ABSTRACT
Over the last three decades, the rise and growth of knowledge-based economies have become central to the economic development of global and local economies. However, the 2008 crisis and the subsequent major economic restructuring process requires that we re-build our understanding of the knowledge economy (KE). Based on the foundations of recent agglomeration economies, the article discusses the relation between the knowledge economies and the spatial reconfiguration of regions and economic restructuring during a period of expansion (2000–2007) and a period of recession (2007–2015) using the case of the Madrid city-region. The article finds that: (1) the KE in metropolitan regions is anchored in a multicore network composed of a large or first-rank city and some medium-sized or second-rank cities; (2) employment in the KE is more resilient to a crisis period than other industries; (3) second-rank cities respond better than first-rank cities to a recession; (4) the city’s hierarchy profile (first- or second-rank) influences the KE’s sectorial composition; (5) the KE locates in relation to a complex combination of agglomeration economies, functions, amenities, and proximity to the metropolitan core. Finally, results also suggest that localisation economies and the cities’ economic functions are important for the KE concentration and specialisation.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1. For a detailed explanation of the KE’s classification framework used in this article, see Appendix 1.
2. We acknowledge that the whole period 2000–2015 hides different phases, from a strong expansion of welfare services during the initial years of the series, to the impact of the international financial crisis in 2008 and the bursting of the Spanish housing bubble, to the restructuring of the financial sector and the worsening of the recession associated with the austerity policies launched since 2010. However, for the sake of simplification, we selected just three moments in the 15-years period: boom, right before crisis, and recession.
3. The 20,000-employee cut-off coincide with that used by McMillen and McDonald (Citation1998), Cervero and Wu (Citation1998), Bogart and Ferry (Citation1999), Hall and Pain (Citation2006), and Romero et al. (Citation2014) for other comparative studies.
4. The paper uses social security employment data by place-of-work. By using employment per capita, employment of both residents and non-residents are considered. On the contrary, participation rates focus only on residents already employed or actively looking for work.
5. Formally, centres specialised in a sector have a Location Quotient greater than 1, while those not specialised have a Location Quotient lower than 1.
6. Centres with HHI values near zero exhibit a uniform distribution over space, meaning that the KE jobs they are specialised in are spread or present in other centres too. HHI values closer to 1 indicate centres with highly concentrated KE jobs in only a few centres.