Classifying administrative divisions by quality of housing

By Dr. Amara Satharasinghe

The Department of Census and Statistics recently released a report titled ‘Classification of Administrative Divisions  According to Quality of Housing – 2012,’ classifying the Administrative Divisions namely District, DS Division and GN Division, according to the  housing quality, an index computed using the provisional housing data already released from the recently concluded Census of Population and Housing.

Basic need

Housing is a basic need. Housing comprises of a whole gamut of shelter, and the attendant infrastructures such as roads, electricity, communication and transportation etc. Good-quality housing is a key element for ensuring a healthy and productive population. Poor housing can lead to poor health. Crowded and low quality housing conditions give rise to poor hygiene by providing places for vermin to breed and transmit diseases via fleas, ticks and other vectors.

Poor household hygiene leads to food and water contamination within the home. Poor indoor air quality leads to respiratory problems and inadequate lighting leads to eyesight problems. Besides physical illness, poor housing can also lead to psychological problems. Stress and related psychological problems are higher for individuals living in poor housing and poverty.

Poor housing caused largely due to poverty in turn leads to poverty not only in terms of economic deprivation but also in terms of poor health and social ill being. Therefore, it is important to assess the quality of housing in a country to help take measures to improve its quality to standards that are necessary and affordable in the national context.

For purposes of designing interventions, it is essential to have assessments at small area level. This publication presents the results of a special study undertaken by the Department of Census and Statistics to produce estimates of housing quality at the level of the GN division.

The study takes advantage of the availability of provisional summaries of data on certain housing characteristics that were prepared at the time of the data collection of the 14th Census of Population and Housing that was conducted in March 2012. Even though the census data are not yet fully available, this information is used to provide much needed data in advance of the release of final census data.  

An index named Housing Quality Index (HQI) is developed to provide a summary measure of the quality of housing. Based on the HQI, the GN divisions are ranked into five categories: very high, high, average, Low and very Low; and these are shown in maps for easy visual examination.

Poverty

While there may be a consensus on an appropriate definition of basic housing, this presumption may not hold sway for poverty. Poverty can be defined or viewed from various perspectives such as income levels and wages, social welfare, assets, access to basic infrastructure, income per capita or affordability. However, evidence shows that there is a direct correlation between housing and poverty. The quality of housing and the standard of living or poverty are covertly or directly proportionate.

Poverty is defined as a multidimensional issue, characterized by the lack of, or limited income and is commonly associated with multiple forms of deprivation and consequences caused by inability to purchase basic goods and necessities. Poverty occurs mainly at the individual or household level but, the most visible evidence of poverty arises when poor families and individuals cluster in an area. These areas which are challenged economically and disproportionately bear the social and economic burden of unemployment, crime, deteriorated housing, and poor health.

Accordingly, the need to provide adequate, suitable and equitable housing has remained a major priority of the government. Adequate housing is one of the effective means to alleviate poverty because shelter is usually the most expensive item for households. It is also a pre-requisite for better health, providing a great amount of saving when one is not sick.

Statistics

Under these circumstances, in order to make informed decisions and to make effectively targeted interventions on improving the quality of housing, statistics on quality of housing for small population groups or communities living in poor quality housing is required.  

Statistics on conditions of housing are compiled through surveys and Population and Housing censuses. But survey data can be used to compile these statistics at district level only. Population and Housing censuses yield such data down to the level of GN division.

The Census of Population and Housing is the largest statistical undertaking in a country. A population census is the only source that provides reliable and detailed statistics on the size, distribution and the composition of population and housing of a country. The 14th Census of Population and Housing of Sri Lanka was conducted in March 2012. The enumeration stage of the Census was carried out in February‐March 2012. Information collected at this Census is of utmost importance to Sri Lanka since this Census covered the entire country after a lapse of 30 years.

As stated earlier, this study was undertaken to develop an indicator called ‘Housing Quality Index’ to measure quality of occupied housing units, at the smallest administrative level of GN division using housing data that have been already released using statistical tools. Percentage of occupied housing units for which principal source of lighting is either the National Grid or a rural power projects, Percentage of occupied housing units having toilets for exclusive use, Percentage of occupied housing units of which permanent materials: bricks, cement blocks/stones or cabook have been used for the construction of walls, Percentage of occupied housing units for which permanent materials: tiles, asbestos, concrete, zinc aluminium sheets or metal sheets, have been used for the construction of roofs, Percentage of occupied housing units which are not ‘raw houses’ ‘line houses’ ‘shanties’ or ‘other types’ are the indicators used to compile the Housing Quality Index.

This report provides maps depicting the spatial distribution of quality of housing across administrative divisions of District, DS and GN to facilitate more user-friendly use of the information.

Classification

In grouping each administrative division into five classes, a statistical tool called ‘Natural Break’ method was applied.  Those classes were labelled as Very high, High, Average, Low, Very Low. This method identifies break points by looking for groupings and patterns inherent in the data. The administrative divisions are divided into classes whose boundaries are set where there are relatively big jumps in the HQI data values by which within class variation is minimised. This ensures administrative divisions in each group are homogeneous with respect to the values of HQI.

This classification was carried out at the GN division level of each district separately so that within district variation of HQI can be compared across GN divisions. Using the average values of HQI, DS divisions and Districts were also classified. No of DS divisions and No of GN divisions falling into housing qualities of Very High, high, average, low and very low by district are shown in the table.

To facilitate the comparison of spatial patterns, standard colour-codings were used in preparing maps. In the same way maps are prepared for the five indicators selected for the study as well. Evidence of validity of this measure was found from field visits.

As stated above, this study provides an index on housing quality ‘HQI” which can be considered as a proxy measure of poverty levels at GN, DS and district levels. For poverty reduction and equity focused development programs it is essential to reach the most marginalised. However, these small area statistics on poverty are hard to obtain for reasons of practicality and cost.

Lack of poverty data for small areas is a conspicuous and often spoken about gap in our knowledge base.  It has been shown that there is a positive correlation between poverty and quality of housing. In the absence of poverty measures at GN level, HQI can provide a proxy measure of poverty which could capture at least some dimensions of poverty. Therefore, HQI can be used to identify small areas at low quality housing units which can correspond with high levels of poverty.

In the report:  ‘Classification of Administrative Divisions by Quality of Housing: 2012,’ spatial variation of HQI and other indicators used for this study have been presented in maps for easy visual examination of housing quality across administrative divisions: District, DS division and GN Divisions together with statistical data tables.



(The writer is Additional Director General of the Department of Census and Statistics.)