wow-inequalities/docs/meeting_2023-11-02.md

Meeting notes 2023-09-15

Questions/issues

Main points: - Currently, two issues of search query: - the scope of results is still very large (~5000 query results & ~1000 sources from snowballing) - is overwhelmingly health-inequality outcome focused

Q: - One of ILO raised points was including Spanish/French studies - I would keep query in English only but also try to screen Spanish/French studies that are in results - I am having some trouble distinguishing between relevant / non-relevant for health outcome category - e.g. policies to prevent STD transmission for sex workers -> probably in since it directly addresses an area of outcome inequality; and focuses on intersectional issues for cis men and trans workers - e.g. influence of socio-economic inequalities to out of hospital medical issues -> probably out since it does not fit into world of work though still focusing on issues of inequality - in reviewing methodology, ILO commented on 'LM-adjacent outcomes' not necessary for inclusion criteria; I am not sure how to understand that comment - we need to measure some outcomes through our indicators and it should be LM adjacent to fix it somewhat to the WoW? - I can not find the commented version of the extended outline anymore, is it still in DB?

Updates

• have been executing the search protocol and started screening results
  • identified around 30 previous reviews which focus on nexus of inequality, world of work and inequality-outcomes

Indicator

[11:41 AM] Miguel Nino Zarazua **PLEASE INCLUDE THE FOLLOWING INEQUALITY MEASURES IN THE DISCUSSION Regarding inequality measures, we need to discuss inequality indices and ratios/shares by families and then group them according to their normative theoretical principles, that is, relative, absolute, or horizontal perspectives. Vertical approaches

1. Gini family (Gini index - a relative measure, AND Absolute Gini index-an absolute measure)
2. Generalized Entropy family (Mean Log Deviation, Theil index, Coefficient of Variation - all relative inequality measures, AND Standard Deviation - an absolute inequality measure.) Please make sure that we when revising the literature, we do not miss any relevant measure.
3. Atkinson Inequality Family
4. Ratios / Top/Bot family (Palma, s80s20 ratio, s40s20 ratio)

As I pointed out earlier, presenting a discussion on the properties of inequality measures is vital because we will need to discuss the implications of focusing on e.g., the Gini using wages or earnings vis-a-vis using the absolute Gini on the same outcomes. Furthermore, we also need to discuss the desirable properties that inequality measures need to satisfy (anonymity, scale independence, population independence, transfer principle, subgroup decomposability, etc.) and the implications of the limitations of some of the indices (e.g., the Gini index is not decomposable by population subgroups, which means that it cannot be used to estimate e.g. earnings inequalities by subnational levels. In this case, the MLD would be a better choice. Horizontal approaches Furthermore, we need to discuss the normative and policy implications of adopting horizontal inequality measures or ratios that compare differences in outcome between groups: Group Gini Group Theil Group CoV Ratios and shares by groups (for example looking at differences between labour force participation by men and women, or by age cohorts). In this sense, the discussion on inequality measures needs to go beyond relative measures and present a discussion in the context of variables and indicators related to the WoW. PLEASE conduct a review of the literature on inequality measures to carry out this task. I have included some papers in the shared Dropbox folder BUT please note that these are not exhausting. Please conduct a review and add ALL the consulted materials in the shared folder and in the Zotero Library. For the purpose of the analysis, the identification of variables in the ILO micro dataset (and ILOSTAT) that capture forms of work AND labour market outcomes is key, as we will need to outline how the ILO could use inequality measures/ratios on the existing variables/indicators to track and monitor progress in work and labour inequalities, AND also identify key informational gaps. Please take a look at the variables and dimensions that are available on the ILO microdata collection (ILO Microdata Explorer) and identify the following:

1. Existing indicators to estimate vertical and horizontal inequality measures
2. What other variables are currently available in main sources of information, namely: a) Population Censuses, b) Labour Force Surveys, c) Household Income and Expenditure Surveys and d) Other relevant Household Surveys could be used to conduct inequality analysis. PLEASE check questionnaires of a-d to do this.
3. Identify key variables / indicators for inequality analysis that are currently NOT available on a-d. Please pay attention to indicators that are relevant for the Wo

ToDo

• now:
  • create summary of how we are doing the review - SUN
    • WHAT we search
    • WHY we search it
    • HOW we screen for/extract from it
  • have a read through comments on outline
  • have document outline/annotated bib for ILO ready eow at the latest
  • look at new resolutions (Oct 2023)
  • include union/trade unionization/collective action in search terms
  • find way of capturing social dialogue in query?
  • [-] create 'task list' for each screening/extraction step:
    • screening:
      • add intervention/outcome/country-region/in-out-criteria (see also screening tasks)
      • add additional information: intersectional, review, method::qualitative
    • extraction:
      • build extraction grid tool (from SPPF, with above policy categorization)
• [-] querying: create way of having one source of truth for search terms; e.g. script to transfer python lists from yaml/json files to python lists and -> output functional query (into script or through cmdline)
• [-] screening:
  • build tool: contains collection of all tags; kept up-to-date for any added; contains checklist for 'tag areas' (outcome; intervention; inequality; region; actor; method; ...)
• [-] extraction:
  • have category of 'measure' extracting type/kind of measure used (Gini/Group Gini/Palma/share/measure of means)
  • focus on tripartism:
    • tag studies with "actor::collective_action" (or actor::trade_union?) tag to highlight role of collective action (agency/structural)
      • e.g. trade union talks; minimum wage; collective bargaining; ...
    • tag studies which highlight e.g. role of individual sectors with "actor::private" to highlight role of private sector/employers (structural)
    • tag studies which talk about concept of "social dialogue", "tripartism" or similar how employers/unions/gov solve issues to highlight soc dialog
      • perhaps find "historical examples which highlight this process, even if we don't include it systematically"
• [-] paper:
  • [-] change citations of older ILO resolutions to 2023 ones in-text