Week of May 7, 2018

1. Self-Referential Metadata: Last week's faiV included a link to an app to automatically extract data from charts. I joked it would be the most clicked link in the history of the faiV and it certainly was the most clicked link of the week, more than doubling the clicks on any other link. It was also the most clicked of the last few months. Second place was a review of The Financial Diaries, that unfortunately I suspect many people couldn't read more than the first page of (honestly, it's great, but it's not $45 for 24 hours great. And does anyone ever pay that? Why?).
In other faiV news, Gisella Kagy got in touch to let me know the link to her paper about differential profits by gender for Ghanaian tailors was the right one; and I ran into Leora Klapper who let me know that she forwards the faiV to many colleagues each week. And yes, both of those are really just an excuse to say writing the faiV often feels like shouting into the void. So if you do see stuff you like, or just appreciate the faiV generally, please do get in touch every now and then to let me know. And it's also OK to let me know when I'm getting too niche, too snarky, or you have something you think should be featured in the faiV.

2. US Inequality: It was at the Dignity and Debt Network inaugural meeting that I ran into Leora, and a bunch of other researchers working on household finance, debt and related matters. It was a sociology conference so I had to get used to a format that wasn't paper-centric, but, of course, my bias is to noticing papers. A particularly interesting one was by Barbara Kiviat and Rourke O'Brien finding that low credit scores lower the likelihood of a job offer for a female applicant and lowers the offered salary to black applicants. 
One wonders how such biases play out in the gig economy. Here's a piece on the growing use of 1-day gigs by restaurants and retail. It's practices like that which can make a job guarantee emotionally appealing. Here's Annie Lowrey on the growing momentum behind (very very) vague proposals for a jobs guarantee among Democrat candidates.  

3. Rotten Kin: I'm going to use that as a very tenuous jumping off point to rotten kin as a factor that I don't think gets enough attention in the political economy of job guarantees and universal basic income. At least I hear often about discrimination and racism as explanations for why people would oppose such policies (leaving aside disputes about the basic economics). But I think almost everyone has a cousin or uncle or sibling that they think it would be bad for to get a cash stipend or would abuse a job guarantee in some way. I think that plays a big part in people's skepticism, even if they don't voice it publicly because it's not a nice thing to say about your family.
Anyway, here's Munir Squires writing in VoxDev about "kinship taxes" on Kenyan firm growth finding a fifth of women and a third of men would be willing to pay, and pay a lot, to hide income from their networks. But that's just the tip of the iceberg. Think also of the ways that husbands and wives buy certain goods to protect income from each other. And then think about the rotten kin tax that Indian textile firms are paying based on Bloom, Mahajan McKenzie and Roberts' work--obviously that tax is less than the perceived tax paid if you hire non-relations as managers, but still.

4. Our Algorithmic Overlords: Guys, it's time for some philosophy theory. Stephanie Wykstra (who is one of those people who gives me feedback on the faiV) has a perspective on defining fairness and how we should think about whether algorithms being used in criminal justice are fair. This is something I think about a good bit, as witnessed in my review of Automating Inequality: I perceive a lot of folks giving human beings a really odd "benefit of the doubt" as more fair when compared to an algorithm, but simultaneously decrying systematic and institutional racism and misogyny (which exist! and are very bad! and we should do something about them!) that are the result of those same human beings beings who are supposedly more fair than the algorithm being racist and misogynist. For instance, check out the bureaucratic nightmare of fighting deportation.
This seems quite relevant: When Will Workers Follow an Algorithm? That's a newish paper from Kohei Kawaguchi who runs an experiment with vending machine operators and finds that by-and-large they won't obey the algorithm. Now I wish someone would go back to the seaweed farmers. This is the way future computational social science is going to find me by the way--references to seaweed farmers and isomorphic mimicry.

5. Some Other Stuff: It's late in the day. I want to go outside. But I know you've been eager to know which country is roundest. JP Morgan Chase Institute has a look at the financial impact of Hurricanes Harvey and Irma through the lens of the daily transactions of consumers and businesses in the affected areas. One of the ways we know that families around the world cope with disasters is remittances/informal transfers. Which is why we should be worried if global banks are cutting off correspondent bank relationships and therefore raising the costs of remittances.   


From Richard Reeves, Katherine Guyot and Eleanor Krause, here's a look at differing definitions of what "middle class" means in the US (at least according to economists and researchers). Source:  Brookings Institute

From Richard Reeves, Katherine Guyot and Eleanor Krause, here's a look at differing definitions of what "middle class" means in the US (at least according to economists and researchers). Source: Brookings Institute

Return to the Weekly faiV