{"id":378753,"date":"2023-09-29T18:00:00","date_gmt":"2023-09-30T00:00:00","guid":{"rendered":"http:\/\/freerangestats.info\/blog\/2023\/09\/30\/mvi"},"modified":"2023-09-29T18:00:00","modified_gmt":"2023-09-30T00:00:00","slug":"the-uns-proposed-multidimensional-vulnerability-index-by-ellis2013nz","status":"publish","type":"post","link":"https:\/\/www.r-bloggers.com\/2023\/09\/the-uns-proposed-multidimensional-vulnerability-index-by-ellis2013nz\/","title":{"rendered":"The UN’s proposed Multidimensional Vulnerability Index by @ellis2013nz"},"content":{"rendered":"\r\n\r\n
\r\n[This article was first published on free range statistics - R<\/a><\/strong>, and kindly contributed to R-bloggers<\/a>]. (You can report issue about the content on this page here<\/a>)\r\n
Want to share your content on R-bloggers? click here<\/a> if you have a blog, or here<\/a> if you don't.\r\n<\/div>\n

The United Nations President of the General Assembly\u2019s High-Level Panel<\/a> on the development of a Multidimensional Vulnerability Index have (September 2023) released an Advanced Unedited Version of their Final Report<\/a>, and along with it the scores on that index for 142 countries. This is relevant for my work; it\u2019s pretty complex; and I wanted to understand it fully so I downloaded it and drew some charts.<\/p>\n\n

A good measure of \u201cvulnerability\u201d has been sought for at least 30 years as a supplement to measures of economic prosperity, with a number of parties (particularly small island developing states) advocating for its inclusion in eligibility decisions for financial and other forms of assistance. Vulnerability has formally been a factor in determining Least Developed Country status for some time now. This High Level Panel was commissioned<\/a> to resolve the contradictions in the various measures out there and address the \u201clack of international consensus on how to characterize and, by extension, mitigate their (small states\u2019) vulnerabilities\u201d.<\/p>\n\n

There is (I think) an awkward contradiction between acknowledging the multidimensional nature of vulnerability and the desire for a single index that can be used for financing decisions. The High Level Panel deal with this by creating an elegant hierarchy where increasingly abstract indexes are created out of other indexes, which themselves are made out of actual indicators that have been scaled from 0 to 100, with 100 meaning the most vulnerable (or least resilient). All the details are in the report linked to above.<\/p>\n\n

Numbers are combined to form the index in the next level up of the hierarchy by taking the square root of the mean of the squares of the original numbers, which has the effect of pushing the average a bit towards the higher numbers in the inputs; this is acknowledgement of the common judgement that you don\u2019t need to be vulnerable against all the inputs to be vulnerable overall. Consider that the arithmetic mean of (2, 6) is 4; the geometric mean is 2.8 (pushed downwards); but the square root of the mean of the squares is 4.5 (pushed upwards). If that \u201c6\u201d meant you were vulnerable to something and the \u201c2\u201d you weren\u2019t to something else, using 4.5 as your combined measure is giving a bit more importance to the variable that you are vulnerable to.<\/p>\n\n

Visualising the results<\/h2>\n\n

First lets look at some results for the Pacific, the area I\u2019m working on in my day job. Here are the values on the final, single Multidimensional Vulnerability Index (MVI) for the 12 Pacific Island countries in the dataset, compared to other countries in grey:<\/p>\n\n<\/object>\n\n

Unlike most of my posts, I haven\u2019t interspersed my text and output with code creating it, because I want the results to be easy to focus on without distraction in talking about how the visualisations are done. But that code is still available in one spot here<\/a>.<\/p>\n\n

A few surprises in that first plot, for example that Vanuatu (often used as a case study of a vulnerable country, that has had a bad run recently with cyclones) is only very slightly above world median vulnerability.<\/p>\n\n

Next level down from the main index we have two important component indexes – \u201cStructural Vulnerability\u201d and \u201cLack of Structural Resilience\u201d. The final index for each country is their average of scores on these two, so let\u2019s look again at the Pacific countries on the two indexes at once, again with the rest of the world\u2019s countries shown in grey:<\/p>\n\n<\/object>\n\n

Hmm ok, that\u2019s interesting; for example we can see that Tuvalu gets it\u2019s relatively high overall score (that we saw in the first plot where it was ranked second most vulnerable of Pacific countries) almost entirely from \u201clack of structural resilience\u201d and in contrast is actually slightly below the median score for \u201cStructural vulnerability\u201d.<\/p>\n\n

The next two charts let us see what the concepts are behind these two indexes.<\/p>\n\n

First, here\u2019s \u201cstructural vulnerability\u201d. There are two levels of concepts I\u2019ve merged into one chart here; the first three (economic, environmental and social vulnerability) are themselves made up of the more detailed remaining nine. We\u2019re still dealing with abstracts, not down to any actual data here. But at least we can see what it is that conceptually is leading to high or low scores. In the case of Nauru for example we see a particularly high score for \u201cexposure to entrance of international forced displacement of people\u201d. Hmm, there\u2019s some complicated history behind that one.<\/p>\n\n<\/object>\n\n

And here\u2019s the concepts making up \u201clack of structural resilience\u201d:<\/p>\n\n<\/object>\n\n

We can see that these \u201cresilience\u201d concepts are about the ability of a society to respond to shocks they will get as a result of vulnerability; for example the lack economies of scale, adequate water supply, or a resilient agricultural system.<\/p>\n\n

Finally, one more level down, here are the actual measureable indicators that are turned into those concepts, and ultimately the index itself:<\/p>\n\n<\/object>\n\n<\/object>\n\n

Visualising the hierarchy of concepts<\/h2>\n\n

But which variables feed in to which concepts, and how to grasp this all at once? To help me understand how the variables get aggregated into concepts, then into higher level concepts, then into two indexes and finally one index, I constructed a network graph to illustrate the hierarchy:<\/p>\n\n

<\/a><\/p>\n\n

This visualisation works for me – at least on a reasonable sized screen. But this, and all of the following network charts, are a bit fiddly to read in-line in the blog so I\u2019d recommend opening them in another tab (with right mouse button) or clicking on them to navigate straight there (and then you can use the back key to come back here to the main text).<\/p>\n\n

This layout suggests a nice option – we can colour-code the nodes (representing variables, concepts or index scores) according to an actual country\u2019s scores, and see what\u2019s pushing the country\u2019s vulnerability scores up and down and how they add up to the overall score. Here\u2019s such visualisations for each of our 12 Pacific countries, grouped by subregion:<\/p>\n\n

Melanesia<\/h3>\n

Papua New Guinea<\/h4>\n

<\/a><\/p>\n\n

Most vulnerable scores (bright yellow) for PNG coming from gender equity, and lack of access to sanitation. Non-vulnerable scores coming from no problem with drylands, coastal zones, and access to freshwater.<\/p>\n\n

Fiji<\/h4>\n

<\/a><\/p>\n\n

Fiji scores higher for vulnerability on exposure to natural hazards, rainfall shocks, low connectivity and low integration wutih international markets.<\/p>\n\n

Vanuatu<\/h4>\n

<\/a><\/p>\n\n

Vanuatu also scores high on vulnerability to natural hazards, low connectivity, and gender equity. But it is scored with a fairly high level of enviornmental resilience, and not too socially vulnerable.<\/p>\n\n

\u2026 and so on. I won\u2019t write commentary on each individual country, but for completion they are all here:<\/p>\n\n

Solomon Islands<\/h4>\n

<\/a><\/p>\n\n

Micronesia<\/h3>\n

Federated States of Micronesia<\/h4>\n

<\/a><\/p>\n\n

Marshall Islands<\/h4>\n

<\/a><\/p>\n\n

Kiribati<\/h4>\n

<\/a><\/p>\n\n

Tuvalu<\/h4>\n

<\/a><\/p>\n\n

Nauru<\/h4>\n

<\/a><\/p>\n\n

Palau<\/h4>\n

<\/a><\/p>\n\n

Polynesia<\/h3>\n

Samoa<\/h4>\n

<\/a><\/p>\n\n

Tonga<\/h4>\n

<\/a><\/p>\n\n

OK, that was interesting. I certainly feel I understand both a) how the index is constructed and b) the ratings it is giving to Pacific Island countries.<\/p>\n\n

\r\n
To leave a comment<\/strong> for the author, please follow the link and comment on their blog: free range statistics - R<\/a><\/strong>.<\/div>\r\n
\r\nR-bloggers.com<\/a> offers daily e-mail updates<\/a><\/strong> about R<\/a> news and tutorials about learning R<\/a> and many other topics. Click here if you're looking to post or find an R\/data-science job<\/a>.\r\n\r\n
Want to share your content on R-bloggers? click here<\/a> if you have a blog, or here<\/a> if you don't.\r\n<\/div>","protected":false},"excerpt":{"rendered":"

The United Nations President of the General Assembly\u2019s High-Level Panel on the development of a Multidimensional Vulnerability Index have (September 2023) released an Advanced Unedited Version of their Final Report, and along with it the scores on that…<\/p>\n","protected":false},"author":1752,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[4],"tags":[],"aioseo_notices":[],"jetpack-related-posts":[],"amp_enabled":true,"_links":{"self":[{"href":"https:\/\/www.r-bloggers.com\/wp-json\/wp\/v2\/posts\/378753"}],"collection":[{"href":"https:\/\/www.r-bloggers.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.r-bloggers.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.r-bloggers.com\/wp-json\/wp\/v2\/users\/1752"}],"replies":[{"embeddable":true,"href":"https:\/\/www.r-bloggers.com\/wp-json\/wp\/v2\/comments?post=378753"}],"version-history":[{"count":2,"href":"https:\/\/www.r-bloggers.com\/wp-json\/wp\/v2\/posts\/378753\/revisions"}],"predecessor-version":[{"id":378754,"href":"https:\/\/www.r-bloggers.com\/wp-json\/wp\/v2\/posts\/378753\/revisions\/378754"}],"wp:attachment":[{"href":"https:\/\/www.r-bloggers.com\/wp-json\/wp\/v2\/media?parent=378753"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.r-bloggers.com\/wp-json\/wp\/v2\/categories?post=378753"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.r-bloggers.com\/wp-json\/wp\/v2\/tags?post=378753"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}