Housing deficits as a frame for housing policy: demographic change, economic crisis and household formation in Indonesia

Abstract

The idea of housing deficit is a common, seemingly objective frame for housing policies that promote increased supply. This paper critically examines the concept through a case study of Indonesia, where different sources report a deficit of between 3 and 14 million dwelling units estimated without a transparent methodology. The wide range of estimates demonstrates the multiple interpretations of the term's meaning. In the paper, changes in household formation trends in urban Indonesia from 1990 to 2007 are used to estimate a quantitative housing deficit. I find the decreasing rate of household formation that is being interpreted as a housing deficit. However, this interpretation is complicated by the country's demographic transition and the high urbanization rate. Further, the abrupt change in household formation occurred around the year 2000, suggesting that the economic and political upheavals following the Asian financial crisis played an important role. Comparison of household formation rates across socioeconomic groups and urban areas shows housing markets also matter, illustrating the complexity of the issue.

Keywords:

• housing deficit
• household formation
• economic crisis
• demographic transition
• Indonesia

Acknowledgements

I would like to express great gratitude to the following people, in no particular order, for either research assistance and/or comments and feedback on this work: Peter Ellis, Arish Dastur, David Dowall, Delik Hudalah, Wilmar Salim, Arief Ramadhian, Renata Simatupang, Nathalia Marthaleta, Agustin Samosir, Isabelle Milbert, and Belinda Yuen. Any errors or omissions are my own.

Notes

¹The idea of a frame in public policy refers to normative and/or cognitive ideas in the foreground of policy debates. For more on ideas in public policy, see Campbell (2002).

²Cities were identified using a definition similar to that used by the United States Census Bureau to define Metropolitan Statistical Areas – administrative regions that include an urban ‘core’ with a population of more than 50,000. The population of the urban ‘core’ was determined using a combination of data from Badan Pusat Statistik (2007) and the size of the urban footprint from GIS data (Schneider, Friedl, & Potere, 2010). The minimum population threshold was set at 75,000 to be consistent with previous research on housing in the country (Struyk et al., 1990). Cities were grouped in categories by population size; large cities have more than one million residents, medium cities between 500,000 and one million, and small have between 75,000 and 500,000. In 2007, there were 15 large cities, 20 medium cities, and 56 small cities.

³In the United States, the census no longer uses the term ‘household head’, which implies a traditional patriarchal household structure. Instead they use the term ‘householder’ which is applied to whoever answers the census question and is not loaded with the meaning of household head. The traditional household head concept is more applicable in Indonesia.

⁴Income data are not readily available for Indonesia so the present study uses expenditures as a proxy for income and socioeconomic status, terms that are used interchangeably.