University of Groningen
Raising the bar (10)
Elhorst, Paul; Abreu, Maria; Amaral, Pedro; Bhattacharjee, Arnab; Corrado, Luisa; Doran,
Justin; Fuerst, Franz; Le Gallo, Julie; McCann, Philip; Monastiriotis, Vassilis
Published in:
Spatial Economic Analysis
DOI:
10.1080/17421772.2019.1553658
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Final author's version (accepted by publisher, after peer review)
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Elhorst, P., Abreu, M., Amaral, P., Bhattacharjee, A., Corrado, L., Doran, J., Fuerst, F., Le Gallo, J., McCann, P., Monastiriotis, V., Quatraro, F., & Yu, J. (2019). Raising the bar (10). Spatial Economic Analysis, 14(1), 1-4. https://doi.org/10.1080/17421772.2019.1553658
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
EDITORIAL
Raising the bar (10)
Paul Elhorst, Maria Abreu, Pedro Amaral, Arnab Bhattacharjee, Luisa Corrado, Justin Doran, Bernard Fingleton, Franz Fuerst, Julie Le Gallo, Philip McCann, Vassilis Monastiriotis,
Francesco Quatraro and Jihai Yu
Abstract This editorial summarises the papers published in issue 14.1 so as to raise the bar in applied spatial economic research and highlight new trends. The first paper applies the Shapley-based decomposition approach to determine the impact of firm-, linkage- and location-specific factors to the survival probability of enterprises. The second paper applies Bayesian comparison methods to simultaneously identify the most likely spatial econometric model and spatial weight matrix explaining new business creation. The third paper compares the performance of continuous and discrete approaches to explain subjective well-being across space. The fourth paper applies a multiple imputation approach to determine regional purchasing power parities at the NUTS3 level using data available at the NUTS2 level. Finally, the last paper constructs a regional input-output table for Japan from its national counterpart using and comparing the performance of four non-survey techniques.
Keywords: survival, well-being, purchasing power, input-output, spatial econometrics JEL classification: C21, C67, I31, O18, M13
Spatial Economic Analysis is a pioneering journal dedicated to the development of theory and methods in spatial economic analysis. This issue contains five papers contributing to these developments. All papers are methodological in nature, illustrate their innovative findings by focusing on an empirical application, and discuss the implications of their findings from a policy point of view and/or the perspective of further research.
The first paper in this issue, by Sohns and Revilla Diez (2018), explains the survival probability of 309 micro-enterprises in three rural Vietnamese provinces over the period 2010-2013, using a three-level mixed-effects parametric model. In addition, a distinction is made between opportunity-driven (174) and necessity opportunity-driven (135) enterprises. The first group is willing to hire non-family employees and to invest more if they observe or are challenged by new opportunities in the market. The second group is more reserved since the main focus is to guarantee a sufficient level of income. The authors attempt to determine the impact of enterprise-specific factors (first level), production and consumption linkage-related factors (second level), and location-specific factors (third level). The latter factors consist of market institutional variables, including state versus non-state owned firms, pro and anti-cyclical external effects, proximity of customers and markets, and access to financial services. To determine the impact of each set of factors, the authors employ the Shapley-based
decomposition approach applied to the R-squared. The enterprise-specific factors appear to be the most important; their contribution to the survival probability amounts to 64.3% of the opportunity-driven and to 64.5% of the necessity-opportunity-driven enterprises. This is followed by respectively 23.0% and 15.4% for the linkage-related factors, and 12.6% and 20.1% of the location-specific factors. Based on these numbers the author make several policy recommendations to foster the survival and growth of micro-enterprises.
The second paper in this issue, by Credit (2018), does not deal with business survival but the related topic of business creation. It studies the relationship between rail transit proximity and the creation of new high technology businesses and it finds that transit proximity has a significant positive impact, given that the region has a relatively mature and extensive transit system, such as those in Boston and Philadelphia. The paper also finds that the exposure variable area provides the most consistent and stable foundation for calculating the expected rates of new business activity compared to other variables, such as population and existing business activities. Finally, the paper argues for the crucial role of spatial dependence when studying the impact of transit proximity on the creation of new high technology business. To investigate this role, the author applies the most advanced techniques that are currently available.
In the latest version of the Encyclopedia of Geographical Information Systems (GIS), Elhorst (2017) points out that revision is needed to the way of thinking about, and the model selection strategies that are used in, most empirical studies to determine the structure of spatial processes, and identifies two promising new approaches. The first, developed by LeSage (2014, 2015), is based on Bayesian comparison methods, and the second, developed by Halleck Vega and Elhorst (2015), is based on taking the spatial lag of X (SLX) model as a point of departure.
The Bayesian comparison method in this paper by Credit is used to test whether the SLX model needs to be extended to a spatial Durbin model (SDM) with a spatial lag in the dependent variable or to a spatial Durbin error model (SDEM) with a spatial lag in the error term. The first model implies that spillover effects are global and the second that they are local. The first occurs when a change in one of the explanatory variables at any location is transmitted to all other locations, even if two locations are unconnected according to the spatial weight matrix describing the spatial arrangement between the units in the sample. By contrast, local spillovers occur at other locations only if two locations are connected to each other according to the spatial weight matrix. Generally, global spillovers are more difficult to justify than local spillovers. Nevertheless, Credit (2018) does find evidence in favour of this type of spillovers, which he explains by the specific nature of knowledge transfers, information exchange and other agglomeration factors. This result is achieved by comparing 54 possible model specifications: 18 weights matrices, ranging from 3 to 20 nearest neighbours, and 3 model specifications, SLX, SDM and SDEM. This contribution is one of the few examples that successfully identifies the most likely candidate for the both the spatial econometric type of model and the spatial weight matrix. Previous examples appeared in Spatial Economic Analysis by Rios, Pascual and Cabases (2017), and in Regional Studies by Da Silva, Elhorst, and Neto (2017).
The third paper in this issue, by Sarrias (2018), endeavours to provide enhanced methodologies to examine subjective well-being (SWB), measured by a binary indicator, and how the
relationship between this indicator and individual characteristics vary over space. Two reasons for this spatial variation are statistical in nature, sampling variation and variables omitted from the model that follow a spatial non-stationary process, but the third and most relevant reason is that people’s preferences for some attributes are intrinsically different across space. Ignoring spatial heterogeneity in consumer preferences and compensation schemes is an acknowledged weakness of many studies, and this paper attempts to address that concern. The author compares two main specifications: (i) a random parameter specification where estimates associated with each covariate are allowed to vary across municipalities according to a normal distribution. This method has similarities with the random coefficient model originally developed by Swamy (1970), and extended with cross-sectional dependence by Pesaran (2006); and (ii) a latent class specification with a pre-specified number of groups, which provides a discrete alternative to parameter heterogeneity. Most of the discussion in the paper focuses on which approach, continuous versus discrete, is better suited to quantify compensating variation for a number of local amenities. The analysis is based on a micro-economic data set of 16,008 individuals between 15 and 64 years of age living in 324 different communes across Chili. The reviewers of this paper especially liked the positioning of the paper in a policy context, i.e., the paper explains how policies to compensate for welfare changes as a result of, for example, environmental changes, may not compensate appropriately if an averaging approach is taken to such relationships.
Spatial heterogeneity is also the topic of the fourth paper in this issue, by Rokicki and Hewings (2018). This paper constructs regional prices for Poland at NUTS2 and NUTS3. Unique raw price data for 300 goods and services are used to calculate annual regional purchasing power parity (PPP) deflators for 16 NUTS2 regions over the 2000-2012 period, following previous approaches developed by Eurostat and the OECD. Based on these indices, similar deflators are estimated for the 66 NUTS3 regions by a multiple imputation approach, a Bayesian Monte Carlo technique. Regions with the highest prices appear to be located in and around big agglomerations (especially Warsaw) and adjacent to the border with Germany. Lower prices are found in the central and eastern part of the country in which the agricultural sector plays a dominant role. Over the period 2000-2011 regional price levels do not show a clear tendency to convergence, although when employing their data imputed at the NUTS3 level the authors do find that price disparities increased in the first years following the EU accession in 2004. When using data at the NUTS2 level they are unable to find this pattern.
Given the lack of information on regional price levels within EU countries, the paper offers a number of interesting policy implications. The main one is that the allocation of structural funding in the EU based on per capita income levels might be biased, as the purchasing power might differ across regions much more than has been accounted for. Notably, rural regions might be overvalued.
The last paper in this issue, by Fujimoto (2018), is part of a series of contributions to Spatial Economic Analysis on input-output models, including for example those by Hermannsson (2016), Hermannsson, Lecca and Swales (2017), and Oosterhaven and Többen (2017). This paper constructs a regional input-output table for Japan from its national counterpart using and comparing the performance of four non-survey techniques, each based on different assumptions regarding cross-hauling to estimate export and imports. The cross-cross-hauling adjusted regionalization method
developed by Többen and Kronenburg (2015), modified by Fujimoto (2015) in a previous study published in Japanese, comes out as the best.
Hopefully, all these five methodological contributions to the literature will reach a broad audience.
References
Credit, K. (2018). Transitive properties: a spatial econometric analysis of new business creation around transit. Spatial Economic Analysis, 1-27. DOI: 10.1080/17421772.2019.1523548
Da Silva, D.F.C., Elhorst, J.P., & Neto, R.d.M.S. (2017). Urban and rural population growth in a spatial panel of municipalities. Regional Studies, 51(6), 894-908.
Elhorst, J.P. (2017). Spatial panel data analysis. In: S. Shekhar, H. Xiong, & X. Zhou (Eds.), Encyclopedia of GIS, 2nd edition (pp. 2050-2058). Cham, Switzerland, Springer International Publishing. LeSage, J.P. (2014). Spatial econometric panel data model specification: a Bayesian approach. Spatial
Statistics 9(2), 122-145.
LeSage, J.P. (2015). Software for Bayesian cross section and panel spatial model comparison. Journal of Geographical Systems, 17(4), 297-310.
Fujimoto, T. (2015). Quantitative analysis of the regional income determinant factors in the remote island economy: Generation and application of regional input-output table. Journal of Rural Economics, 86(4), 257-272 (in Japanese).
Fujimoto, T. (2018). Appropriate assumption on cross-hauling national input–output table regionalization. Spatial Economic Analysis, 1-23. DOI: 10.1080/17421772.2018.1506151
Halleck Vega, S., & Elhorst, J.P. (2015). The SLX model. Journal of Regional Science 55(3): 339-363. Hermannsson, K. (2016). Beyond intermediates: the role of consumption and commuting in the
construction of local input–output tables. Spatial Economic Analysis, 11(3) , 315-339.
Hermannsson, K., Lecca, P., & Swales, J.K. (2017). How much does a single graduation cohort from further education colleges contribute to an open regional economy? Spatial Economic Analysis, 12(4), 429-451.
Oosterhaven, J., & Többen, J. (2017). Wider economic impacts of heavy flooding in Germany: a non-linear programming approach. Spatial Economic Analysis, 12(4), 404-428.
Pesaran, M.H. (2006). Estimation and inference in large heterogeneous panels with a multifactor error structure. Econometrica, 74(4), 967–1012.
Rios, V., Pascual, P., & Cabases, F. (2017). What drives local government spending in Spain? A dynamic spatial panel approach. Spatial Economic Analysis, 12(2-3), 230-250.
Rokicki, B., & Hewings, G.J.D. (2018). Regional price deflators in Poland: evidence from NUTS-2 and NUTS-3 regions. Spatial Economic Analysis, 1-18. DOI: 10.1080/17421772.2018.1503705
Sarrias, M. (2018). Do monetary subjective well-being evaluations vary across space? Comparing continuous and discrete spatial heterogeneity. Spatial Economic Analysis, 1-35. DOI: 10.1080/17421772.2018.1485968
Sohns, F., & Revilla Diez, J. (2018) Explaining micro-enterprise survival in rural Vietnam: a multilevel analysis. Spatial Economic Analysis, 1-21. DOI: 10.1080/17421772.2019.1535184.
Swamy, P.A.V.B. (1970). Efficient inference in a random coefficient regression model. Econometrica, 38(2), 311-323.
Többen, J., & Kronenberg, T. (2015). Construction of multi-regional input-output tables using the Charm method. Economic Systems Research, 27(4), 487-507.
