Annex A. Survey method and implementation

Households were selected using sequential random sampling, which consists of the following stages:

• Households are selected by means of random numbers associated with serial numbers of households in the list.

• Households already selected cannot be selected again.

• Households are selected and interviewed strictly according to the order of generating the random numbers.

• Households, which failed to be interviewed (not found, refusal, etc.), are replaced with the next ones, according to the order of the random numbers.

• Selection is repeated until a required number of interviews is reached.

Interviewers used the Kish grids to select respondents at household level. The interviewers had a Visit Registration Form and a Respondent Selection form together with each questionnaire. The procedure of selection of respondents consisted of the following stages:

• All household members eligible for the survey (aged 18-64) were sorted by gender (primary sorting key) and then by age (secondary key).

• Each of them was assigned a serial number.

• A respondent was determined according to the Kish grid.

• If the determined respondent was not eligible by quota, the next household member was chosen for the survey.

In general, respondents had no trouble understanding the questions in the questionnaire. Interviewers, however, faced several issues while carrying out the interviews:

• Some of the respondents refused to take part in the survey. Reasons included:

  • Respondents were busy and had no time for answering questions.

  • Questions devoted to credits and loans made many respondents feel uncomfortable and caused aggression and unwillingness to answer the questions, especially in the southern parts of Kyrgyzstan.

  • Some respondents showed disinterest in green economy and sustainable development.

  • Some respondents refused to give a phone number for security reasons.

• The presence of heavy doors, surveillance cameras and fences in Bishkek made it difficult for interviewers to physically access households. Moreover, in many parts of Bishkek, warnings about frequent thefts and attacks are posted at the entrance doors of the porches. They ask people not to open the door to strangers. Interviewers had to find the house committee and explain to them about the survey and its purposes.

• Due to the arrest of Kyrgyzstan’s former president, political instability occurred during the fieldwork stage. This created life-threatening and therefore unfavourable survey conditions in Bishkek and the Chui region.

• During the fieldwork, local people had conflict with Chinese workers in Solton Sary. This caused dangerous and unfavourable conditions for survey conditions in the Naryn region.

Household access to banking facilities varies depending on where people live in Kyrgyzstan (rural or urban settings). Therefore, the survey makes an important contribution to represent the geographic distribution of the population (Annex B further elaborates this point). The survey has, however, several limitations.

First, the survey excludes people under the age of 18. While there are legal reasons for this, the population under 18 makes up more than one-third of the population (National Statistical Committee of the Kyrgyz Republic, 2019[1]). Results might look different if the population between 16 and 18, who are allowed to earn money, had been included in the study. In particular, results on use of mobile banking might look different.

Second, someone within the household might have a bank account or have taken out a loan, but the respondent may have interviewed another household member. The survey design did not take this into account. This should be kept in mind when interpreting results on use of bank accounts.

Third, socio-economic characteristics such as income or type of occupation were not included in the quota to select respondents. Readers should therefore not interpret these as representative of the national population.

Lastly, this report has focused on households. To some extent, household finances are closely interlinked with those of small business holders and individual entrepreneurs. Findings on financial literacy, services and products will have some predictive power for Kyrgyz firms and potentially a higher degree of precision compared to sampling firms. Nevertheless, the sample only captures issues related to the business environment to a limited extent.

The data gathered here, despite their limitations, make an important contribution to the empirical evidence on access to financial products and services and their use in Kyrgyzstan.