Hacker News Reader

Comment by mc-lovin | Hacker News Reader

This is a great article. Not because it proves a particular point (we should accept the evidence no matter where it leads) but because of the clarity of the article, and the fact that it accurately represents the academic paper: on a cursory reading of the academic paper, the randomization is truly as good as they sell it in the linked article.

I think this is potentially a big step forward in our understanding of development. It challenges the assumption that the main thing holding back development is a lack of investment in human capital (training) rather than a lack of access to physical capital.

The only issue I have is whether you can expect the same results from one-off payments, as an ongoing basic income, and whether we can expect the same results for the whole population as the group that was selected, who seem to form a very specific group.

View on HN

If I recall correctly, it has been shown that Basic Income Guarantees[1] (free cash without means testing besides citizenship) works incredibly well for both developed and developing countries. This article comes as no surprise, but is well-written and compelling which is a great boon to BIG. Despite stereotypes of poor people, basic income is used very effectively with little administrative cost (not for booze and gambling? how bizarre!): - People work more and for a higher wage (exceptions: recently pregnant mothers and students) - Domestic abuse plummets - People are unemployed LONGER, but use that time to find a job they like more (or create their own!)

This form of welfare has been supported by people on the right (Friedman, Hayek) and left (Russell). I'm really disappointed it isn't used to greater effect. Voters and tax-payers have simply no faith in those damned poor people to make the right decision it seems.

Here's a great story from in Canada, where it was referred to as Mincome [2]

BIG also works well in India, where women, in particular, benefit[3]. A further benefit is it is highly corruption resistant. There are no tests, forms, etc. If you're a citizen, you get cash. Done.

Please, anyone, show me some research that shows BIG doesn't work well. Cheers

[1] http://en.wikipedia.org/wiki/Basic_income_guarantee

[2] A town without poverty? http://www.dominionpaper.ca/articles/4100

[3] http://www.soas.ac.uk/news/newsitem84314.html

You are making very big claims without adequate references. The study shows the effects of a one-off payment, to a highly selective group (note that I'm not implying their randomization was bad, but that both treatment and control groups were very special relative to the whole population), an a particular developing country.

You may personally believe that this is part of a general phenomenon that applies to highly developed countries, to all groups of people in that country, and in the long term not just the short term, but you have presented little evidence for the claims in your first paragraph (I read all the links).

You seem to view BIG as somehow unique, when in fact it is functionally equivalent to progressive taxation plus welfare, when welfare is only conditional on income. It is also hard to argue against because it treats redistribution, which is a sliding scale, as something absolute: it suggests an (imaginary) bar for "basic" income.

I personally believe systems where welfare is primarily conditional in income, but weakly conditional on searching for work or getting training (such as Australia's) are optimal for developed countries. The government can be far more generous when it adds some strings to welfare, than when it is simply giving out money. The only issue is whether countries with different governmental frameworks can copy Australia's model.

In short, you seem to be overlooking what I consider the real interest of this study, which is its specific relevance to developing nations.

>Please, anyone, show me some research that shows BIG doesn't work well. Cheers

That's a very asymmetric way to think about the evidence for and against a particular policy.

edit: The specific problem with your evidence, is that on a such a broad topic, I would expect to see an academic review of the literature. What you provided was [1] wikipedia, [2] an interesting example with no control, and [3] something that wasn't an actual academic study, but a description of a program and it's alleged benefits without any peer-reviewed study demonstrating these benefits.

> I would expect to see an academic review of the literature. What you provided was [1] wikipedia

I expect this is why you're seeing downvotes.

If you look at that article, you will find it does not provide evidence for any of the claims in the post. That's not the fault of Wikipedia, as I explained, this is an extremely broad and complex issue, and most academics won't directly cover it because "basic income" is too vague a concept for them.

But that was the point I was making in the first place: you cannot cite three sources and then claim to have the evidence on your side, when these three sources don't provide the evidence that would be needed.

Personally, I assumed that whatever followed "If I recall correctly" would likely not be expertly sourced. I did look at that Wikipedia article-- it has 102 references, many of them offline. So hell if I know if there's evidence in there, but I'm certainly not complaining that I don't know where to read more for myself.

But like the GP-- It's something I want to believe, it's very well-credentialed, and I've seen some encouraging but inconclusive evidence. So naturally as a skeptic the main thing I'm looking for now is the strongly negative evidence, if such a thing can exist. Do you have any leads?

I don't want to get too bogged down in the details of this thread, so I'll just tell you straight up why I don't support the idea of a basic income guarantee. Hopefully doing so will indirectly address the issue of evidence that we were discussing.

The main problem is that the BIG is an unnecessary and unhelpful concept. I believe the sociology is that people like to believe in causes, rather than fine grained facts or principals, and so BIG is attractive as a cause in spite of not being a useful concept.

First, BIG confuses the issues around redistribution. Assuming welfare is only conditional on income, the sum total of taxation, welfare, BIG and others, can be thought of as a negative taxation system. BIG obscures this because it implies there is something special about redistribution through a "guaranteed" income. E.g. economists are very concerned about the incentive effects of welfare, because every dollar you earn results in less welfare. BIG obscures this by focusing on the flat portion (the guaranteed income) and drawing attention away from the more progressive income taxes that would be needed to pay for it.

Second, BIG is different from many current welfare systems because it is unconditional on being part of a special group (e.g. a parent, farmer, elderly) or certain actions (e.g. education, training, looking for work in good faith). I consider this to be a bad thing, in that it is too extreme in the direction of being unconditional, at least for the developed world. Both in terms of ethical fairness, political feasibility, and economic incentives, giving away money for nothing has strong limits on it. Adding a few conditions, like ensuring people are really looking for work, strikes a much better balance. So again, BIG takes a sliding scale decision (how much should welfare be conditional on certain actions, and should it be conditional on things other than income) and makes it seem like a binary decision by only providing one extreme as an alternative to the status quo.

btw, below I give a well argued article which claims BIG is a useful concept.

www.usbig.net/pdf/manyfacesofubi.pdf

I don't understand your problem with a basic income guarantee. You say that it is unhelpful, but it is a very clear description of the mechanics of redistribution. It is much easier to explain a "basic income" to my mother than a "negative income tax" -- and yes she understands that anything provided by the government must be paid for by taxation.

The unconditional aspect of a basic income allows individuals to accumulate capital. Many welfare systems require recipients to sell their cars, property, and maintain bank accounts at a near zero balance. They also clawback employment earnings at punitive tax rates of 50-100% (often higher than the top-marginal rate). Since welfare is often very low, those supplementing their income are often driven to working under-the-table for cash. This creates a welfare trap that keeps poor people poor.

Means checking requires an expensive bureaucracy, is paternalistic, and I think unnecessary.

If you read the research results for the Canadian and US experiments you will find that people want to work, even with a basic income. The slight negative incentives that were discovered can be attributable to two basic factors: 1. some women chose to stay home to care for children or elderly family members -- this has an obvious societal benefit, and 2. if the experiment is conducted in a non-saturation site (i.e. one fraction of the population gets a basic income while the other does not) then there will be an economic distortion affecting low income workers. Employers are driven to employ those without a basic income, who are more desperate for work. The same would happen if you tried to enforce a minimum wage on half the population. Employers would hire those without a minimum wage.

If you are looking for the real problems with the basic income, it is easy to find. The cost to raise everyone above the poverty line is _very_ high. One would have to consider the size of the family unit receiving funds and adjust for economies of scale. Nevertheless, it could be done. By contrast, a basic income at the rate of many existing welfare systems would be very affordable and would result in many collateral savings such as reduced health care and administration costs.

Also, despite the cost of a basic income, it has the support of many Nobel-prize winning economists because it increases utility. I have not yet seen a model for optimal income taxation that does not point to a negative tax / basic income. See, for example, textbooks on public economics or the classic paper by Mirrlees,

http://www.econ.yale.edu/~dirkb/teach/pdf/mirrlees/1971%20op...

While you make some interesting arguments for a basic income guarantee, I think your post would be better as a top level comment.

You don't seem to have addressed anything I wrote past the first two paragraphs (which were a kind of introduction). E.g. I explained precisely what I meant by the concept being "unhelpful".

Reply on news.ycombinator.com