Datafication,development and marginalised urban communities: an applied data justice framework

ABSTRACT

The role of data within international development is rapidly expanding. However, the recency of this phenomenon means analysis has been lagging; particularly, analysis of broader impacts of real-world initiatives. Addressing this gap through a focus on data’s increasing presence in urban development, this paper makes two contributions. First – drawing from the emerging literature on ‘data justice’ – it presents an explicit, systematic and comprehensive new framework that can be used for analysis of datafication. Second, it applies the framework to four mapping initiatives in cities of the global South. These initiatives capture and visualise new data about marginalised communities: residents living in slums and other informal settlements about whom data has traditionally been lacking. Analysing across procedural, rights, instrumental and structural dimensions, it finds these initiatives deliver real incremental gains for their target communities. But it is external actors and wealthier communities that gain more; thus, increasing relative inequality.     

Big Data for Development: A Review of Promises and Challenges

The article uses a conceptual framework to review empirical evidence and some 180 articles related to the opportunities and threats of Big Data Analytics for international development. The advent of Big Data delivers a cost‐effective prospect for improved decision‐making in critical development areas such as healthcare, economic productivity and security. At the same time, the well‐known caveats of the Big Data debate, such as privacy concerns and human resource scarcity, are aggravated in developing countries by long‐standing structural shortages in the areas of infrastructure, economic resources and institutions. The result is a new kind of digital divide: a divide in the use of data‐based knowledge to inform intelligent decision‐making. The article systematically reviews several available policy options in terms of fostering opportunities and minimising risks.     

School ethnic diversity and White students’ civic attitudes in England

The current paper focuses on White British students in lower secondary education and investigates the effect of school ethnic diversity on their levels of trust and inclusive attitudes towards immigrants. Use is made of panel data of the Citizenship Education Longitudinal Study (CELS) to explore these relationships. Ethnic diversity is measured with the proportion of students in a grade identifying with a minority. In agreement with contact theory, the paper initially finds a positive relation between diversity and inclusive attitudes on immigrants. However, this link disappears once controls for social background, gender and prior levels of the outcome are included in the model. This indicates that students with particular pre-enrolment characteristics have self-selected in diverse schools and that inclusive attitudes have stabilized before secondary education. Diversity further appears to have a negative impact on trust, irrespective of the number of controls added to the model.     

Common Factor Analysis versus Principal Component Analysis: A Comparison of Loadings by Means of Simulations

Common factor analysis (CFA) and principal component analysis (PCA) are widely used multivariate techniques. Using simulations, we compared CFA with PCA loadings for distortions of a perfect cluster configuration. Results showed that nonzero PCA loadings were higher and more stable than nonzero CFA loadings. Compared to CFA loadings, PCA loadings correlated weakly with the true factor loadings for underextraction, overextraction, and heterogeneous loadings within factors. The pattern of differences between CFA and PCA was consistent across sample sizes, levels of loadings, principal axis factoring versus maximum likelihood factor analysis, and blind versus target rotation.     

Robust weighted generalized estimating equations based on statistical depth

Most of the longitudinal data contain influential points and for analyzing them generalized and weighted generalized estimating equations (GEEs and WGEEs) are highly influenced by these points. An approach for dealing with outliers is having weight functions. In this article, we propose some new weights based on the statistical depth. These weights express centrality of points with respect to the whole sample with a smaller depth (larger depth) for the point far from the center (for the point near the center). The proposed approach leads to robust WGEE. These approaches are applied on two real datasets and some simulation studies.     

Internal resource waste and centralization degree in two-stage systems: An efficiency analysis

Internal resource waste refers to the waste in the intermediate resources between the upstream stage and downstream stage in a production or service system. This study examines a system with a two-stage structure, in which the outputs from the first stage are taken as the inputs for the second stage. Two-stage systems can exist in centralized, decentralized, or mixed organizational modes. In this paper, we propose two-stage DEA models considering a degree of centralization that makes it possible to measure internal resource waste in different system modes. Some managerial insights are tested and verified from the perspective of efficiency analysis. We find that: 1) when there is only one intermediate measure in a centralized two-stage system, internal resource waste can be eliminated completely, and 2) a higher degree of centralization in a two-stage system can lead to less internal resource waste and more expected outputs. Finally, we present a numerical example and two practical real-world examples that illustrate our approach and findings.     

Non-radial profit performance: An application to Taiwanese banks

This research modifies the directional Russell measure (DRM) of Fukuyama and Weber (2009) [1] to decompose the Nerlovian profit efficiency in Chambers et al. (1998) [2] so as to obtain a generalized measure that completely excludes technical inefficiency from allocative inefficiency. Based on such a decomposition, we further develop a new slack-based and profit-oriented productivity indicator, combining the Nerlovian profit measure with the conventional Luenberger productivity indicator (LPI), in order to provide a full picture of the sources of productivity change. Productivity change, based on the profit boundary, is decomposed into four components: the change in technical efficiency; the change in allocative efficiency; the shift of technology; and the price effect from outputs and inputs. This decomposition provides a more complete picture of the sources of productivity change. The above indicator is used herein to measure the productivity change of Taiwanese banks in terms of profit.     

Social welfare support and homicide: Longitudinal analyses of European countries from 1994 to 2010

The purpose of this research is to explore the extent to which retrenchment in welfare support is related to homicide trends across European countries between 1994 and 2010. Using a longitudinal decomposition design that allows for stronger causal inferences compared to typical cross-sectional designs, we examine these potential linkages between social support spending and homicide with data collected from a heterogeneous sample of European nations, including twenty Western nations and nine less frequently analyzed East-Central nations, during recent years in which European nations generally witnessed substantial changes in homicide rates as well as both economic prosperity and fiscal crisis. Results suggest that even incremental, short-term changes in welfare support spending are associated with short-term reductions in homicide—specifically, impacting homicide rates within two to three years for this sample of European nations.     

Stochastic efficiency analysis with a reliability consideration

Stochastic Data Envelopment Analysis (DEA) models have been introduced in the literature to assess the performance of operating entities with random input and output data. A stochastic DEA model with a reliability constraint is proposed in this study that maximizes the lower bound of an entity׳s efficiency score with some pre-selected probability. We define the concept of stochastic efficiency and develop a solution procedure. The economic interpretations of the stochastic efficiency index are presented when the inputs and outputs of each entity follow a multivariate joint normal distribution.     

大数据利益相关者的利益矛盾及其伦理治理

在一切皆可数据化的条件下,数据成为最宝贵的资源。在大数据利益相关者共同体中,由于分工的不同和地位的差异,大数据搜集者、大数据使用者与大数据生产者相互之间必然处于尖锐的矛盾之中。因此,为了实现数据共享的健康运行和促进大数据时代的顺利发展,有必要进行伦理治理,为大数据利益相关者制定出相应的伦理原则。