Two applications of wild bootstrap methods to improve inference in cluster‐IV models

Microeconomic data often have within‐cluster dependence, which affects standard error estimation and inference. When the number of clusters is small, asymptotic tests can be severely oversized. In the instrumental variables (IV) model, the potential presence of weak instruments further complicates hypothesis testing. We use wild bootstrap methods to improve inference in two empirical applications with these characteristics. Building from estimating equations and residual bootstraps, we identify variants robust to the presence of weak instruments and a small number of clusters. They reduce absolute size bias significantly and demonstrate that the wild bootstrap should join the standard toolkit in IV and cluster‐dependent models.     

A Partially Heterogeneous Framework for Analyzing Panel Data

This article proposes a partially heterogeneous framework for the analysis of panel data with fixed T. In particular, the population of cross‐sectional units is grouped into clusters, such that slope parameter homogeneity is maintained only within clusters. Our method assumes no a priori information about the number of clusters and cluster membership and relies on the data instead. The unknown number of clusters and the corresponding partition are determined based on the concept of ‘partitional clustering’, using an information‐based criterion. It is shown that this is strongly consistent, that is, it selects the true number of clusters with probability one as N→∞. Simulation experiments show that the proposed criterion performs well even with moderate N and the resulting parameter estimates are close to the true values. We apply the method in a panel data set of commercial banks in the US and we find five clusters, with significant differences in the slope parameters across clusters.     

Factors shaping the international knowledge connectivity of industrial clusters: a comparative study of two Latin American cases

Abstract

Recent research has emphasized that success of industrial clusters is not only driven by intra-cluster knowledge sharing (‘local buzz’) but also by externally sourced knowledge (‘global knowledge pipelines’). This article examines the factors that determine the channels through which clusters connect with global knowledge pipelines depending on the structure of the global value chain within which they are inserted, their knowledge base and their stage of evolution. Building on a comparative case study of the salmon farming cluster in Chile and the software cluster in Costa Rica, we adopt an evolutionary perspective based on historical analysis to better understand how the configuration of clusters’ international knowledge linkages shifts over time. Our findings suggest that (i) the more hierarchical the global value chain structure, the less room for knowledge co-creation between local and foreign actors; (ii) clusters relying on analytical knowledge bases opt for more formal and coordinated links with high involvement of public actors, whereas in clusters relying on synthetic knowledge bases, international knowledge interaction is based on less formal links mainly between business actors; and (iii) as clusters evolve the channels through which they connect with foreign knowledge increase in number and new ‘hybrid’ varieties develop.     

Building Innovation Excellence of World Class: The Cluster as an Instrument of Spatial Governance in the European Union

European spatial governance is becoming an intriguing mix of ideas from the economic, political and cultural spheres. This article asserts that, in the EU's spatial planning, the cluster is increasingly part of a hybrid spatial politics, here named the ‘cluster gaze’, based on the interplay of innovation‐oriented political rationality and spatial governance. To study this process, the article provides an empirical investigation into selected EU documentation. The investigation is based on two perspectives. First, the cluster is analysed as a mediating instrument to stimulate and rescale transnational market developments in the EU. Second, the cluster is studied as an instrument of spatial management — one that builds on a business managerial ethos and endorses a specific hierarchical spatial imaginary and a cluster evidence base to assess the productivity and efficiency of European clusters. Both of these perspectives pave the way for a European ‘politics of cluster excellence’ that is about the constant sharpening of cluster practices, continuous evaluation and ranking, and the enhancement of cluster performance to rise from the ‘European league’ towards the ‘world class’.     

Dynamics of Inventor Networks and the Evolution of Technology Clusters

Clusters are important drivers of regional economic growth. Although their benefits are well recognized, research into their evolution is still ongoing. Most real‐world clusters seem to have emerged spontaneously without deliberate policy interventions, each cluster having its own evolutionary path. Since there is a significant gap in our understanding of the forces driving their evolution, this study uses a quantitative approach to investigate the role of inventor collaboration networks in it. Inventor collaboration networks for 30 top‐performing American metropolitan clusters were constructed on the basis of patent co‐authorship data. The selected clusters operate in hi‐tech fields: information technology, communications equipment and the biopharmaceutical industry. Starting from a widely accepted hypothesis that the ‘small‐world’ structure is an optimal one for knowledge spillovers and promotes innovation effectively, the authors statistically tested the impact of ‘small‐world’ network properties on cluster innovation performance proxied by patent output. The results suggest that the effect of the small‐world structure is not as significant as theorists hypothesized, not all clusters benefit from the presence of inventor collaboration networks, and cluster performance can be affected by policy interventions. Our analyses also suggest that cluster typology moderates the impact of inventor network properties on cluster innovation performance.     

Strategic compensation in South Korea's publicly traded firms

This strategic compensation model based on contingency theory of organizations proposes that a fit between the organization's business environment and compensation system will affect the firm's performance (Rockmore, 1991; Rockmore and Scarpello, 1994). This proposition was tested with a set of Korean sample data of seventy-five publicly traded companies within thirty-four three-digit Standard Industrial Classification codes.

The cluster analysis applied to the 7-item task environment uncertainty components resulted in two group classifications that face relatively ‘more certain’ and ‘more uncertain’ environments. Principal component factor analysis of the 8-item compensation system components resulted in three compensation system factors and subsequent cluster analysis classified firm pay plans into two clusters, which we labelled ‘more traditional’ and ‘more innovative’. Discriminant analysis confirmed the resulting classifications of both task environment and compensation system types. With both ROA and EPS measures, significantly more ‘fits’ (firms in stable environment that use traditional compensation system or firms in uncertain environment that use innovative compensation system) showed financial gains compared with ‘nonfits’. This financial impact of the contingency was more salient for those firms that face uncertain environment. The same results were obtained when ANOVA was performed.     

The diffusion of exporting in Brazilian industrial clusters

The present research aimed at understanding the process by which firms in a cluster start to export based on systemic interactions, and the process of diffusion of exporting as a business strategy within the cluster. Diffusion was defined, following Rogers’ (1995: 5) seminal work, as ‘the process by which an innovation is communicated through certain channels over time among the members of a social system’. The research method used was industry case studies and the unit of analysis selected was the cluster. Two manufacturing industries were chosen to be investigated, and within each geographic area clusters were identified as the origin of dynamic export growth in the industry. Players in each industrial cluster, as well as other significant actors, were interviewed. Extensive secondary data research was done to study clusters’ historical development. Detailed analysis and a comparison of the experiences permitted the extraction of some general conclusions concerning the similarities and differences between the clusters in terms of the adoption and diffusion of exporting. Results showed that the diffusion of exporting in an industrial cluster is quite similar to the dissemination of technical innovation. Social ties were important to facilitate the diffusion of exporting in one of the clusters studied. Also, the role of domestically-owned flagship firms in leading the internationalization process proved to be important in only one of the clusters, while the role of external actors was fully supported in the two industries studied. Finally, a number of support institutions, private and public, interfered in different stages of the internationalization process. In both industries, the federal government had only a late and limited impact on export initiation and development.     

How do respondents process stated choice experiments? Attribute consideration under varying information load

The popularity of stated choice (SC) experiments has produced many design strategies in which researchers use increasingly more ‘complex’ choice settings to study choice behaviour. When the amount of information to assess increases, we wonder how an individual handles such information in making a choice. Defining the amount of information as the number of attributes associated with each choice set, we investigate how this information is processed as we vary its ‘complexity’. Four ordered heterogeneous logit models are developed, each for an SC design based on a fixed number of attributes, in which the dependent variable defines the number of attributes that are ignored. We find that the degree to which individuals ignore attributes is influenced by the dimensionality of the SC experiment, the deviation of attribute levels from an experienced reference alternative, the use of ‘adding up’ attributes where feasible, the number of choice sets evaluated, and the personal income of the respondent. The empirical evidence supports the view that individuals appear to adopt a range of ‘coping’ strategies that are consistent with how they process information in real markets, and that aligning ‘choice complexity’ with the amount of information to process is potentially misleading. Relevancy is what matters. Copyright © 2006 John Wiley & Sons, Ltd.     

Exploring the international linkages of the euro area: a global VAR analysis

This paper presents a quarterly global model combining individual country vector error‐correcting models in which the domestic variables are related to the country‐specific foreign variables. The global VAR (GVAR) model is estimated for 26 countries, the euro area being treated as a single economy, over the period 1979–2003. It advances research in this area in a number of directions. In particular, it provides a theoretical framework where the GVAR is derived as an approximation to a global unobserved common factor model. Using average pair‐wise cross‐section error correlations, the GVAR approach is shown to be quite effective in dealing with the common factor interdependencies and international co‐movements of business cycles. It develops a sieve bootstrap procedure for simulation of the GVAR as a whole, which is then used in testing the structural stability of the parameters, and for establishing bootstrap confidence bounds for the impulse responses. Finally, in addition to generalized impulse responses, the current paper considers the use of the GVAR for ‘structural’ impulse response analysis with focus on external shocks for the euro area economy, particularly in response to shocks to the US. Copyright © 2007 John Wiley & Sons, Ltd.     

The Limit Distribution of the Bootstrap for the Unit Root Test Statistic when the Residuals are Dependent

In this article, the unit root test for the AR(1) model with dependent residuals is considered. We adopt a bootstrap procedure to bootstrap the residuals with bootstrap sample size m less than the size n of the original sample. Under the assumptions that m → ∞ and m/n → 0, the convergence in probability of the bootstrap distribution function is established. Research supported by National Natural Science Foundation of China (No. 10471126)     

Linking learning with governance in networks and clusters: key issues for analysis and policy

In this paper we analyse the relationship between governance and learning in clusters and networks. In particular, we see these two key elements as interdependent, suggesting that, under particular circumstances, such interdependence may drive clusters and networks towards a dynamic development trajectory. A pure ‘governance perspective’ makes the development of any locality dependent on the system of powers which exists within the locality or across the global value chain. In parallel, a pure ‘competence-based approach’ focuses mainly on the capabilities of actors to learn and undertake activities. In contrast, we open the prospects for an interdependent relation that may change the actual competences of actors as well as the governance settings and dynamics in networks and clusters. When supported by public policies, the learning process may have the potential to modify the governance environment. Simultaneously, the learning process is intrinsically influenced by economic power, which may seriously affect the development prospects of clusters and networks. This is why an intertwined consideration of both aspects is necessary to promote specific approaches to learning and to design appropriate policies. In this paper we offer two preliminary case studies to clarify some of these dynamics: the first taken from the computers cluster in Costa Rica and the second from an Italian bio-pharmaceutical firm and its production network. The first case study refers to the software cluster that was created from scratch in Costa Rica thanks to an enlightened government policy in coordination with new local enterprises and an important foreign direct investor; while the second reflects on the ability of an individual company to create a network of relationships with large transnational companies in order to acquire new competences without falling into a subordinate position with respect to its larger partners.     

A Comparison of Sequential and Information‐based Methods for Determining the Co‐integration Rank in Heteroskedastic VAR Models

In this article, we investigate the behaviour of a number of methods for estimating the co‐integration rank in VAR systems characterized by heteroskedastic innovation processes. In particular, we compare the efficacy of the most widely used information criteria, such as Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) , with the commonly used sequential approach of Johansen [Likelihood‐based Inference in Cointegrated Vector Autoregressive Models (1996)] based around the use of either asymptotic or wild bootstrap‐based likelihood ratio type tests. Complementing recent work done for the latter in Cavaliere, Rahbek and Taylor [Econometric Reviews (2014) forthcoming], we establish the asymptotic properties of the procedures based on information criteria in the presence of heteroskedasticity (conditional or unconditional) of a quite general and unknown form. The relative finite‐sample properties of the different methods are investigated by means of a Monte Carlo simulation study. For the simulation DGPs considered in the analysis, we find that the BIC‐based procedure and the bootstrap sequential test procedure deliver the best overall performance in terms of their frequency of selecting the correct co‐integration rank across different values of the co‐integration rank, sample size, stationary dynamics and models of heteroskedasticity. Of these, the wild bootstrap procedure is perhaps the more reliable overall as it avoids a significant tendency seen in the BIC‐based method to over‐estimate the co‐integration rank in relatively small sample sizes.     

Patterns of technological learning within the knowledge systems of industrial clusters in emerging economies: Evidence from China

Through an interview-based exploratory study and a follow-up survey-based quantitative analysis, this paper investigates the technological learning pattern in terms of structure and mechanisms of interaction within the knowledge system of two industrial clusters in China. Unlike the recent studies that suggest that industrial cluster comprises disconnected leader-centered communities, we argue that the different leader-centered communities within the knowledge systems of industrial clusters are not disconnected from each other. Instead, those communities are inter-connected through the so-called ‘knowledge spanning mechanisms’. Regarding the interaction dimension of technological learning pattern, this paper argues that in analyzing learning behavior in the knowledge networks of industrial clusters, it is necessary to synthesize the learning opportunity perspective and the absorptive capacity perspective to better understand and explain the similarities and dissimilarities in technological learning behavior among different cluster types, across cognitive subgroups, and between product innovation and process innovation. Our study reveals that in the context of emerging countries, the following four factors are decisive for technological learning opportunities inside the knowledge networks of industrial clusters: the underlying complexity of technology in clusters, the inter-connectedness between product and process, path dependency in knowledge searching, and the incremental nature of a cluster’s technological development.     

Estimates of technical inefficiency in stochastic frontier models with panel data: generalized panel jackknife estimation

Estimates of technical inefficiency based on fixed effects estimation of the stochastic frontier model with panel data are biased upward. Previous work has attempted to correct this bias using the bootstrap, but in simulations the bootstrap corrects only part of the bias. The usual panel jackknife is based on the assumption that the bias is of order T ⁻¹ and is similar to the bootstrap. We show that when there is a tie or a near tie for the best firm, the bias is of order T ^−1/2, not T ⁻¹, and this calls for a different form of the jackknife. The generalized panel jackknife is quite successful in removing the bias. However, the resulting estimates have a large variance.     

The wild bootstrap,tamed at last

The wild bootstrap is studied in the context of regression models with heteroskedastic disturbances. We show that, in one very specific case, perfect bootstrap inference is possible, and a substantial reduction in the error in the rejection probability of a bootstrap test is available much more generally. However, the version of the wild bootstrap with this desirable property is without the skewness correction afforded by the currently most popular version of the wild bootstrap. Simulation experiments show that this does not prevent the preferred version from having the smallest error in rejection probability in small and medium-sized samples.     

UK consumer attitudes concerning environmental issues impacting the agrifood industry

Consumer concern for the environment has been on the increase in recent years. This concern has created many opportunities, as well as threats, for all businesses but particularly for those involved in the agrifood industry. These concerns can and do change considerably in a short space of time so it is essential that consumer surveys are conducted on a frequent basis to monitor these changes. This paper presents the findings of one such survey based on the findings of a consumer attitude study concerning environmental issues relating to the agrifood industry. The data is subjected to a cluster analysis in order to segment the respondents and make the findings more meaningful. The paper concludes that there are four groups of consumers (the ‘angries’; the ‘not‐quite‐so‐angries’, the ‘anxious‐but‐amiables’ and the ‘Panglossians’) and that while a number of subjects constitute issues for consumers the one that unites most consumers, and which will probably have the greatest impact on the environment, is genetic modification. Consumer resistance to genetic modification is found to be high and it is suggested that as a result farmers will continue to employ traditional farming practices in the future. Copyright © 2000 John Wiley & Sons, Ltd. and ERP Environment.     

Cluster emergence: combining pre-existing conditions and triggering factors

This article argues that the emergence of regional clusters relies on both necessary pre-existing conditions for cluster appearance in general and triggering factors that cause clusters to emerge in particular places. This approach is used to analyse two ‘critical cases’; the emergence of the synthetic-knowledge boat building industry in the Arendal area in Norway from the mid-1950s and the analytical-knowledge cancer medicine industry in Oslo around the year 2000. Although the industries and the contexts are otherwise very different, the framework turns out to be useful in interpreting the emergence of the two clusters. However, the specific pre-existing conditions and the triggering factors differ between the two cases. The Arendal boat building industry emerged through the combination of traditional boat building skills and exogenous knowledge of the use of new plastic material, while the Oslo cancer medicine industry built on indigenously-developed scientific knowledge. The framework is useful in putting anecdotal evidence of cluster emergence due to the efforts of entrepreneurs into a wider analytical framework of the preconditions necessary for entrepreneurs to succeed.     

Combination of Tests for Cointegration in Cross‐Correlated Panels

In this paper, Cheng and Sheng's (2017) combination of ‘combinations of P‐values’ (CCP) is extended to a combination of more than two tests and applied for cointegration testing in cross‐correlated panels. In a Monte Carlo experiment, power and size of the different combinations of combinations are investigated. If uncertainty about the panel configuration is taken into account, the results indicate that a multi‐test combination can minimize power losses. Furthermore, the usefulness of the combinations studied is illustrated by an application to international interest rate linkage. Cross‐sectional dependencies in both the simulation and the empirical studies are accounted for by using the block bootstrap.     

Changing consumer behaviour through eco-efficient services: an empirical study of car sharing in the Netherlands

In this paper ‘innovation’ is discussed as a strategy to change consumer behaviour to reduce the environmental impact of consumption. Innovation aims to improve the ‘production process’ of certain end results or ‘units of service’ for the consumer, which is defined as the ‘consumption technology’. This concept includes the entire organization of how these units of service are produced. ‘Eco-efficient services’ are introduced as an innovation strategy, based on theoretical considerations through which more eco-efficient consumption technologies and less environmentally damaging consumption behaviour could be realized. This strategy is explored through an empirical consumer behaviour study on commercial car sharing services in the Netherlands. Based on an inventory of the environmental effects of this approach, as exemplified by the car sharing schemes, it can tentatively be concluded that the eco-efficient services concept offers an interesting framework for reconsidering the current consumption practice and for generating innovative activities in industry. © 1998 John Wiley & Sons, Ltd and ERP Environment.     

Towards an Asian model of human resource management? A comparative analysis of China,Japan and South Korea

There has been much discussion regarding the possible decline and weakening of national systems of human resource management (HRM). Yet, culture and institutions are often cited as the major stumbling blocks to this outcome. Such constraints may dissipate in the analysis, particularly if ‘regional clusters’ are considered where geographic and cultural closeness and ‘openness’ to similar economic pressures exist. In such cases, it might be postulated that convergence in HRM will occur. We argue that China, Japan and South Korea represent such a cluster. This article examines these countries, to see if a degree of convergence is taking place and if it is towards an identifiable ‘Asian’ model of HRM. A model of change is presented that distinguishes between levels of occurrence and acceptance. Details of the development and practice of HRM in each country are then set out. The article ends with a discussion and implications section and a brief conclusion.