Abstract: This paper is concerned with a common question for economic geographers - why do similar industries cluster together? Since the 1970s, this enquiry gained impetus as a result of ‘declining regions’ and the contrasting success of a handful of ‘glamour‘ regions. A common characteristic of these ‘new economic spaces’ was intense interaction between local firms and institutions, with a strong element of ‘untraded interdependency’, particularly in the form of knowledge exchange and innovation. In the cut-throat world of the global economy, it was argued that these characteristics would lead to competitive advantage. I examine the key propositions of this approach in an Australian context and argue they are unconvincing. I then turn to an alternative approach and suggest that in an age of ‘ubiquities’, only labour remains embedded locally. Is labour then, the bedrock of industrial clusters? I develop this idea by revisiting regulation theory – and a case study of an ‘unglamorous’ but resilient manufacturing sector in Melbourne’s north.