How do Ghana’s landfills affect residential property values? A case study of two sites in Accra

Abstract

Like many countries in sub-Saharan Africa, Ghana lacks well-engineered sanitary landfill sites. Increased urbanization and concomitant real estate growth lead landfills to compete with residential land use, resulting in closer proximity between landfill sites and residential neighborhoods. The effects of landfills on the property values of nearby residential communities have been the subject of much debate in the developed world, where state-of-the-art and environmentally well-engineered landfills are common. However, academic and other research is inconclusive on the effects of landfills on property values in the developed countries. This paper addresses this knowledge gap by exploring the effects of landfills on residential property values in Ghana, using the Oblogo and Mallam landfills in Accra as a case study. Our analysis indicates that while landfills do depress nearby residential property values, the effects are contingent on property location relative to the level of urbanization in a community, and year of completion and total costs of property development.

Keywords:

• Accra
• landfills
• property values
• hedonic pricing models
• OLS regression

Notes

1. See Appiah-Kubi (2004) for a detailed discussion of District Assemblies’ (local governments) noncollection of property taxes in Ghana.

2. Because of the incremental nature by which houses are developed in Ghana, it is very common to find people occupying uncompleted buildings while development of the structure proceeds.

3. Includes Ga East, Ga West, Ga Central, and GA South Municipalities.

4. City local governments are referred to as either Metropolitan Assemblies or Municipal Assemblies. Metropolitan Assemblies are for cities/large towns with populations of over 250,000 while Municipal Assemblies are defined as towns with populations of 90,000–250,000.

5. The size of residential properties and materials used give some indications of property value, socio-economic background of property owners, and the classification of the neighborhood as “Class A.”

6. One Ghana Cedis (GhC) was equivalent to US$1.4 at the time of survey (June–July 2009).

7. The price variable (total cost incurred in putting up the property) per our estimate warrants inclusion in the model. Aside from the fact that the initial costs of construction affect the current sale value, a test for correlation between the independent variables (multicollinearity) using the variance inflation factor (VIF) did not reveal any multicollinearity problems. The mean VIF for the regression estimate is 4.16, which is distant from the standard multicollinearity threshold of >10. Indeed, if the price variable is removed, only LSIZE (the size of the land) is significant in the model and the R² drops. This implies that the price variable contributes to the overall model fit and its omission engenders omitted variable bias.

8. This is advocated in Ghana’s medium-term development policy frameworks including the Growth and Poverty Reduction Strategy (GPRS II) and Ghana Shared Growth Development Agenda (GSGDA) (National Development Planning Commission [NDPC], 2005, 2010).