This paper reviews theories about the economic effects of land value taxation as well as research which suggests hypotheses addressing the disparate circumstances of central cities such as Pittsburgh, suburban cities such as McKeesport, and relatively isolated cities such as New Castle. In order to test those hypotheses, specified a general econometric model of the housing market is specified. The model is adjusted to fit the circumstances of each city, and then the adjusted models using time-series data for each city is estimated. The periods of study for each city cover spans of time during which there were both increases in the tax rates applicable to land and decreases in the tax rates applicable to improvements. Incentive effects of decreases in the tax rate on buildings are expected to encourage housing development in Pittsburgh and, possibly, New Castle, but not in McKeesport. Liquidity effects of increases in the tax rate on land may encourage housing development in the three cities. All three cities employ land value taxation as an economic development tool and as a means for helping to stem or reverse the loss of population. By encouraging the construction of housing, land value taxation may help to attract households that would otherwise locate in other jurisdictions. Although Pittsburgh has had land value taxation since 1913 (Williams 1962), McKeesport and New Castle did not adopt such a tax system until 1979 and 1982, respectively.