Members of the IDEAMAPS Network are performing a systematic literature review of frameworks of urban deprivation. The goal of this review is to synthesise domains and indicators which might define urban deprivation across low- and middle-income countries (LMICs), and ultimately inform what data are incorporated into IDEAMAPS models and processes.
The review includes literature from both the physical and social sciences, integrating ideas and datasets that are often discussed and understood in highly silo-ed disciplines. We present a summary of our preliminary findings and version 1 of our Domains of Deprivation framework for feedback from you all, members of the IDEAMAPS Network. We expect this framework to be iterated and improved with your input over the coming months before we submit the review and our final framework to a peer-reviewed journal for publication. Your feedback is greatly appreciated, and will be acknowledged in the published paper.
We searched for the following terms in Scopus, a database that includes over 36,000 peer-reviewed journals from life sciences, social sciences, environmental sciences, and health sciences. Note that an asterisks (*) indicates any word ending.
In Scopus, we identified 2,447 articles using the above search terms after excluding publications in languages that could not be covered by the team. 322 of these articles had a title or abstract that indicated an urban deprivation framework was described and/or applied. Our team then read the full text of each of the 322 articles to determine whether it was about an urban deprivation framework. If so, we recorded a number of metadata attributes, for example the stated geographic coverage of the article, and the domains and indicators of urban deprivation as described by the authors. This process resulted in 90 articles eligible for inclusion in the literature review from Scopus.
Given the diversity of disciplines working on issues related to urban deprivation, we are aware that our search terms might have missed some key articles. Therefore, we also performed a "snowball" process whereby we reviewed all titles of cited works in the eligible articles, and included any additional articles, reports, or books that described and/or applied an urban deprivation framework. In total, 118 articles are included in our systematic literature review.
To synthesise this literature into a single urban domains of deprivation framework, we iteratively developed a coding framework from all 1,897 indicators described in the 118 articles, and then grouped these terms into key domains. The coding framework served as the basis for our conceptual framework with eight domains and 64 sub-domains of deprivation.
domains_of_deprivation_framework_v1Download
Most existing urban deprivation frameworks (84%) were developed from a single country context, and concentrated on the UK, countries historically linked with the UK including India, South Africa, and the US, or China. The majority of articles (86%) provided an applied example in which census (61%) or survey data (59%) were a main source of indicator data, while Earth Observation data (12%) played a lesser role.
A clear trend in the literature is that many existing frameworks have been shaped by the data that are available to the authorship team; social scientists tended to define framework domains around household census and survey data (e.g. household assets, recipients of government programmes), while physical scientists tended to define framework domains around indicators that were measurable with satellite imagery or other Earth Observation data (e.g. air quality), basic government infrastructure data (e.g. roads), or volunteered geographic information (e.g. facility locations).
The IDEAMAPS Network Domains of Deprivation framework aims to reflect a full spectrum of issues that define, and result from, deprivation across LMIC cities. However, further validation and improvement of our framework are needed by speaking with slum dwellers themselves and local city planners to ensure that we have not fallen in the same trap as researchers before us by emphasising those issues for which we have information, and overlooking key issues for which new data sources are urgently needed.
