The Afghanistan Multiple Indicator Cluster Survey (AMICS) was carried out in 2010-2011 by the Central Statistics Organisation (CSO) of the Government of the Islamic Republic of Afghanistan in collaboration with United Nations Children’s Fund (UNICEF). Financial and technical support was provided by UNICEF.
MICS is an international household survey programme developed by UNICEF. The Afghanistan MICS was conducted as part of the fourth global round of MICS surveys (MICS4). MICS provides up-to-date information on the situation of children and women, and measures key indicators to monitor progress towards the Millennium Development Goals (MDGs), the Afghanistan National Development Strategy (ANDS) and other internationally agreed upon commitments.
Download the full report here: AMICS 2010/11 Report (PDF)
AMICS Fact Sheets
I. Afghanistan and the Global MICS Surveys
II. Comparing the AMICS Data with Other Data
III. The AMICS Findings
IV. The AMICS Methodology
V. Producing the AMICS
I. Afghanistan and the Global MICS Surveys
The Afghanistan Multiple Indicator Cluster Survey (AMICS) is a nationally representative sample survey that presents data on the social, health, and educational status of women and children in Afghanistan. It was conducted in 2010-2011 by the Central Statistics Organisation (CSO) of the Government of the Islamic Republic of Afghanistan, with the technical and financial support of UNICEF. The survey is based on the need to monitor progress towards goals and targets emanating from recent international agreements such as the Millennium Declaration and the Plan of Action of A World Fit For Children. It further helps track progress towards the Afghan Government’s policy commitments to reduce poverty and support the wellbeing of women and children, such as the commitments made through the Afghanistan National Development Strategy (ANDS).
The purposes of the MICS surveys are to provide up-to-date information for assessing the situation of children and women in Afghanistan, providing evidence for future action and programming design, and to inform relevant policies and interventions. MICS surveys can also contribute to the improvement of data and monitoring systems in countries where they are conducted, strengthening technical expertise in the design, implementation, and analysis of such systems.
Childinfo.org is a website that serves as an information resource about MICS surveys. It contains information about the methodology of MICS surveys, training in MICS, and it also hosts a collection of MICS surveys from countries around the world. The AMICS 2010/11 also explains more about MICS in Chapter 1: Introduction.
No. This is the third MICS survey conducted in Afghanistan. The first was in 1997, and the second was in 2003. However, the MICS methodology has been refined since then, and the Afghan Government’s capacity to carry out large, complex surveys has greatly improved over the last decade.
The two surveys are comparable. The methodology used was the same in each case. However, the previous MICS 2003 survey in Afghanistan could not reach all regions of the country, unlike the AMICS 2010/11, which collected data in every province of the country.
MICS is a global survey that seeks to generate data that is comparable across countries, as well as to measure progress over time within countries. However, each country’s situation is unique and the survey methodology may be adjusted to reflect local conditions and norms. Modest changes to the methodology can help generate findings that are relevant to local contexts. These may include the addition or subtraction of indicators, as appropriate, the removal or addition of questions or slight changes to the ways that questions are asked by surveyors, to help ensure that the questions are clearly understood by respondents. In the AMICS 2010/11, any changes to the standard MICS survey are clearly indicated. An example of a change made to the MICS survey applied in Afghanistan is that a subsample was administered to test blood in some households for anaemia, because there was a need to know the extent of anaemia among women and children in the country.
The Pentavalent vaccine is a combination of five vaccines: diphtheria, pertussis, tetanus, hepatitis B and haemophilus influenza. Although the Pentavalent vaccine was introduced in Afghanistan in 2009, there has not yet been any change to reflect this in the vaccination card issued by the Ministry of Public Health. Therefore, interviewers recorded only the DPT vaccination during the field data collection, as accurate information on the coverage of all five vaccines included in the Pentavalent vaccine was not available at the time of the survey.
II. Comparing the AMICS Data with Other Data
The purpose of the AMICS report is to present primary quantitative data and to highlight the survey’s major findings, as well as to point out important inequities evident from the data. The main content of the report are data tables as well as brief descriptions of highlights from the data. However, since the AMICS presents data by a wide range of background characteristics for nearly 80 unique indicators, it is beyond the scope of the AMICS report to undertake any detailed analysis of all of the findings. However, the AMICS findings serve as valuable information for other agencies and for research scholars to undertake their own qualitative analyses, in combination with other information at their disposal (such as knowledge of health service coverage in a region showing low performance on health indicators, for example).
It is beyond the scope of the AMICS report to compare findings from Afghanistan to other countries. However, as a global survey, MICS reports from other countries may be consulted by those interested in comparing Afghanistan’s findings to findings from elsewhere. Please visit www.childinfo.org to download MICS reports for other countries.
Different methodologies and definitions were used for each of these two surveys. However, the findings of both surveys should be considered when determining indicators.
On immunization indicators, the AMICS findings show lower coverage than demonstrated by NRVA’s data. A survey reflects the situation of the time period in which it was conducted. Experts have suggested that the differences in the two surveys may be due to routine immunization coverage being weak. Household surveys mainly measure routine immunization, as opposed to individual immunizations.
For differences in findings between the NRVA and the AMICS on the indicator of the population’s access to improved water supply and to improved sanitary facilities, differences are mainly due to the use of different definitions. The MICS uses UNICEF’s and WHO’s Joint Monitoring criteria while NRVA uses its own national criteria. In order to be comparable, indicators must use shared definitions and methodological procedures. While many indicators are comparable between the NRVA and AMICS, not all are.
Some of the survey methods are also different, such as the sampling frame. The AMS uses the same eight regions as the AMICS; however, the AMS reached 87% of the country, which is somewhat lower than the AMICS’ coverage. In the AMS, approximately one third of the Southern population was not surveyed due to insecurity, thus Southern rural populations were not fully accounted for in the AMS. Another important difference is that the AMS used both direct and indirect methods in surveying child mortality, while the AMICS used only the indirect method.
Together, these two surveys provide a valuable set of up-to-date information on human development in Afghanistan.
III. The AMICS Findings
The AMICS findings should inform the delivery of aid and development in a way that responds to disparities, inequities or inadequate impact from previous efforts. The AMICS findings can assist Afghan Government services, and aid and development programming, to be targeted and effective by being based on evidence for need.
There are high regional disparities shown in the data, with particularly pronounced disparities evident in the Southern region; and in Eastern, Central Highlands and West regions for some indicators. There are also great variances found between rural and urban areas. The single most predictable indicator of status of almost all other indicators is the education level of the mother. Overall, the findings show that some progress is evident, yet several indicators remain at alarming levels and are cause for concern.
The speed of reduction for the Under-5 Mortality Rate (U5MR)in children and infant mortality rate(IMR) in other, much more stable countries in the region is less than 4% per year over the past two decades. Meanwhile, Afghanistan’s experience of violent conflict, minimal social services in many areas, low child immunization coverage, and high child malnutrition levels suggest a slower speed of reduction in comparison to other countries in the region. For this and other possible reasons, it is likely that the U5MR and IMR are under-estimated in the AMICS.
In general, estimation of child mortality is complex, especially in Afghanistan. Any single survey’s data on mortality cannot be reflected as a true value, unless a series of data from different surveys are found to be comparable, and thus validated. Generating accurate estimates of child mortality poses a considerable challenge because of the limited availability of high-quality data from many developing countries. Complete vital registration systems are the preferred source of data on child mortality because they collect information as events occur and they cover the entire population. However, many developing countries lack fully functioning vital registration systems that accurately record all births and deaths. For more information about estimating child mortality, please refer to recent UN publications on the estimation of child mortality, such as “Step-by-step Guide to the Estimation of Child Mortality” New York: United Nations (1990).
Disparities by region were found across indicators, to varying degrees. Some regional disparities are very pronounced. For instance, the West region has the lowest prevalence of appropriately iodized salt. The availability of soap is lowest in the Central Highlands region, which also had other indicators of concern. Better treatment practices of diarreoa are observed in the South. However, overall, most indicators show higher disparities in the South region. The reasons for this are multiple, and would require further analysis beyond the scope of the AMICS to ascertain the causes of these disparities.
This finding may be due to the application of the ILO’s definition, which says that at least one hour of economic activity per week among children aged 5-11 years is considered to count as child labour, while at least 14 hours of economic activity per week for children aged 12-14 years qualifies as child labour.
Measuring the maternal mortality rate is not part of the standard global MICS survey. This is because collecting data on maternal mortality is not suitable for a household sampling survey. The measure used to calculate maternal mortality is the number of maternal deaths per 100,000 live births. A sampling survey is not able to obtain a great enough sample size as required in this measure.
IV. The AMICS Methodology
Several steps were taken to assure data quality. Numerous individuals and agencies with different areas of expertise were involved in the design and execution of the AMICS survey, to help ensure that the survey followed internationally recognized standards for carrying out household surveys. A carefully designed questionnaire was field-tested to ensure effectiveness. Supervisors, editors and interviewers were trained in field data collection process prior to going into the field. The 66 local data collection teams (each consisting of six interviewers, two editors and one supervisor) were supervised throughout the data collection, and completed questionnaires were verified routinely. The CSO undertook other monitoring activities during the data collection period through their provincial offices and national office. To help ensure access to female respondents in the context of Afghanistan, each data collection team included a minimum of three female interviewers (and one female editor).
The sample for the AMICS was designed to provide estimates for a large number of indicators on the situation of children and women at the national level, for urban and rural areas, in eight regions. Originally, 13,468 households were selected for the survey, and 98.5% of these households were surveyed. A stratified two-stage sample design was used, where the primary sampling units are the enumeration areas (EAs), each with an average of about 185 households (and a total of 21,194 EAs). In the second stage, 30 households were selected from each EA. This is a manageable number of households per EA to work with in a survey, given the high costs of transportation, fieldwork logistics, and the need to seek a cost-effective cluster size.
It is common that respondents may sometimes not be able to answer a question included on the questionnaire. This may be due to the passage of time and a respondent not being able to correctly recall information, or a respondent may not wish to answer a question. In the AMICS report, however, there is little data missing. Specifically, for all tables mentioning the background characteristic of mother’s educational level, up to a maximum of seven cases are missing out of 14,872 cases, and for all tables mentioning the background characteristic of household head’s educational level, up to maximum of 11 cases are missingout of 13,116 cases. For this reason, the sums for each educational level do not equal the total number of cases shown in the tables where these background characteristics are shown. This fact is also mentioned in the report (Chapter 2).
V. Producing the AMICS
A survey such as the AMICS is a complex undertaking that demands the skills and expertise of a variety of different professions. The AMICS was overseen by a technical committee consisting of representatives of the Central Statistics Organization (CSO) and of UNICEF. To help produce the AMICS survey, cartographers, fieldwork trainers, data collection teams in each province, and a large team of data processors are all involved at different stages. Appendix B of the report provides a detailed list of all of the personnel involved in carrying out the AMICS. The CSO also consulted different governmental ministries and agencies at various stages of the survey and report drafting. For example, consultations were held in May 2012 when a draft of the report was shared with numerous government ministries, such as the Ministry of Education, the Ministry of Public Health, and the Ministry for Labour, Social Affairs, the Martyred and Disabled, as well as with other UN agencies.
A survey of the scope and size of the AMICS is typically a multi-year undertaking. The data collection began in October 2010 and was completed in May 2011. Processing the data using the double-entry method takes several months, which is followed by data cleaning. Tabulation and analysis is then undertaken, and the preparation of a report for public consumption is the final time-consuming step in this process.
The AMICS survey was designed to reflect the unstable conditions in which it would be carried out. It was made a priority to collect nationally representative data, and therefore, it was decided that the AMICS would need to include coverage of insecure areas. Therefore, precautionary measures in the design and delivery of the survey were taken to mitigate the risks of coverage being restricted due to insecurity.
A reserve sample of EAs was selected within each stratum to serve as possible replacements in extreme cases where the security situation for an original sample EA made it difficult to enumerate. A total of 102 sample EAs were selected as possible replacements. During fieldwork, 423 of the original 516 sample EAs were enumerated, and 26 replacement EAs were enumerated; while the remaining 67 sample EAs were not replaced due to particularly unsafe conditions for surveyors. Therefore the final sample in the AMICS data file includes 449 sample EAs. This led to a small reduction in the effective sample size; however, the sample is still considered to be nationally representative.
The AMICS 2010/11 was funded by UNICEF. UNICEF also provided some technical assistance to the CSO.