This project is maintained by amartrics
Strachan, Glenda, et al. “Gender (In)Equality in South Asia: Problems, Prospects and Pathways.” South Asian Journal of Human Resources Management, 2015. https://doi.org/10.1177/2322093715580222.
This editorial from the South Asian Journal of Human Resources discusses a preeminent social issue that has deeply impacted the economic and developmental wellbeing of South Asia: gender inequality. Examining the topic of gender-based inequity through an economic lens – and simultaneously answering the question of how they are connected via correlative unfreedoms – offers a unique perspective on this problem’s causes as well as steps that can be taken to mitigate its deep-rooted, wide-reaching effects. In “Gender (In)Equality…,” authors Glenda Strachan, Arosha Adikaram, and Pavithra Kailsapathy utilize geospatial data about employment rates, pay gaps, occupational segregation, and union involvement to connect different countries’ financial struggles and successes to their support of women in the workplace. Their research indicates that women as a heterogenous group experience “insecure employment” (Strachan et al.), unfair pay, horizontal and vertical gender segregation in corporate environments, and low-status jobs in unorganized labor sectors at a rate far higher than their male counterparts. South Asia in particular has recorded some of the lowest employment rates for women worldwide, with women representing a shockingly low 20 percent of all employees in Pakistan, India, and Bangladesh. Outdated notions of gendered occupations force women in the area to primarily hold caretaker jobs and avoid the corporate ladder altogether, acting as a barrier to developmental achievements like fully-equal political participation (women make up barely 3% of politicians and C.E.O.s in India) and egalitarian market opportunities. Beyond the formal workplace environment and the bureaucratic realm, South Asian women also routinely participate in economic activity as “unpaid family workers,” but receive little to no compensation for their contributions to familial rates of gross domestic product (Strachan et al.). For example, in Sri Lanka, 20.5% of women – but only 3% of men – in the region are considered “unpaid family workers.” Citing data from the likes of the World Economic Forum and the Gender Gap Index as well as business reports, H.R. journals, and comparative diversity studies in the specific context of the South Asian region, the authors of “Gender (In)Equality…” make a strong case for the direct connection between economic and social unfreedoms. While their data is not entirely infallible – and the largely survey-based method of its collection increasingly acknowledged as somewhat outdated in the data science community – the margin of error is slim enough and the sampled data wide enough to prove there are concrete roadblocks for Southeast Asian women along the road to development. As implied by Amartya Sen in his book Development as Freedom, both the social and economic aspects of development mentioned in this article must be monitored carefully in order to dismantle the unfreedoms that prevent women from achieving them.
Vaitla, Bapu, et al. “Big Data and the Well-Being of Women and Girls – Applications on the Social Scientific Frontier.” Data2x, Apr 2017. https://data2x.org/resource-center/big-data-and-the-wellbeing-of-women-and-girls/.
In Bapu Vaitla’s comprehensive report on the implementation of “big data” for the furthering of gender equity in southeast Asia, the results of several unique data collection methods demonstrate staggering imbalances between the “well-being” of men and women in the region. Gender-based social unfreedoms dominate the region of southeast Asia in the form of patriarchal norms and the silencing of female opinions. But beyond the more visible indicators of lacking “wellness” and development for women in the region (e.g. socioeconomic statistics, mapped access to healthcare facilities), Vaitla argues that a more insidious hallmark of failed equity in the context of development remains hidden in the form of a disarming lack of care for the issue of gender equality among southeast Asian populations. Data collected by the United Nation’s Global Pulse program through a post-2015 study on the geographic concentrations of online conversations demonstrates that the topic is of great import to tech users in Nepal, Bangladesh, and India. Out of seven quantitative categories of Internet activity regarding the topic of “gender inequality between men and women,” Nepalese tech users fell into the most concentrated or frequent category of discussion, while Bangladesh and India were listed in the second-highest category. (In comparison, data collected from the United States placed it in the 4th category, while most European countries ranked in the 6th or 7th categories.) With so many social taboos surrounding open, offline dialogue about this issue in these countries, its rise in prevalence as a topic of online conversation may, at first glance, seem to indicate significant social progress. However, Vaitla’s report demonstrates that the topic of gender equality in southeast Asia is still viewed as an exclusively female issue, even online. Further examination of the UNGP study’s results reveal that female Twitter users in the region of southeast Asia are consistently more likely to discuss topics like “equality between men and women,” “protection against crime and violence,” and “freedom from discrimination and persecution” (Vaitla et al. 22) than their male counterparts. Based on the predictions and filtered results of the machine-learning technology used to disaggregate the online patterns of all Twitter users in the study, men in the same geographic area appear to care more about “protecting forests, rivers, and oceans” than any of the aforementioned topics. In the broader context of human development, this disparity of interest raises serious concerns. As Amartya Sen mentions in his book, Development as Freedom, popular opinion and free agency are both vital factors in the process of social transformation, as they have a direct influence public policy (Sen 18). The fact that the issue of social equity is viewed as unimportant to half of the region’s population poses a significant problem on the road to development; to eliminate social unfreedoms, gender discrimination must be openly discussed by all members of a society.
Bosco, C., et al. “Exploring the high-resolution mapping of gender-disaggregated development indicators.” J. R. Soc. Interface, 2017. http://dx.doi.org/10.1098/rsif.2016.0825.
Some articles about gender-specific development issues tend to focus on qualitative rather than quantitative data to support their authors’ points. However, this report utilizes an approach that is entirely different, highlighting a unique method of data collection used to quantify oft-overlooked or “invisible” social data related to gender inequality in Bangladesh, Kenya, Nigeria, and Tanzania. One of the biggest problems faced by aid organizations and community-based startups aiming to equalize social gender imbalances stems from a lack of usable data. Using “gender-disaggregated high-resolution maps” (C. Bosco et al. 1) produced by a combination of Bayesian generalized linear models, artificial neural networks, and repeated cross-validation processes, C. Bosco and his co-authors have developed a modelling technique able to accurately predict social, health, wealth, and resource-based inequity across entire geographic regions, which could resolve many of the pitfalls associated with localized data collection. While other, older methods of organizing and analyzing geospatial data are still relevant and usable within the context of human development (see “Gender (In)Equality…”), the results generated by the model discussed in this article speak for themselves. Instead of relying on nationwide surveys and simplified census data, the high-resolution mapping of development indicators combines geospatial covariates with correlating, pre-established data pools to more accurately predict social, health, and economic phenomena in both tiny pockets and large swathes of land. Though there are some drawbacks associated with the use of this method, C. Bosco and his team managed to chart literacy, stunting, contraception use, and income rates on various geographic scales with a relatively small margin of error “in the range of 2-30% explained variance…” – variance largely attributed to gaps in the project’s original referential datasets (5). For instance, the high-resolution map of Bangladesh produced by this method was able to display the geographic clustering of women with little to no education alongside geospatial concentrations of stunted children, concisely linking two distinctive indicators of development. Seeing as how almost 40% of Bangladesh’s population under the age of 5 are reported to have stunted growth – and considering that children with uneducated mothers “are more than twice as likely to be short for their age” (2) – this map, and similarly modelled depictions of data, could prove to be an extremely useful tool in the demonstration and prediction of eventual progress towards a more aggregated future in terms of gendered development.
Bosak, Keith and Schroeder, Kathleen. “Using Geographic Information Systems (GIS) for gender and development.” Development in Practice, Apr 2005, vol. 15, no. 2, pp. 231-247. JSTOR, https://www.jstor.org/stable/4030084?seq=1#metadata_info_tab_contents.
“Using Geographic Information Systems (GIS) for gender and development,” a peer-reviewed article written by Keith Bosak and Kathleen Schroeder, discusses in detail exactly what its title implies: the wide-scale use of land-use oriented mapping programs to further research and progress within the realm of gendered development. As a method of organizing information, geographic information systems are highly-accurate and descriptively expansive, tying together different types of variables – including spatially referenced, video archival, and demographic data – with locational attributes in order to map and manipulate them. Without enough information, though, G.I. systems can be of no use, as in the case of gender-disaggregated development in southeast Asia. One of the major issues with ensuring equity for women in “undeveloped” locales in this region like Nepal, Bangladesh, and Sri Lanka stems from the lack of high-quality data collected on the topic. As Bosak and Schroeder quickly point out, this problem can be directly linked to low rates of involvement amongst women in the field of data science, which has led to the long-term incubation of inherent bias in modern data collection tools. However, even when women are directly involved with the collection of data – as in the case of a GSI-based breast cancer research project conducted in Long Island, New York in 2002 (Bosak and Schroeder 234) – much of it ultimately ends up being filtered out or disregarded by the intrinsically-biased criteria of geographic databases. This conclusion begs the question: how can feminist-minded data scientists, geographers, and developmental experts hope to a) eliminate gender-based bias from such set-in-stone methods of data collection, and b) alleviate the massive discrepancies between the available data on female populations in “developed” versus “undeveloped” parts of the world? Bosak and Schroeder’s own research, funded by the Consultative Group on International Agricultural Research, presents a partial solution to this complex question by suggesting the combination of traditional GIS with contextual information, a process they tested “to determine ‘hot-spots’ of female poverty in Nepal, Bolivia, and Malawi” (231). Most household surveys, which are typically referenced by G.I. systems to calculate geotagged poverty measures, “mask inequality within households” and amongst family members, failing to recognize important data like income disparity between husbands and wives (234). Recognizing this bias, researchers can adjust their analysis of data to generate more sophisticated indicators of development based on gendered experiences in GIS databases, a technique Bosak and Schroeder applied to their own research on the women of Nepal’s Eastern Terai region. Though the area is regarded as the country’s “most agriculturally productive women,” the female workers and housewives living there are recorded to have the “lowest average body mass of any region” (234). Relying on national data or even non-disaggregated regional data to examine malnourishment amongst Nepalese farming populations, a GIS would not be able to predict these results, indicating its shortcomings as a solely quantitative modelling tool. Similar deviations between the fiscal, physical, and social health of men and women across southeast Asia are also often overlooked by “old-school” methods of data collection. By layering a series of localized partial perspectives over geospatial data collected using an approach of positivist epistemology, however, Bosak and Schroeder argue that a “more objective representation of reality” can emerge (235), allowing for a more accurate vision of development within the context of southeast Asian cultural norms.
Lopes, Claudia A. and Bailur, Savita. “Gender Equality and Big Data – Making Gender Data Visible.” United Nations Entity for Gender Equality and the Empowerment of Women, Jan 2018. UN Women, https://www.unwomen.org/en/digital-library/publications/2018/1/gender-equality-and-big-data.
In this UN Global Pulse report, Drs. Claudia Lopes and Savita Bailur present an in-depth review of the benefits and drawbacks of integrating “big data” with sustainable development goals, focusing especially on the field’s relationship to the progress of SDG 5 – the “[achievement] of gender equality and [empowerment] of all women and girls” (Lopes and Bailur 2). As noted by researchers Keith Bosak and Kathleen Schroeder, data science as an industry possesses great potential for the betterment of the female experience, specifically through the use of geospatial mapping and predictive modelling to help aid organizations and government programs implement tailored developmental programs in “third world” countries. Lopes and Bailur cautiously dismantle this optimistic outlook, arguing that there is an unignorable trend of invisibility when it comes to female-focused data. Part of this issue stems from a lack of information availability – on a global scale, men have far more opportunities to access to information and communications technology, with the disparity between female and male internet users in countries like Bangladesh and Nepal clocking in at a concerning rate of over 30% (4). The problem is more extensive than a simple matter of web connectivity, however; the continued use of outdated tools that rely on biased methods of compiling geospatial information has led to massive gaps in accuracy, quality, and legitimacy between datasets specific to men and women, impeding the elimination of developmental problems with notably gendered aspects. Rather than trying to change these tools of the past, Lopes and Bailur seem convinced that recent technologies spawned by the gradual evolution of data science from a strictly statistical realm of inquiry focused on quantifying the human condition to a more fluid, contextually-based field of analysis will be influential in improving the lives of women in underdeveloped regions. The authors quote political scientist Constance Citro to emphasize this conclusion, stating that development projects “must move from [the] sample survey paradigm of the past 70 years to a mixed data sources paradigm for the future” (5). Insofar as to what a “mixed data sources paradigm” entails, Lopes and Bailur point to a 2014 study in Indonesia, where a group of data scientists used neural networks attuned to pick up on keywords across social media posts to analyze “real-time… discrimination against women in the workplace” (16). The fact that this project, which pulled data from the unfiltered byproducts of digital behavior in real-time and contextualized qualitative knowledge of the researched region’s cultural norms, succeeded in plotting socioeconomic discrimination where others did not is no coincidence, and speaks to the power of human influence in accurately mapping, modelling, and understanding gender-based data. Lopes and Bailur emphasize that this type of “situational awareness” (2) is key to producing better datasets for the use of human development. This approach can help data scientists note important progress towards the achievement of different “tiers” of SDG 5 by tracking the proportion of women subjected to spousal abuse through social media updates and mobile phone surveys (Tier 2), analyzing the discrepancies in ownership rates of agricultural land between men and women by overlapping survey data and satellite imagery (Tier 3), and using voice recognition technologies to report on women’s involvement in politics via radio programs such as C-SPAN and international equivalents (Tier 1) (7). By applying an integrated method of human and machine analysis to data collection, the gender-based gap in data collection may finally be bridged, allowing real-world disparities in women’s social, economic, physical, and mental health to be alleviated as well.
Final Word Count: 2398