
deprivation according to each type of fuel – solid and liquid –, in order to inform policy
more accurately. What is the justification behind considering ‘liquid’ fuel as a deprivation? Is
it kerosene poisoning or fire risk, but LPG does not have those risks? Solid fuel has a health
and climate justification. The inclusion of ‘liquid’ fuel needs to be clearly justified. Is there a
policy aiming at only having electric heating or any of the type?
Sewerage (sanitation): a problem with this indicator is that the uncensored headcount ratio
(percent of population who is deprived in the indicator) is 40.8%; whereas the censored
headcount ratio (percent of population who is poor and deprived in the indicator) is 15.6%.
The incidence of multidimensional poverty (H) is 23.6%. So, 66.1% (15.6/23.6) of the poor
are deprived in sewerage. But 33% ((40.8-15.6)/(100-23.6)) of the non-poor are also deprived
in sewerage. Therefore, this is not discriminating poor from non-poor. A suggestion would
be to look into the deprivations of the poor vs non-poor and see if you can ‘tighten’ the
definition, so it definitely indicates poverty.
4. LIVING STANDARDS/FINANCIAL INCLUSION DIMENSION
Household debt: it is recommended to include a recall period in the indicator definition in
Table 1. That is, “a person 15+ failed to pay rent/mortgage, loan or utility services twice or
more”, during which period of time? As this indicator has a weight of 1/8 – which is the
highest weight among all the indicators13 – and has a high censored headcount ratio (13.0%
– see Figure 2), its contribution will be very visible. Therefore, it will come up as an important
indicator and, for this reason, it needs a good justification.
Unemployment: it’s not clear from the definition if the persons are ‘not searching for a job’.
Are housewives considered deprived, or retired people or people living with disabilities? It
is recommended to clarify this information in the description of the indicator in Table 1, in
order to have a complete information on who could be identified as deprived or not.
Recommendations on the follow-up process
As part of the process of developing a national MPI for Kazakhstan, it would be recommended that:
Two committees be created:
o Technical Committee: composed of the BNS team and any other institution
supporting the BNS team on the technical process of development of the national
MPI
o Steering Committee: composed of the policymakers and main stakeholders that
will use the measure for policy and/or can provide relevant feedback on its structure,
according to their expertise of the national context.
The Technical Committee reviews the indicators according to the technical
recommendations mentioned above, to the extent of data availability.
13 Because all the dimensions are weighted equally and all indicators are weighted equally within each dimension
(nested weights), and the dimension of Living Standards/financial inclusion has only two indicators, this makes
that these two indicators have the highest weights of all the indicators.
23
The Technical Committee produces updated results at the national level and conducts a
disaggregation analysis by different sub-groups of population (regions, area, age cohorts, …)
that are relevant for the context of Kazakhstan and for which data is representative.
The Technical Committee presents these updated results to the Steering Committee
The Technical Committee reviews the measure according to the feedback of the Steering
Committee.
Once the measure is agreed, the Technical Committee conducts detailed analysis on
disaggregation, robustness, redundancy and multidimensional versus monetary poverty (if
data is available) 14 and writes an MPI report that could contain: a description of the process
of development of the national MPI; a short description of the method used; a description
of the data; the structure and its components (indicators, dimensions, weights, deprivation
cut-offs and poverty cut-off), adding the normative justifications behind each elements; the
main results; and policy implications.
Once the report is finalised, it is recommended to organise an official launch, at the national
and international spheres, of the MPI as an official and permanent statistic to measure
multidimensional poverty in Kazakhstan and guide policy action towards its eradication.
Technical next steps
It is recommended that the Technical Committee completes the Massive Open Online
Course (MOOC) on “Designing a Multidimensional Poverty Index (2022)”, prepared by
UNDP and OPHI, which is now available in Russian and in a self-paced version under the
following hyperlink: https://www.learningfornature.org/ru/courses/designing-a-
multidimensional-poverty-index-2022-2/
OPHI conducts an online session to cover redundancy, robustness, and changes over time
analysis: due to the limited time available these topics were not covered during the in-person
training workshop and may need to be covered online;
OPHI conducts an online session to review the interpretation of the results of the MPI for
Kazakhstan and cover the communication strategies for dissemination of the MPI concepts
and results to wide audiences;
The Technical Committee needs to complete the syntax for the deprivation matrix, update
the aggregation, dimensional breakdown, and disaggregation syntax according to their
indicators and sampling variables, and perform missing values analysis;
The Technical Committee updates the measure following feedback from the Steering
Committee;
14 If time and data permits, it could be possible to also perform an analysis on the evolution of multidimensional
poverty in Kazakhstan over time. This would require specific considerations in data construction and ensure
that a) the data source has the same sampling design; b) the indicators of the measure are harmonised
throughout each year of the survey – i.e., they are defined in the same way in all the years of the survey. This
ensures comparability of results over time.
24
Upon agreement on the final structure with the Steering Committee, a detailed analysis of
the MPI results at the national level, decomposition of poverty by indicators and
disaggregation analysis by sub-groups of population, would be performed by the Technical
Committee. An analysis on the evolution through time (changes over time), and an analysis
on the complementarity of both the MPI and the national income measure, could also be
performed.
Robustness and redundancy tests: robustness tests using the current structure for different
weights and poverty cut-offs are needed, as well as a redundancy analysis to show that each
indicator adds new and relevant information to the whole MPI structure;
Writing of the MPI report to include: the detailed process of developing the MPI for
Kazakhstan, the structure, and the normative decisions around which it is built (summary of
the information included in the policy table), the detailed final results, policy implications of
the results and conclusions.
Policy next steps
The BNS will send the policy table and work to the policymakers in other ministries for their
more detailed feedback on how the structure of the MPI can be improved.
Coordinate another meeting with the policymakers to present the results of the updated
measure.
Upon agreement on the final structure, final detailed analysis and finalisation of the report,
it is recommended to organise an official launch, at the national and international spheres,
of the MPI as an official and permanent statistic to measure multidimensional poverty in
Kazakhstan and guide policy action towards its eradication.
Conclusions
This report has provided a summary of the main activities that took place during the Technical
Workshop on Multidimensional Poverty in Nur-Sultan in the beginning of July 2022, as well as the
main challenges and learnings that were drawn from this experience. A succinct but detailed
presentation of the Alkire-Foster method and its practical application through precise steps, is
covered in the second section of this report. The structure of the pilot national MPI for Kazakhstan,
together with the preliminary results at the national and subnational level are also detailed in the third
section. And the report ends with a set of actionable recommendations and next steps towards the
development and finalisation of the MPI for Kazakhstan.
The main outcome of this project is the learning and knowledge that the BNS team has gained in
terms of computation, analysis, and justification, both from a technical and policy point of view, of
the structure of the pilot national MPI for Kazakhstan and its importance as an official public policy
statistic. The intensive work that the entire BNS team has carried out during the week-long training
workshop, has contributed to a very substantial progress in the process of developing the national
MPI for Kazakhstan. The different angles of progress range from the technical and interpretative
understanding of the Alkire-Foster method, used worldwide to measure multidimensional poverty;
to the detailed study of the rationale behind the policy decisions behind each of the indicators
considered in the structure of the national MPI; to working with the data to calculate the indicators
and obtain the preliminary results of the national MPI; to its presentation and communication to a
large group of policy makers from different ministries and governmental institutions.
25
A key recommendation that emerges from this project is the need to reconsider the indicators that
measure the level of satisfaction and replace them with more objective indicators that more accurately
measure the deprivations that poor people in Kazakhstan experience. A number of steps, both
technical and policy, still need to be carried out as part of the process of developing and finalising
the Kazakhstan MPI. Nonetheless, this report enables to already underline that the strong working
capacity that characterises the BNS team, and the knowledge acquired during the implementation of
this project, will allow the team to continue the development of the MPI towards its
institutionalisation as an official and permanent public policy statistic in Kazakhstan.
26
Appendix
Agenda of the Technical Workshop
Oxford Poverty and Human Development
Initiative
http://www.ophi.org.uk | [email protected]
Oxford Dept of International Development,
Queen Elizabeth House, University of Oxford
Technical Workshop
« Towards a Multidimensional Poverty Index (MPI) for Kazakhstan »
4th July – 8th July, 2022
Aim The aim of this workshop is to provide a conceptual and technical introduction to
multidimensional poverty measurement with a strong emphasis on the Alkire-Foster
(AF) method. The discussion will revolve around the implementation and use of
multidimensional measures for policy purposes. By the end of the workshop
participants will also have explored HBS data and generated some optional
indicators and structures for the MPI in Kazakhstan.
Audience The course will target statisticians and technical experts from Kazakhstan, including
participants from the Kazakhstan Statistics Office and other relevant organizations.
Participants must have previous knowledge of Stata or SPSS. They should have
access to computers with the software already installed (Stata or SPSS will not be
provided).
Objectives At the end of the workshop, participants will:
1) Understand why and how multidimensional poverty measures add value to
previous poverty approaches and can be used for informing policymaking.
2) Understand how to design, compute, and analyse a Multidimensional Poverty
Index (MPI), using the Alkire-Foster method.
3) Understand some possible uses of MPIs for policy and learn about the
opportunity to engage with the Multidimensional Poverty Peer Network
(MPPN) to participate in regular sessions on methodological aspects and use of
MPIs in other countries.
Format This training will be delivered in person in Nur-Sultan, Kazakhstan at the National
Bureau of Statistics Office. The language of teaching is Russian. Hence the sessions
will be divided in two parts: 1) participants will be shown videos in class with either
Russian subtitles or simultaneous Russian translation, followed by a Q&A session
with the (English-speaking) instructor with simultaneous or consecutive translation;
2) practical sessions organised in working groups will be developed with
simultaneous or consecutive English-Russian translation
Facilitators Dr. Juliana Milovich
Researcher | Oxford Poverty & Human Development Initiative (OPHI),
University of Oxford
27
Monday 4th July 2022
Lecture 1:
The Alkire-Foster
method
(1 hour)
9 :00 – 10 :00
• 9:00-9:30 Projection of videos (30min)
o “The Alkire-Foster method”, by James Foster (14m)
o “Interpretation of MPI”, by Sabina Alkire (14m)
• 9:30-9:35 Presentation of highlights to discuss (5min)
• 9:35-10:00 Discussion & notes to exchange questions and answers on Day 2 (25min)
Highlights to discuss
during the lecture
✓ What are the steps to calculate an MPI?
✓ How do we interpret the MPI? and the incidence? and the intensity?
✓ What is the difference between uncensored and censored headcount ratios?
Working groups
session (1.5 hours)
10:00 – 11:30
• 10:00-11:30 Exercise on paper – AF method (90min): the participants (divided into working
groups of 3 people max.) will work on Exercise 1 of the AF method
11:30-11:45 Tea break
Lecture 2:
Data & Indicators
(1.25 hours)
11:45 – 13:00
• 11:45-12:05 Projection of videos (20min):
o “Dimensions and Indicators”, by Jakob Dirksen (4min)
o “Indicator issues”, by Usha Kanagaratnam (12min)
o “Missing value”, by Rizwan ul Haq (4min)
• 12:05-12:10 Presentation of highlights to discuss (5min)
• 12:10-13:00 Discussion & notes to exchange questions & answers on Day 2 (50min)
Highlights to discuss
during the lecture
✓ What are in theory the most important dimensions and indicators for measuring
multidimensional poverty in Kazakhstan? Which is their reference population?
✓ Which is the justification to consider them? How are they related to policy documents? And
how can policy impact them?
13:00 – 14:00 Lunch break
Discussion session
14:00 – 17:00
(3 hours)
• 14:00-17:00: Discussion around the relevant indicators for the Pilot MPI of Kazakhstan and the
indicators that can be calculated with the HBS data (180min)
Goals for the day:
1. Learn the Alkire-Foster method and its steps to measure multidimensional poverty
2. Identify the relevant indicators for the MPI of Kazakhstan
3. Identify indicators that could be calculated with the data
Tuesday 5th July 2022
Practical session
(1.5 hours)
9.30 – 11:00
• 9:30-9:45 Introduction of the participants and presentation of the agenda (15min)
• 9:45-10:10 From Lecture 1: Review of the Alkire-Foster method to ensure that everyone
understands it before entering into the computation of the deprivation matrix (25min)
• 10:10-10:35 From Lecture 1: Discussion, questions, and answers (25min)
• 10:35-11:00 From Lecture 2: Discussion, questions, and answers (25min)
11:00 – 11:15 Tea break
Practical
session
(1.75 hours)
11:15 – 13:00
• 11:15-12:45 Presentation of the main steps and codes in Stata to build two examples of
indicators of the deprivation matrix using the file ‘dofile_0_dataprep_VF2-KAZ2022’
(90min)
• 12:45-13:00 Discussion, questions and answers (15min)
13:00 – 15:00 Lunch break
Working group session
(2 hours)
15 :00 – 17:00 hours
• 15:00-17:00 Beginning of group work (120min): participants are divided in two groups:
o Less technical participants: focus on policy-related activities, completing the sheet
called “Session 1 – Setting out” of the ‘My MPI tracker’ file with the relevant
indicators. Writing down:
a. Indicators, their deprivation cut-offs, and applicable population
b. Justification of the relevant indicators – link to National policy
documents
c. Linking indicators and deprivation cut-offs to SDGs and existing
policy priorities
d. Identifying policy actors to be engaged for each specific indicator
o More technical participants (computer centre): focus on preparing the HBS data
to start building the deprivation matrix in SPSS
28
Tasks during the session:
✓ Start completing the first tab “Session 1 – Setting out” of the excel file “My MPI tracker”
✓ Prepare the data of the HBS survey to start building the indicators of the deprivation matrix
Goals for the day:
1. Make sure all the participants have learned the Alkire-Foster method and its steps to measure multidimensional poverty
2. Learn the key consideration required to compute the indicators
3. Link the relevant indicators for the MPI of Kazakhstan with policy documents, priorities, SDGs and policy institutions
4. Define the deprivation cut-offs of the indicators and the applicable population
5. Prepare the data of the HBS survey to start building the indicators of the deprivation matrix
Wednesday 6th July 2022
Working group session
(1.5 hours)
9:30 – 11:00 hours
• 9:30-11:00 Beginning of group work (90min): participants are divided in two groups:
o Less technical participants: focus on continuing completing the sheet called “Session
1 – Setting out” of the ‘My MPI tracker’ file with the relevant indicators. Writing down:
a. Indicators, their deprivation cut-offs, and applicable population
b. Justification of the relevant indicators – link to National policy
documents
c. Linking indicators and deprivation cut-offs to existing policy
priorities
d. Identifying policy actors to be engaged for each specific indicator
o More technical participants (computer centre): focus on translating the variables, from
Russian to English, in the dataset that need to be used to build the indicators for the
MPI
Tasks during the session:
✓ Continue filling in the first tab “Session 1 – Setting out” of the excel file “My MPI tracker”
✓ Translate the variables, from Russian to English, of the different datasets to build the indicators
for the MPI
11:00 – 11:15 Tea break
Working group session
(1.75 hours)
11:15 – 13:00
• 11:15-13:00 Continue working on the previous activities (115min)
Tasks during the session:
✓ Continue filling in the first tab “Session 1 – Setting out” of the excel file “My MPI tracker”
✓ Finalize translation of the variables, from Russian to English, of the different datasets to build
the indicators for the MPI
13:00 – 14:00 Lunch break
Working group session
(3 hours)
14:00 – 17:00
• 14:00-17:00 Continuity of working groups (180min): participants are divided in two groups:
o Less technical participants: focus on the previous policy activities
o More technical participants (computer centre): focus on merging the different datasets
and produce the final database to start building the indicators of the deprivation matrix
Tasks during the session:
✓ Fill in the first tab “Session 1 – Setting out” of the excel file “My MPI tracker”
✓ Produce a final single database of the HBS survey, in Russian and also translated from Russian
to English
Goals for the day:
1. Link the relevant indicators for the MPI of Kazakhstan with policy documents, priorities, SDGs, and policy institutions
2. Produce a final single database of the HBS survey, in Russian and also translated from Russian to English, that will be
used to start building the indicators of the deprivation matrix
29
Thursday 7th July 2022
Practical
session (1.5 hour)
9:30 – 11:00
• 9:30-10:45 Presentation of main steps and codes in Stata on aggregation (75min)
• 10:45:-11:00 Questions and answers (15min)
11:00 – 11:15 Tea break
Practical
session (1.75 hour)
11:15 – 13:00
• 11:15-12:45 Presentation of main steps and codes on dimensional breakdown and
disaggregation analysis (90min)
• 12:45-13:00 Questions and answers (15min)
13:00 – 14:00 Lunch break
Working group
session
(3 hours)
14:00 – 17:00
• 14:00-17:00 Continuity of group work (180min): participants are divided in two groups:
o Less technical participants: focus on
a. finalizing the policy-related activity on the indicators, completing the
sheet called “Session 1 – Setting out” of the ‘My MPI tracker’ file with
the relevant indicators.
b. building the presentation for the policymakers meeting of Friday
8th July
o More technical participants (computer centre): focus on coding the indicators
to build the deprivation matrix using the HBS data from Kazakhstan –
identify the decisions about deprivation cut-offs, applicable population, missing
values, etc., (and how to translate them from Stata to SPSS)
Tasks during the session:
✓ Finalize building the policy table on the indicators and their justifications
✓ Build the presentation for the meeting with the policymakers of Friday 8th July
✓ Create the deprivation matrix in SPSS and Stata
Goals for the day:
1. Learn the steps to estimate the incidence and the intensity of multidimensional poverty and the MPI, and learn how to
interpret the results
2. Learn the steps to decompose poverty levels by indicators and disaggregate results by regions, and learn how to interpret
the results
3. Finalize the computation of the deprivation matrix in SPSS and in Stata
Friday 8h July 2022
Working group
session
(1.5 hours)
9:30 – 11:00
9:30-11:00 The BNS team finalises the presentation for the meeting with the policymakers,
incorporating the preliminary results of the pilot MPI for Kazakhstan (the incidence and intensity of
multidimensional poverty, the MPI, its decomposition by indicators and the contribution of each
indicator to overall poverty in each region).
Meeting with
stakeholders
(2.5 hours)
11:00 – 13:30
✓ OPHI does a presentation on multidimensional poverty, policy uses and country examples
✓ BNS does a presentation on the process of development of the MPI for Kazakhstan and its
preliminary results
✓ Discussion to receive feedback from stakeholders
✓ UNICEF presents analysis on child multidimensional poverty measurement
✓ Final discussion
✓ Closing remarks
13:30 – 14:30 Lunch break
Closing the workshop:
Wrap-up meeting
(1.5 hour)
14:30 – 16:00
• Discussing impressions from the meeting and any further doubts
• Agreeing on next steps
• Closing the week of work
Goals for the day:
1. Present the preliminary results of the MPI for Kazakhstan to key stakeholders and receive their feedback
2. Discuss impressions from the meeting
3. Agree on next steps
30
List of institutions present during the policy meeting of Friday 8th July 2022
Agenda of the meeting with policymakers on Friday 8th July 2022 (in Russian)
11.00-11.15 Приветственное слово. Белоносова Н. Е. -Директор департамента статистики
труда и уровня жизни БНС
Цели и задачи встречи
Знакомство
№ Full name Position and organization
Foreign experts, consultants:
Dr. Juliana Milovich Oxford Initiative for Poverty Reduction and Human Development
(OPHI), University of Oxford, researcher
Rafkat Hasanov Independent Consultant of the UN Economic Commission for Europe
Department of Labor Statistics and Living Standards
Bureau of National Statistics
Natalia Belonossova Director of the Department of Labor Statistics and Living Standards
Daurenbekova Gulzhan
Kulgazievna
Deputy Director of the Department of Labor Statistics and Living
Standards
Department of Living Standards Statistics
Amerzhanova Marzhan
Yerzhanovna
Head of the Department
Kereybayeva Samal Baizakovna Chief expert of the Department
Makshaeva Aizhan Sovetovna Chief expert of the Department
Office of Household Survey Statistics
Malikova Dana Erkenovna Chief expert of the Department
Zhumanova Nagima Askarkyzy Chief expert of the Department
1. Kasenova Rymzhan Beibitovna Expert of the Department
Information and Computing Center of BNS
2. Dilmagambetova Gulnar
Seipenovna
Chief Specialist of the Department of Information Support of
Household Statistics
3.
Kyndybai Laura
Chief Specialist of the Department of Information Support of
Household Statistics
Analytical Center of the Information and Computing Center of BNS
4. Bayzakov Nauryz Aybarovich Econometrician-analyst
5. Khamitzhan Abylaykhan
Aitbayuly
Data Analyst
6. Kerembayev Alpamys Aidarovich Business Analyst
7. Kerembayev Anuar Tolegenovich
Керембаев
Project Manager
Ministry of Labor and Social Protection of the Population of the Republic of Kazakhstan
8. Zhabagina Galiya Myrzabekovna Deputy Director of the Department of Social Assistance
9. Kurmankulova Asiya
Kadyrnyazovna
Head of the Department of Targeted Social Assistance
Institute of Economic Research" of the Ministry of National Economy of the Republic of Kazakhstan
10. Toktasyn Bakbergen
Bakytzhanuly
Analyst of the Center for Macroanalytics and Forecasting
Agency for Strategic Planning and Reforms of the Republic of Kazakhstan
11. Pernebayeva Zhuldyz Usenovna Director of the Department of Social Sphere
12. Maratkyzy Ulzhan Chief Expert of the Department of Social Sphere
31
11.15-11.35 Международная практика измерения многомерной бедности. Сабина
Алькаир,
Директор Оксфордской инициативы по борьбе с бедностью и человеческому
развитию
Джулиана Милович, исследователь Оксфордской инициативы по борьбе с
бедностью и человеческому развитию
11.35-12.10 Подходы для разработки пилотного национального индекса многомерной
бедности. Маржан Амержанова, руководитель управления статистики уровня
жизни
Самал Керейбаева, главный эксперт управления статистики уровня жизни
12.10-12.15 Комментарии к пилотному проекту ИМБ. Рафкат Хасанов, консультант ЕЭК
ООН
12.15-13.00 Обсуждение. ЮНИСЕФ, Государственные органы
Policy table: Justification for the Indicators of the Pilot National Multidimensional Poverty Index
Indicator Rationale
What are the SDG
targets related to
measurement?
What SDG indicators allow
tracking this indicator?
What government agencies/programs
are working on this indicator? And
how can they influence it?
What political actors should
be involved before the launch
of the MPI?
Education
4.1 By 2030, ensure that
all girls and boys
complete free, equitable
and quality primary and
secondary education
leading to relevant and
effective learning
outcomes
4.2 By 2030, ensure that
all girls and boys have
access to quality early
childhood development,
care and pre-primary
education so that they
are ready for primary
education
Level of satisfaction with the
quality of education
Strategy 2050: Section 4. Knowledge and
professional skills are key landmarks of the
modern education, training and retraining system
NDP 2025: National Priority 3: Quality
Education, Objective 3. Improving the quality of
education.
National project Quality Education
"Educated Nation": Objective 1. Ensuring the
availability and quality of preschool education and
training.
Objective 2. Improving the quality of secondary
education: reducing the gap in the quality of
education between regions, urban and rural
schools in Kazakhstan
NDP 2025 and the National project: both
documents have the same strategic indicator - the
level of satisfaction of the population with the
quality of preschool / secondary education.
4.1.1 Proportion of children and
young people (a) in grades 2/3; (b) at
the end of primary; and (c) at the end
of lower secondary achieving at least a
minimum proficiency level in ( i )
reading and (ii) mathematics, by sex
16.6.2 Proportion of population
satisfied with their last experience of
public services
MOES: teacher training, improvement of
education infrastructure
ASPiR, MOES, MNE,
UNICEF, charitable
foundations and associations in
the education sector
Level of satisfaction with the
affordability of education
Strategy 2050: Section 4. Knowledge and
professional skills are key landmarks of the
modern education, training and retraining system
NDP 2025: National priority 3: Quality education,
Objective 1: Ensuring access and equality in
education.
National project Quality education "Educated
nation": Objective 1. Ensuring the availability and
quality of preschool education and training.
Objective 3. Providing schools with a comfortable,
safe and modern educational environment.
Several outcomes, including:
- Coverage of children with additional education,
- Coverage of children with special developmental
disabilities by psychological and pedagogical
support and early correction
4.1.1 Proportion of children and
young people (a) in grades 2/3; (b) at
the end of primary; and (c) at the end
of lower secondary achieving at least a
minimum proficiency level in ( i )
reading and (ii) mathematics, by sex
16.6.2 Proportion of population
satisfied with their last experience of
public services
MOES: building schools and other
educational institutions on the principles
of inclusiveness
ASPiR, МОES, MNE,
UNICEF, charitable
foundations and associations in
the field of education
33
Preschool attendance Strategy 2050: Priorities in education: (1)
Kazakhstan needs to switch to new methods of
preschool education.
NDP 2025: Objective 1. Ensuring access and
equity in education. The physical availability of
places in preschool institutions will be ensured at
the rate of a potential 100% enrollment of
children.
National project Quality education "Educated
nation": Objective 1: Ensuring the availability of
quality and pre-school education.
Model of preschool education and unbringing:
chapter 2, para 7 The state policy in the system of
preschool education is aimed at ensuring
accessibility.
4.2.2 Participation rate in organized
learning (one year before the official
primary entry age ), by sex
4.2.2.1 Readiness for school (
percentage of children attending the
first grade of primary school who
attended a pre-school in the previous
year )
МОES: financial support for households
and construction of preschool institutions,
improvement of the registration system
for children, training of teachers,
improvement of the infrastructure of
preschool education
ASPiR, МОES, MNE,
UNICEF, charitable
foundations and associations in
the field of education
Health and environment
3.8 Achieve universal
health coverage,
including financial risk
protection, access to
quality essential health-
care services and access
to safe , effective ,
quality and affordable
essential medicines and
vaccines for all
Level of satisfaction with the
quality of health care services
Strategy 2050: section 3. New principles of social
policy – social guarantees and personal
responsibility. One of the key priorities in
healthcare: to provide affordable medical services
at high standards of care.
NDP 2025. Priority 2. Accessible and effective
healthcare system. Objective 2. Increasing the
availability and quality of medical services.
Strategic target indicator - Level of satisfaction of
the population with the quality and accessibility of
medical services provided by medical institutions
National project "Quality and affordable
healthcare for every citizen "Healthy Nation":
Priority 1. Affordable and high-quality medical
care. Objective 1. Ensuring wide coverage of the
population with health services. Strategic indicator
4 - Level of satisfaction of the population with the
quality and accessibility of medical services
provided by medical institutions
3.8.1 Coverage of essential health
services
MH: training doctors, improving
healthcare infrastructure
MH, ASPiR, WHO,
Foundations and associations in
the healthcare sector
34
level of satisfaction with the
availability of health services
Strategy 2050: section 3. New principles of social
policy – social guarantees and personal
responsibility. One of the key priorities in
healthcare: to provide affordable medical services
at high standards of care.
NDP 2025. Priority 2. Accessible and effective
healthcare system. Objective 2. Increasing the
availability and quality of medical services.
Strategic target indicator - Level of satisfaction of
the population with the quality and accessibility of
medical services provided by medical institutions.
National project "Quality and affordable
healthcare for every citizen "Healthy Nation":
Priority 1. Affordable and high-quality medical
care. Objective 1. Ensuring wide coverage of the
population with health services Strategic indicator
4 - the level of satisfaction of the population with
the quality and accessibility of medical services
provided by medical institutions
11.6 By 2030, reduce the
adverse per capita
environmental impact of
cities , including by
paying special attention
to air quality and
municipal and other
waste management
3.8.1 Coverage of essential health
services
MH: construction of hospitals and other
health care facilities
MH, ASPiR, WHO,
Foundations and associations in
the healthcare sector
Inability to access health care
services
Strategy 2050: section 3. New principles of social
policy – social guarantees and personal
responsibility. One of the key priorities in
healthcare: to provide affordable medical services
at high standards of care.
NDP 2025. Priority 2. Accessible and effective
healthcare system. Objective 2. Increasing the
availability and quality of medical services.
Strategic target indicator - Level of satisfaction of
the population with the quality and accessibility of
medical services provided by medical institutions.
National project "Quality and affordable
healthcare for every citizen "Healthy Nation":
Priority 1. Affordable and high-quality medical
care. Objective 1. Ensuring wide coverage of the
population with health services Strategic indicator
4 - the level of satisfaction of the population with
the quality and accessibility of medical services
provided by medical institutions
3.8.1 Coverage of essential health
services
MH: construction of hospitals and other
health facilities, training of doctors,
improvement of health infrastructures
MH, ASPiR, WHO,
Foundations and associations in
the healthcare sector
35
Cleanliness of the territory
adjacent to housing (absence of
household garbage (waste))
Concept for the transition of the Republic of
Kazakhstan to a "green economy" for 2021-
2030: section 3.5. Waste management system.
Action plan for the implementation of the
Concept for the transition of the Republic of
Kazakhstan to a "green economy" for 2021-
2030.: section 6.1, paragraphs 38-40
11.6.1 Proportion of municipal solid
waste collected and managed in
controlled facilities out of total
municipal waste generated, by cities
MEGNR: implementation of programs
for the recycling of MSW
MEGNR
Air purity Concept for the transition of the Republic of
Kazakhstan to a "green economy" for 2021-
2030: Section 3.6. Reduced air pollution.
Action plan for the implementation of the
Concept for the transition of the Republic of
Kazakhstan to a "green economy" for 2021-
2030: section 5, paragraphs 35-37
3.9.1 Morality rate attributed to
household and ambient air pollution
9.4.1.1 CO2 emissions per capita
11.6.2 Annual mean levels of fine
particulate matter (eg PM2.5 and
PM10) in cities (population weighted)
13.2.2 Total greenhouse gas emissions
per year
MEGPR RK: implementation of
programs to reduce emissions in the main
economic sectors of Kazakhstan
MEGNR
Drinking water quality The State Program for Housing and
Communal Development "Nurly Zher":
objective 2 "Rational provision of the population
with high-quality drinking water and sanitation
services"
The action plan for the implementation of the
Concept for the transition of the Republic of
Kazakhstan to a "green economy" for 2021-
2030 also contains section 1.1.1 Provide water to
the population including several activities (1-3)
3.9.2 Mortality rate attributed to
unsafe water, unsafe sanitation and
lack of hygiene (exposure to unsafe
Water, Sanitation and Hygiene for All
(WASH) services)
6.1.1 Proportion of population using
safely managed drinking water
services
CCHPU: construction of drinking water
sources, improvement of water supply,
reduction of depreciation of water supply
Committee for Construction
and Housing and Public
Utilities of the Ministry of
Investment and Development
of the Republic of Kazakhstan
Housing and living
conditions
11.1 By 2030, ensure
access for all to
adequate, safe and
affordable housing and
basic services and
upgrade slums
6.1 By 2030, achieve
universal and equitable
access to safe and
Non-compliance with living
standards
NDP 2025: National Priority 1. Equitable social
policy. Objective 2. Ensuring social well-being.
Effective social support will be provided to
address housing issues.
The State Program for Housing and
Communal Development "Nurly Zher" for
2020-2025: The goal of the program is to increase
the availability and comfort of housing and
develop housing infrastructure: Objective 1.
Implementation of a unified housing policy,
Objective 3. Modernization and development of
the housing and communal sector. Target
indicator - by 2025 to ensure 26 sq. m per one
household member
11.1.1 Proportion of urban
population living in slums, informal
settlements or inadequate housing
11.3.1 Ratio of land consumption rate
to population growth rate
CCHPU: implementation of housing
programs
Committee for Construction
and Housing and Public
Utilities of the Ministry of
Investment and Development
of the Republic of Kazakhstan
36
Ability to maintain heat at a
sufficient level
State Program for Housing and Communal
Development "Nurly Zher for 2020-2025":
Section 5.3.1 Modernization of the heat supply
sector focuses on the improvement of the tariffs.
Objective 3. Modernization and development of
the housing and communal sector: Activities 18-
26. Outcome indicator 1 - Wear and tear of
heating networks.
Law "On Housing Relations"and, in particular,
Article 97. Payment for housing from the State
Housing Fund and housing rented by a local
executive body as a private house, and the
provision of housing assistance to low-income
families (citizens)
affordable drinking
water for all
6.2 By 2030, achieve
access to adequate and
equitable sanitation and
hygiene for all and end
open defecation, paying
special attention to the
needs of women and
girls and those in
vulnerable situations
6.3 By 2030, improve
water quality by reducing
pollution, elimination
dumping and
minimizing release of
hazardous chemicals and
materials, halving the
proportion of untreated
wastewater and
substantially increasing
recycling and safe reuse
globally
6.4 By 2030,
substantially increase
water-use efficiency
across all sectors and
ensure sustainable
withdrawals and supply
of freshwater to address
water scarcity and
substantially reduce the
number of people
11.1.1 Proportion of urban
population living in slums, informal
settlements, or inadequate housing
CCHPU: changing the structure of
consumption, improving equipment for
heating
Committee for Construction
and Housing and Public
Utilities of the Ministry of
Investment and Development
of the Republic of Kazakhstan
Source of drinking water The State Program "Nurly Zher": Objective 2
"Rational provision of the population with high-
quality drinking water and sanitation services"
The action plan for the implementation of the
Concept for the transition of the Republic of
Kazakhstan to a "green economy" for 2021-
2030 also contains section 1.1.1 Provide water to
the population including several activities (para 1-
3)
6.1.1 Proportion of population using
safely managed drinking water
services
3.9.2 Mortality rate attributed to
unsafe water, unsafe sanitation and
lack of hygiene (exposure to unsafe
Water, Sanitation and Hygiene for All
(WASH) services)
CCHPU: construction of drinking water
sources, improvement of water supply,
reduction of depreciation of water supply,
Committee for Construction
and Housing and Public
Utilities of the Ministry of
Investment and Development
of the Republic of Kazakhstan
Sewer access The State Program "Nurly Zher": objective 2
"Rational provision of the population with high-
quality drinking water and sanitation services",
para
3. Coverage of the population with wastewater
treatment in cities.
Activities 15 and 17 of the program.
6.2.1 Proportion of population using
(a) safely managed sanitation services
and (b) a hand-washing facility with
soap and water
3.9.2 Mortality rate attributed to
unsafe water, unsafe sanitation and
lack of hygiene (exposure to unsafe
Water, Sanitation and Hygiene for All
(WASH) services)
CCHPU: construction, improvement of
infrastructure, reduction of depreciation
of water disposal and sewerage
Committee for Construction
and Housing and Public
Utilities of the Ministry of
Investment and Development
of the Republic of Kazakhstan
37
Personal internet access NDP 2025: National Priority 8. Building a
diversified and innovative economy.
Objective 10. Development of infrastructure and
digitalization of basic sectors of the economy: To
reduce the digital inequality, work will be carried
out to improve the quality of the Internet in all
settlements with a population of 250 people and
more. Considering urbanization and economic
feasibility issues, the remaining villages with a
population of less than 250 people will be
connected to the Internet. 100% of regional
centers and cities of republican significance will be
covered by high-speed 5G Internet.
National project "Technological breakthrough
through digitalization, science and
innovation". Priority 6. Internet quality and
information security. Objective 1. Providing 100%
of citizens with high-quality Internet.
Development of broadband networks, transition
of technology to 3G-4G. Improvement of IT
technologies. Providing access to hard-to-reach
and sparsely populated areas.
suffering from water
scarcity
17.8.1 Proportion of individuals using
the Internet
17.6.1 Fixed Internet broadband
subscriptions per 100 inhabitants, by
speed6
MDDIAI: Expansion of coverage of
communication networks, increase of
digital literacy of the population.
Ministry of Digital
Development, Innovations and
Aerospace Industry of the
Republic of Kazakhstan
Standards of living
8.5. By 2030, achieve full
and productive
employment and decent
work for all women and
men, including young
people and persons with
disabilities
and strengthen their
integration into value
chains and markets
9.3 Increase the access
of small-scale industrial
and other enterprises, in
particular in developing
countries, to financial
services, including
affordable credit, and
Unemployment Strategy 2050: New principles of social policy.
Modernization of the labor policy:
Fourthly, we should modernize employment and
salary policy.
NDP 2025: Objective 1. Productive employment.
The strategic target indicator - unemployment rate
Comprehensive plan "Program for increasing
the income of the population until 2025":
Section II. Increasing the income of the
population through the creation of new jobs
contains measures (from 7 to 22) aimed at creating
jobs and reducing unemployment.
The target indicator - unemployment rate
8.5.2 Unemployment rate, by sex, age
and persons with disabilities
MLSPP: Expansion of program funding.
Implementation of state programs in the
field of productive employment and
entrepreneurship.
MLSPP, ILO, ASPiR RK
38
Household debt on payments Comprehensive plan "Program to increase the
income of the population until 2025": Priority
III. Protection of the purchasing power of the
population's income, p. 36. Action: Adoption of
individual plans of banks for working with
problem loans to resolve the problem debts of
citizens on loans.
their integration into
value chains and markets
1.2 By 2030, reduce at
least by half the
proportion of men,
women and children of
all ages living in poverty
in all its dimensions
according to national
definitions
1.2.1 Proportion of population living
below the national poverty line, by
sex and age
ARDFM: development of individual plans
for working with problem loans to resolve
the debts of citizens.
Agency for Regulation and
Development of the Financial
Market
