Assumptions

Knowledge-intensive companies are recognized as value-creating agents. In fields such as urban economy and economic geography, the location of companies has been a recurring subject. The location of these companies and the conditions that define this location in either the urban or intra-urban system are relevant issues. In the urban system, these activities take place primarily in the largest cities or in those where the local conditions propitiate their establishment. Within cities (the central topic addressed in this article), urban or localization economies have been discussed to explain the existence of a certain spatial pattern. Another aspect proposed is the existence of de-concentration of these activities from the center to the periphery of cities. Despite being a topical subject, this has been barely explored in developing countries, likely as a consequence of the availability of information together with certain reluctance to analytical frameworks. The objective of this article is to contribute to the knowledge about the intra-urban location of knowledge-intensive services in the Valley of Mexico Metropolitan Zone (VMMZ).

Data and Methodology

Company data were obtained from the 2019 National Directory of Economic Units (DENUE). We used Subsectors 51 to 56 of the 2018 North American Industrial Classification System (NSCIAN). These include information activities in mass media, financial services, real estate, professional, corporate, and business-support services. The number of economic units in knowledge-intensive services were grouped by postal code, i.e. we established the total number of companies for each postal code in the VMMZ.

The covariables (measured at the postal code level) that would explain the location of these companies were: average education level, economic diversity (Shannon entropy index), density of economic units (log-transformed), population density (log-transformed), distance to Mexico City center downtown square (log-transformed), distance to freeways broader than 8 lanes (log-transformed), distance to the nearest subway station (log-transformed), and distance to the nearest park or urban area greater than 400 m2 (log-transformed).

Due to the significant number of postal codes with no knowledge-intensive companies, we decided not to derive estimates using a least-squares linear regression. Instead, we used Poisson, binomial-negative, and Hurdle regressions in their “original” and excess-zero versions. These have the advantage of being more precise in the presence of “long tails” and categorical variables.

Results

In spatial terms, at least three clusters of knowledgeintensive companies were identified, mainly in the center and north-west areas of the city and in the vicinity of the airport.

The variables accounting for the greatest explanatory power were density and diversity of companies, and average education level. However, the results are relevant regarding the difference between the zeros component (location vs. non-location) and the number of companies. That is, variables such as education level influence the location of companies, but not necessarily the number of companies. On the other hand, the number and diversity of companies, as well as their distance to freeways and to subway stations are significant variables in relation to the number of companies. In other words, the presence of other companies and the transportation and mobility variables influence the location of other companies.

Discussion and conclusions.

Knowledge-intensive activities have been consolidated as among the most relevant in urban economies. In developing countries, efforts are still needed both regarding the importance of these activities and regarding patterns of inter- and intra-urban location. This paper intends to identify some of the relevant factors that affect the location of these types of activities. The location of these companies is influenced by the density and diversity of companies, as well as by transportation and mobility infrastructure, that is, they behave according to a logic of urbanization economies and take advantage of the urban “environment” developed mainly in the most consolidated areas of the city. An additional factor to consider is that the quality of employment (measured through the education level) was found to be a variable that influenced only the location decision but not the number of knowledgeintensive companies. This would mean that skilled labor is relevant, but not necessarily in situ. Finally, urban space quality (measured through distance to green areas) is not statistically significant, which contrasts with the findings in other studies.

This is a first approach to this topic from a spatial standpoint and still requires further in-depth studies.