Outstanding teardown of how disposable income thresholds get weaponized as poverty metrics. The 60% of median threshold treats inequality and deprivation as interchangeable when they're completely different phenomena, a kid at 59% of median in Denmark has way better material conditions than one at 61% in Romania but only one gets counted as poor. I work in policy research and the conflation drives me nuts because it makes cross-country comparisons useless. The severe material deprivation data showing Spain at 13th instead of 36th is the real story,but nobody quotes that becuase "Spain not actually terrible" doesnt get clicks.
I did not know how UNICEF was defining childhood poverty, but now that I do know I agree it is very misleading and seems statistically illiterate. It reminds me of an administrator I once worked with who wanted employees to be disciplined whose performance scores were below the median score, since having such a score meant that they were worse than the average employee. But of course 50% of employees will be in the lowest 50%--that's what 'median' means*. (*While not to be confused with "what 'means' means," or "what 'having means' means," as the UNICEF authors also seem to have done.)
Relatedly, some countries are going to score worse than other countries on a ranking of childhood poverty--no matter how accurately childhood poverty is measured--because that's what a ranking is. But that doesn't mean that the worse-scoring countries are necessarily doing a bad job--especially when we are already looking at relatively well-off countries. We'd do well to reflect on the massive progress that's shown by poverty markers including "not going on holiday once a year" and "not having a car for personal use" rather than "having a BMI <17" or "not having access to potable water" or "not being vaccinated against any disease."
Finally, so much respect for your defense of the Canary Islands from doom-mongering analysts with moving goalposts. I feel like you and I are on opposite sides of the Atlantic, on different islands, but still fighting the same fight of using statistics correctly to stand up for the distinct values of our homes.
> But that doesn't mean that the worse-scoring countries are necessarily doing a bad job--especially when we are already looking at relatively well-off countries.
It also annoys me that by publicizing this flawed measure, international organizations miss the opportunity to promote standardized multidimensional measures (focusing on material conditions, but also education and health), such as the MPI: https://ourworldindata.org/grapher/multidimensional-poverty-index-mpi
Instead, UNECE (UN Economic Commission for Europe) recommends the use of national multidimensional measures of poverty:
"Recommendation 27: Following the SDGs, countries should implement multidimensional poverty measures to complement existing monetary measures of poverty. These multidimensional measures should be tailored to the national context and policy priorities, and be tracked over time." (https://unece.org/sites/default/files/2021-01/ECECESSTAT20204.pdf).
UPDATE: Here's a multidimensional poverty index that actually permits comparison between countries, produced by OPHI:
Outstanding teardown of how disposable income thresholds get weaponized as poverty metrics. The 60% of median threshold treats inequality and deprivation as interchangeable when they're completely different phenomena, a kid at 59% of median in Denmark has way better material conditions than one at 61% in Romania but only one gets counted as poor. I work in policy research and the conflation drives me nuts because it makes cross-country comparisons useless. The severe material deprivation data showing Spain at 13th instead of 36th is the real story,but nobody quotes that becuase "Spain not actually terrible" doesnt get clicks.
> a kid at 59% of median in Denmark has way better material conditions than one at 61% in Romania but only one gets counted as poor.
Exactly. Couldn't have said it better.
> but nobody quotes that becuase "Spain not actually terrible" doesnt get clicks.
I think that's the most straightforward explanation.
Very nice research and analysis, Javiero!
I did not know how UNICEF was defining childhood poverty, but now that I do know I agree it is very misleading and seems statistically illiterate. It reminds me of an administrator I once worked with who wanted employees to be disciplined whose performance scores were below the median score, since having such a score meant that they were worse than the average employee. But of course 50% of employees will be in the lowest 50%--that's what 'median' means*. (*While not to be confused with "what 'means' means," or "what 'having means' means," as the UNICEF authors also seem to have done.)
Relatedly, some countries are going to score worse than other countries on a ranking of childhood poverty--no matter how accurately childhood poverty is measured--because that's what a ranking is. But that doesn't mean that the worse-scoring countries are necessarily doing a bad job--especially when we are already looking at relatively well-off countries. We'd do well to reflect on the massive progress that's shown by poverty markers including "not going on holiday once a year" and "not having a car for personal use" rather than "having a BMI <17" or "not having access to potable water" or "not being vaccinated against any disease."
Finally, so much respect for your defense of the Canary Islands from doom-mongering analysts with moving goalposts. I feel like you and I are on opposite sides of the Atlantic, on different islands, but still fighting the same fight of using statistics correctly to stand up for the distinct values of our homes.
> But that doesn't mean that the worse-scoring countries are necessarily doing a bad job--especially when we are already looking at relatively well-off countries.
It also annoys me that by publicizing this flawed measure, international organizations miss the opportunity to promote standardized multidimensional measures (focusing on material conditions, but also education and health), such as the MPI: https://ourworldindata.org/grapher/multidimensional-poverty-index-mpi
Instead, UNECE (UN Economic Commission for Europe) recommends the use of national multidimensional measures of poverty:
"Recommendation 27: Following the SDGs, countries should implement multidimensional poverty measures to complement existing monetary measures of poverty. These multidimensional measures should be tailored to the national context and policy priorities, and be tracked over time." (https://unece.org/sites/default/files/2021-01/ECECESSTAT20204.pdf).
UPDATE: Here's a multidimensional poverty index that actually permits comparison between countries, produced by OPHI:
- Report: https://ophi.org.uk/sites/default/files/2025-11/B62_Child_poverty_2025_3.pdf
- Table 3 Age Results MPI 2025: https://ophi.org.uk/global-mpi/2025