Instead of fruitlessly trying to find the poor, the government can easily make a list of the rich
Although we know the poor are starving in Bangladesh, we are struggling to get relief to their doors because we don’t really know who or where they are. In a previous article (Here’s how the PM can stop the rice thieves) published in Dhaka Tribune I argued it is impossible to make an accurate database of the poor in our country where the vast majority work in the informal sector. So I advocated moving to short-term emergency universal food rations for everyone, abandoning pointless attempts at poverty targeting.
Developments in the past week though show targeting may be non-negotiable for the government. Therefore, in this article, I argue that instead of making a list of the poor, the government will be more successful if it makes a list of the rich.
The government has tried really hard to find the poor
For the past seven years, the Bangladesh Bureau of Statistics has tried to make a National Household Database (NHD), which the World Bank called the country’s first “poverty registry”. Even though the project’s original Tk328 cr budget has bloomed to Tk727 cr, and the deadline was pushed from 2017 to 2021, the database is incomplete. Whilst BBS’ mediocrity is partially to blame, the real problem is the project is actually trying to do the impossible.
The poor, as I pointed out in my last article, are a dynamic group. People move into and out of poverty for all sorts of reasons, including income and health shocks. As a result, any poverty registry is rendered obsolete very soon after it’s made.
Moreover, the NHD used Proxy Means Test (PMT) to identify the poor. This is because due to high informality, we don’t have people’s salary data to verify who’s poor. So, across the entire country, PMT uses proxies—such as housing type, presence of TV, owning farm animals, etc—to try to guess how rich or poor a household is. You can imagine how arbitrary, time consuming and inaccurate this exercise is! “You can do this stuff for limited research purposes, but not to actually administer tax payer funds,” insisted a senior Bangladeshi civil servant I interviewed for my PhD. Globally, errors in PMT range between 50%-93%, i.e. usually PMT gets it wrong.
So, it’s no wonder the poverty registry flopped. Sadly, instead of learning a lesson from this failure, policymakers have doubled down on targeting since the pandemic hit us. The reason the rice is being misappropriated, they maintain, is because we don’t have the right database. If only we did, the rice would reach the right people. It is unclear though why simply having an accurate list would make the rice thieves deliver rice to the right door any more diligently than having the wrong list.
Forget the poor, find the rich
Forced to halt the OMS rice sales due to widespread looting, the PM wisely announced 5 million new ration cards for the poor. A commendable and timely move, it can help many people. Unfortunately, subsequent statements indicate these cards will be distributed based on a new list of the poor that party functionaries will hurriedly create now. Given we failed to make this list after trying for the last seven years, I have grave doubts we will suddenly pull off this miracle in the next few weeks.
In my previous article, I outlined the causal mechanism by which I believe giving emergency food rations to everyone will help the PM tackle the rice thieves and get desperately needed food to the hungry. To counter the myth that we don’t have enough for all, I also discussed how we can source this rice.
If this is totally unacceptable to policymakers though, instead of fruitlessly trying to find the poor, the government can easily make a list of the rich. So instead of giving ‘only to the poor’, the government will give to everyone ‘except for the rich’. Called affluence testing, it is used by countries from UK-Canada to South Africa-Mongolia.
Affluence testing and community surveillance
Affluence testing is very useful because the rich in a country are far easier to identify than the poor. You don’t have to investigate if someone owns a Tesla or what their mobile phone behaviour is, the government already has the necessary information in income tax files, bank records, property registration documents, etc. Just put it together, and you immediately know who to exclude from receiving relief. Whilst UNICEF has previously used “a maximum income/resource threshold of more than 200% of the national minimum wage” to identify the rich, the government can set whatever level it feels appropriate now. Whoever comes with those National ID cards will be turned away whilst everyone else gets rice rations.
But not everyone has NID, right? That’s true, there are many people in the floating population who don’t. But remember, the rich do not line up for rations; if someone’s actually standing in those long lines under the sun, consider them self-reporting their poverty.
Furthermore, to make the system more robust, the government can help improve community surveillance by announcing, perhaps via TV, how much rice they’ve sent to which locality. Alongside, just like it sends lists of the poor to the localities, it can instead send lists of the rich (to make everyone happy, call the list “Friends of the people”). This twin strategy can make it easier for communities to monitor if rice is being misappropriated. It will allow us to face the crisis as a society, where each citizen is an active agent, instead of a passive recipient.
Will affluence testing stop the rice thieves totally? I doubt it. But neither will identifying the poor, which is what the government is trying to do now. In contrast to poverty targeting, affluence testing is at least faster and more efficient. Given the twin pressures of hunger and Covid-19, time is of the essence. Also, I do not think making a list of the rich based on existing records will disincentivise declaring wealth in the future – normal rules of the game are suspended in this emergency.
Worry about who we’ve left out, not who we’ve mistakenly let in
Recent studies show there has been a catastrophic loss in income across most groups. So I echo Dr Amartya Sen, Dr Martin Ravallion, Dr Abhijit Banerjee and a host of other distinguished scholars when I urge the government to worry less about inclusion errors and focus instead on exclusion errors. There is no point trying to find 10 million of the ‘most needy’ to give to – right now, to survive, we must expand social protection both horizontally and vertically, i.e. we must give more to more people. We don’t need to be concerned some of the undeserving will get some rations, most people are deserving right now and the fatal problem is we are missing the deserving.
I still hold that universal food rations, at best geographically targeted, are the most feasible solution given current constraints. Nevertheless, in this article, I have proposed if we can’t accept that solution, we must at least stop trying to find who and where the poor are. If we haven’t succeeded in this task after trying for seven years, we won’t succeed now. Instead, let’s find out who the rich are and give relief to everyone else besides them.
Nabila Idris is a PhD student at the University of Cambridge where she researches Bangladesh’s social protection policymaking.