Your two quotes are referring to two different survey counts. Let me explain how this works: it’s impossible to get an exact count down to every individual, you survey a population sample and then make estimates based on your population sample. This can be done by both government and private agencies.

Here are the two different surveys from the article broken down:

The NSSO survey, done by the Indian government, not officially released, but leaked to the media. All this article tells us about the results of this survey is that “poverty had increased”. No years, no numbers, no percentages. Just three words.

The CMIE survey, done by a private company, used by the World Bank. We at least have the numbers for this survey, which says poverty has more than halved from 2011 to 2019. However, the World Bank already assumed that poverty would be decreasing rapidly, and put out global poverty estimations based on rapidly decreasing poverty before

When you quoted the World Bank as saying

The global poverty headcount in 2018 is revised slightly up from 8.7 to 8.9 per cent

That was referring to CMIE survey results being less optimistic than original World Bank projections.

I did make a mistake, I didn’t credit you for the NSSO survey, which did indeed support your point. I think that’s why you misread the conclusion of the CMIE survey. I think it’s sus af the government won’t release it, but I’m also a scientist. All we have from it is, “poverty had increased”. Imagine I presented to you a new article that said nothing about poverty rates in China except “poverty had increased”, how much would you value that piece of information?