An index that triggers a metric ton of annoyance

How is that possible? Well, I’m also interested in this GHI. Because this, and other such indices that sometimes make the news, reflect an Indian reality that anyone in this country can see with an eye-opener. The reality is no less than world class highways or the second richest person on the planet. Given this, why Tasty’s reaction to GHI?

For example, the Indian government has an October 15 press release titled “Global Hunger Report 2022 – Index is an inaccurate measure of hunger and suffers from serious methodological issues” (https://bit.ly/3yVWaiB). Consider what it contains. Part of the index, it states, is “based on an opinion poll conducted on a very small sample size of 3,000 … a small [sic] sample for a country the size of India.” Indeed, how can a sample of 3,000 accurately reflect a nation of 1.4 billion people?

This seems like a very serious objection. Except that the statisticians doing the survey know it doesn’t make sense. Why so? Population sampling is a well understood statistical exercise. There is a sample size involved, but also the level of error and confidence in your results that you are comfortable with. As you can imagine, these three numbers are related by a formula. For example, you might be comfortable with an error margin of 4% and a confidence level of 95%—these are typical numbers for pre-election opinion polls. Let’s say you cast a poll like this and find that 60% of your sample prefers candidate B over candidate T. This means that if you cast the same ballot 20 times with the same sample size, in 19 of those polls (95% confidence level) you would find that between 56% of the sample and 64% (60% +/- 4. % margin of error) between B is preferred over T.

If it’s confusing, the thing is, you can be pretty confident about those results – because the 95% and 4% figures. But note: I didn’t mention sample size. For those error and confidence figures, the formula says you need a sample of about 600 people. not anymore. But note: I didn’t even mention the size of the entire population. The truth is that it doesn’t matter. When you have a large population, increasing your sample size beyond a certain size—perhaps counter-intuitively—makes only a slight difference to your results.

For example, let’s say you are dissatisfied with the 4% margin of error and want to reduce it to 3% – which is better than most electoral opinion polls. This would require you to increase your sample from 600 to just 1,000 people. And what about GHI’s sample size of 3,000? With that number, and sticking to a confidence level of 95%, the formula gives us a margin of error of 1.8%—far more accurate than most opinion polls claim.

But note again that India’s population of 1.4 billion is not included in these calculations. That number doesn’t have much effect on the accuracy of opinion polls.

The conclusion of all this? Criticizing the GHI for its sample size of 3,000 is, at best, misinformed and misleading. Government statisticians should have known better.

Another objection to the GHI in the press release is as follows: Starting in March 2020 – the outbreak of the pandemic – India has run the “world’s largest food security programme”. People, and “has been extended to December 2022.” The idea is that “the poor, needy and vulnerable families/beneficiaries do not suffer due to non-availability of adequate food grains in times of economic distress.”

To underline how widespread this practice is, the press release offers a small influx of other numbers. For example, we are told that the program distributed 112.1 million metric tons (metric tons) of food grains, which is a . Is equal to 3.91 trillion subsidy. (Also: those two particular numbers show that the cost of one kilo of food grains falls on the government 35).

The flood of numbers continues. Through the pandemic, “supplementary nutrition” of 5.31 million tons of food grains was provided to approximately 77.1 million children … and 17.8 million pregnant women and lactating mothers. Break down those figures to find that each of these people (children, pregnant and lactating women) was distributed about 56 kg of grain in the 2.5 years after the pandemic began. That is, about 2 kg per person per month. Make of that what you will.

The question, however, is whether such large numbers indicate that the Global Hunger Index is flawed? That is, if India runs the world’s largest food security program, giving free food to nearly 60% of Indians because they can’t afford it—does it essentially comply, the press release said. , “There is absolutely no reason why the level of malnutrition of the country should increase”? That we can’t rank so low on the GHI?

No, because both have nothing to do with each other. The GHI is based on specific malnutrition measures. Food distribution over 2.5 years doesn’t necessarily change—or at least, we don’t have any data that shows such a change. Furthermore, the size of the program, in itself, has no effect on those measures. It’s so big because we have the world’s largest population, and certainly the world’s largest population needs food security.

In fact, I leave you outlining that last thought. Yes, we have the largest food security program in the world. But what does it say about India around 2022 that 80 crore Indians—nearly 60% of us—are so poor that they need five kilos of free food grains?

Dilip D’Souza, once a computer scientist, now lives in Mumbai and writes for his dinner. His Twitter handle is @DeathEndsFun.