New theoretical work on spatial concentration of industry – particularly the ‘new economic geography’ – has significantly helped our understanding why some regions develop more than others, why cities arise and where they are located. However, this work rarely incorporates Adam Smith's observation that spatial differences in economic activity also reflect variations in physical geography, which make some places more productive than others at particular times; nor has it accommodated regional development policy – the use of economic incentives to attract industry to particular locations. A full theory of regional development would integrate theories of agglomeration economies with physical geography and with public economics.