2902.0 - Census Update (Newsletter), Sep 1999  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 20/10/1999   
   Page tools: Print Print Page Print all pages in this productPrint All

Census taking in Central Asia

Martin Butterfield, Director of Census Processing & User Services, recently spent three weeks in Kyrgyzstan as part of a United Nations delegation and has this to report:

The United Nations, has over the years, issued a series of handbooks and technical reports intended to assist countries in conducting population and housing censuses. In 1998, the Australian Bureau of Statistics' Census Program drafted a manual on how to manage a census for the United Nations Statistics Division (UNSD). To maximise the value of the manual in the round of censuses being conducted in 2000, the draft manual was intended for use in workshops run by UNSD in 1999.

The first of these workshops was run in Bishkek, the capital of Kyrgyzstan in May/June 1999. Seven countries participated in the workshop: Kyrgyzstan, Kazakstan, Tadjikistan, Uzbekistan, Azerbaijan, Turkmenistan and Mongolia.

Five members of the UN team presented at the workshop and all came from different countries (including Australia, Indonesia and Italy). The members brought with them census-taking experiences from an even wider range of countries, including: Laos, China, Palestine, Yemen, South Africa, Ethiopia and Colombia.

A dominant factor in the workshop was the transitional nature of the Central Asian societies. Until the early 1990s, they had all been part of the Soviet Union and were now experiencing the process of transition to a market-based economy. An immediate benefit of this transition is that there is considerable pride in undertaking, and participating in, their first national census. This also provides a focus for the State Statistics Committee to ensure that response rates are high.

As a considerable proportion of the seven countries' populations are unemployed, allowances must be made when developing recruitment and remuneration practices. Unemployed people who are involved in the census receive a full wage for their workload. State employees such as teachers (who made up about 50% of the collection workforce in Kazakstan), on the other hand, only receive a 'bonus' because they continue to receive their normal salary.

The countries all present interesting logistical challenges in conducting a census. Kyrgyzstan and Tadjikistan are very mountainous while Turkmenistan is 80% desert. In all countries the roads could be expected to be of very poor quality due to low maintenance. Fuel for trucks is very expensive (as a proportion of total costs of the census, transport was about 9% in Central Asia compared to approximately 3% in Australia).

Further challenges arise in these countries due to the presence of a significant nomadic population. This leads many countries to run their censuses in winter as the population is more readily countable. By way of contrast, Tadjikistan is planning to undertake its high altitude census (areas above 4,000 metres) in October, before mountain passes get snowed in for winter.

The forms are generally prepared in both Russian and the relevant national language. Once the data is collected, a wide range of means are envisaged for capturing data from the forms. These range from full character recognition, as intended for Australia, to a hybrid system based on use of a mouse to enter data through pick lists.

A particularly innovative technique used to boost response rates in Kyrgyzstan was to present prizes to selected people who had registered to complete a census form. The high point of this was the inducement offered by the Mayor of Bishkek: one lucky respondent will score a new apartment by filling in their form!