[TLDR (too long didn’t read): If you are reading this, chances are it behooves you.This HR Reader collects assorted critiques of why statistics can and do neglect HR issues –thus HR missing a potential ally’s help in its struggle. I confess I have not here given it some of the praise it may deserve otherwise. For a quick overview, just read the bolded text]. Traducir/traduire los/les Readers; usar/utiliser deepl.com
—Statistics should reflect reality –not political agendas.
1. We live in a world where misinformation spreads rapidly and data manipulation is easier than ever. Public statistical institutions are weakened by chronic underfunding and political interference. Without independence and adequate resources, the integrity of economic and other data they generate –and the decisions made with these data (if any…) — is compromised. Apart from manipulation*, neglect of statistical quality can have serious consequences. They can also produce misleading or entirely fabricated results: nowhere more important than in the realm of human rights (HR). Numbers can beclose but incorrect, which is arguably more dangerous than being wildly wrong (plausible errors are harder to detect and more likely to mislead). Only commitment to transparency builds trust. Legal frameworks must protect these public statistical institutions from external interference allowing them to choose methodologies and publish findings based purely on professional judgment. (Bert Kroese)
*: Paraphrasing Stalin, one citizen without a job is a drama, two million unemployed are a statistic; …and statistics are made by us.
[Note I do not delve here into the realm of polling activities that also have great potential for malign distortions].
Accurate and timely data are not just a bunch of metrics –they are a lifeline. But gaps are real and so is the urgency to clean up the act
—Data analyses in our times must depart from current realities rather than from abstract, machine-generated, ahistorical pre-conceptions. (Jomo Sundaram)
2. A focus on documentation and verification with attention given to turning professionally arrived-at statistics into advocacy tools is supposed to uncover the deeper patterns of reality. But quantifying harm(or potential harm), does not, by itself,help to bring about change in better decision-making by duty bearers. Statistics or other information just simply does not bring about positive change when vested interests are at the core of decision-making. Only in counted cases, can analytical data eventually transform HR practice, for example, helping to build legal cases, providing powerful support for policy change or broadening public engagement with community-generated HR narratives. These data could, but seldom do, estimate the true scale of HR violations, even when direct evidence is for everybody to see. Instead, only partial data are presented if not downrightly suppressed even when this information can be crucial for establishing the magnitude of harm in a court of justice.
3. The promise data scientists try to sell to us comes with profound challenges. Human rights practitioners must navigate a landscape pockmarked by ethical dilemmas and practical constraints. The ethics of using digital evidence must address the increased risk of profiling victims and communities, exposing them to surveillance, retaliation, or loss of privacy. Universities are uniquely positioned to facilitate assessments, deep analyses and the development of practical tools that HR activists can indeed adopt and apply.
4. The integration of data science into legal case-building is not a panacea—technology cannot replace the fundamental work of listening to communities, building trust, and advocating for systemic change! Only when used thoughtfully and responsibly, can statistics and data accelerate our ability to document, prove, and remedy violations of international law. (Laurel Fletcher et al)
5. Bottom line here, we should also note that replacing human tragedies with statistics also contributes to the new normal. Human life is a quality, while the number of lives or deaths is a quantity. But in this case, the key is to have the power to prevent quantity from becoming a new quality. (Boaventura de Sousa Santos)
Let us face it: Human rights are not in the statistical Richter Scale
–The reality of HR violations is often not captured by statistics.
6. Data do not speak truth by themselves. Data are not neutral: they reflect power. When analyzed through a rights-based lens, statistics rarely expose who profits from harmful economic policies and who pays the price.** (CESR) Yes, ‘smoking’ reality is a major devious use of statistics. Yes, we thus need a systematic monitoring of development statistics: not tomorrow, but today since the pressure to produce reliable HR statistics is amply warranted. [Beware: rich countries tend to perpetuate the myth that statistical expertise is the prerogative of the few (…in the North)].
**: We have good reasons to be skeptical about statistics; after all, it is notable that the 17th century term for what is now called statistics was ‘political arithmetick’. (Richard and Patrice Jelliffe)
7. Ah! and also: Please forget about the 3rd and 4th decimal points of statistics and examine the whole numbers of the issues that matter. (Alan Berg) Too often over-detailed statistics are used to distract us from the big picture.
More risks
— Population statistics show us demographic trend lines that can distort the true situation of HR of claim holders. (Stuart Gillespie)
8. Statistics, I think, contribute to the casual way in which they airbrush–out the realities of so many people. Statistics too often fall back on population averages and trend lines on a graph. They fail to put the focus on those left behind*** –or rather, actively held back– by structures and systems that are built to maximize profits for a few and leave out ensuring equity. These ‘average statistics’ imply that, since the average shows progress,we just need to stay on the same path, i.e., the path that has brought us to this implied improvement in human progress. “We just need to continue doing what we are doing, but do it faster…” meaning we do not really need to change what we do —ergo, incrementalism not transformation; …accelerated business-as-usual leaving the-you-know-who behind.
***: Perhaps not facetiously, the average is not the temperature in the belly button when the head is in the oven and the feet are in the refrigerator.
9. I just simply do not buy the above. I rather think that much data presented by (some) statistics are almost designed to dampen down any discussions of the–politics-of-it-all. This skews the political discourse in ways that lead to false narratives.**** We are then bound to have a trust problem.***** A trust deficit is thus justified. [If there is a trust deficit, why not address it?!]. Beware: challenging such ‘feel good’ statistics is not ideological. Our argument is fed by evidence and by experience of the real world. Human rights ultimately require using a proper diagnosis of the problem as a springboard into transformative change –this is the foundational point here. (S. Gillespie)
****: This also often leads to questions about optimism/pessimism like –do you see a glass that is half-full or half-empty? What if you see a glass that is cracked and leaking? Maybe we need a new glass…
*****: Trust reminds me a bit of virginity: once it is lost, it is over. (Louis Casado).
10. Many diagnostic narratives expressed in numbers are out of joint, because the story begins in the middle (or the end) with symptoms –not at the beginning with the underlying, structural causes. To figure the solution, we need to dig deeper into structures and systems. We cannot have a revolution without reliable revelations. (Rupa Marya, Raj Patel)
Claudio Schuftan, Ho Chi Minh City
Your comments are welcome at schuftan@gmail.com
Postscript/Marginalia
Here a bunch of my iron laws pertaining to this topic:
—Is the UN a temple of trustworthy statistics? (just asking…)
- In statistics, the perfect is the enemy of the good.
- In times of widespread deception, the mere fact of statistics telling the truth is revolutionary. (George Orwell)
- John Mauldin (well-known financial expert and New York Times best-selling author) calls statisticians (and economists) ‘meteorologists’, given the imprecisions of their predictions.
- ‘Speaking truth to power’ in the context of statistics is about bringing HR data and information depicting violations and abuses of HR clearly and unambiguously to HR activists for them to bring it to those in power so as to demand change. (adapted from Amnesty International)
- Decision makers are often bureaucrats at the service of power; some call them red tape merchants; they ignore the conflicts of interest in gathering and presenting statistics and thus become complicit. (adapted from S. Gillespie)
- Specific HR statistical information collection is hard to fund; the other, ‘faceless’ information, is easier to fund –or to buy. (adapted from P. Ackerman-Leist)
- Statistics ought to be a champion science of ‘do no harm’. But, since science and politics are two different realms, the logic of data and processes of collecting them are quite different: data can thus do harm.
- Cold statistics (statistics without a heart) serve not like wise future-builders, they notoriously leave out ethical and HR issues of such a future.
- In the search of accuracy, triangulation is often used in cold statistics. Does this then include indicators of HR? I do not think so. (In the context of HR, triangulation ought to refer to always crosscheck the information with at least two other independent sources in order to verify its accuracy –and this may be by using ‘warm’ (often qualitative) data. (Amnesty) ‘New air’ is only found if one looks for it, and
- There are statistical data that are prepared-for and used-by politicians’ speeches and statements that, if not distorted, often are somewhat empty of content or out of context (rings a bell?).