No Time Like the Present
1. SMEs: So this is kind of old, at least in faiV terms. But it's new to me, and a good illustration of one of the fundamental ideas that underpins how I look at all research/interventions related to SMEs: Reality has a surprising amount of detail. The point the author is making is quite different from what I take from it, so let me explain a bit more. Figuring out how to run a small business, in most contexts where we care about helping people running small businesses--developing countries, marginalized groups or areas in developed countries, other people markets and regulation have failed--is really, really hard because there is a surprising amount of detail at every step in the process. Product, location, competition, marketing, production, accounting, financing, investment--all of them involve a surprising amount of detail, and lots of little ways to get things wrong. But with so much detail it's hard to figure out if something is going wrong, much less what specific thing is going wrong.
At this surprising level of detail we tend to throw programs that either only address one small detail (e.g. incentives for formalization), or lots of details spread out across many tasks (e.g. business training). In both cases we see small or negligible effects for the most part (in part because most impact evaluations of training don't have nearly enough power to detect the size of change we could reasonably expect).
That's a fairly long disquisition to set up that the next faiVLive will be on the topic of SME business training specifically. On February 20th, at 10am Eastern, David McKenzie and I will discuss what we know about SME performance, management, survival and especially training. Register to join us here.
Finally, while I remain one of the holdouts against the term "financial health" (more on that another day), here's a report from my old colleague Piotr Korynski, now at The Microfinance Centre, looking at the application of financial health to SMEs. It's definitely worth a read to start peeling back layers on the surprising level of detail required to really understand what is happening inside SMEs.
2. Cash: At this point I feel like any discussion of the death of cash should come with a mandatory voiceover of Mark Twain saying "Reports of my death have been greatly exaggerated." Here's Olivier Usher from Nesta on 2020 being a tipping point in the "cash crash." There are some interesting data points here, and more importantly, some important questions about how payment mechanisms affect behavior, or allow others to control behavior.
The virtual voiceover to this particular death of cash pronouncement is from New York City, where the city council just yesterday approved a regulation requiring all businesses in the city to accept cash as payment. That means that 3 of the 15 largest cities in the US, as well as the entire state of New Jersey have banned the death of cash.
3. Financial Inclusion: Financial inclusion, like cash, has frequently been confined to the dustbin of history in recent years, in favor of other terms. As I mentioned I still prefer inclusion (while noting the irony of the name of the research center I manage) but the reasons that others don't are fair and reasonable. One of the main reasons "inclusion" replaced "access" was the recognition that opening lots of dormant accounts really shouldn't count for anything. But shifting terms didn't really blunt the criticism. Here's Bhavana Srivastava and co. from MSC on when financial inclusion is not inclusive for women, and how to change that. Here's IDEO.org on essentially the same topic, looking at what it will take to include women in the financial system in Tanzania, Bangladesh, Kenya, Nigeria, Pakistan and India. And here's Mayada El-Zoghbi on why measures of access and inclusion don't square up with each other.
Bobbi Gray of the Grameen Foundation also has some problems with financial inclusion (sort of)--here's her list of financial inclusion "notions that must die." Of particular note is the third: financial inclusion is always positive. Keep that one in mind while you read this piece on "financial inclusion will see mass market adoption in 2020." If you're wondering what that means, I'm not sure you'll gain much insight from reading it--it's another in a long line of proclamations that "new data" is going to solve all the problems of financial inclusion. But their is one particular sentence that meant I had to link it: "one can only hope that common-sense regulations will enable these technological advances to deliver on their promise of greater financial inclusion." There are so many ways to read that sentence! And most of them aren't encouraging, but are probably right.
To illuminate that somewhat obscure criticism, here's a piece on a highly effective, yet illegal, way to make lending fairer to women. There is no such thing as "common-sense" regulation. This stuff is really, really hard--this would be a good time to go back to the link to Mayada's piece above and read it if you haven't.
4. Corrupted Economy: For some reason I had the song "Tale as Old as Time" faintly echoing in my head as I read the several pieces from the past few weeks about how a variety of mostly American consulting firms had helped Isabel dos Santos loot Angola for the last decade or so. If you haven't read any of the pieces that have appeared in a variety of media, facilitated by the International Consortium of Investigative Journalists, here's the overview story from the New York Times. Here's a piece from Quartz about how dos Santos managed to launder so much money despite the Anti-Money Laundering regulations that cost poor people sending remittances so much. And scroll down in Matt Levine's column last week to the heading of "Consultants" for a useful discussion of the regulatory power that has been delegated imperfectly to banks. It's all a perfect corrupted economy story: there are different rules for the rich and the poor.
For instance, here's a new report on what DoorDash actually pays it's workers when you take into account the time workers spend and the expenses they take on. Using actual data supplied by DoorDash drivers, they calculate the average worker is earning $1.45 an hour, and a third of workers are actually losing money by taking the gigs.
Finally, I've been following the US "Deaths of Despair" story for awhile, but every time I read something new, I end up more confused rather than less. For instance, here's a piece stating that deaths of despair are only rising for white women in the American South.
5. Our Algorithmic Overlords: Speaking of operating by different rules, here's the amazing story of an AI contest gone wrong. The contest seems to have been to develop a machine learning system to predict adoption rates for animals in shelters, but that's not really important. The important part is what the "winner" of the contest did: stole the outcomes for the private data that participants would be judged against, created a many layered application to hide the fact that all they were doing was spitting out the stolen answers, including getting some of the answers wrong so they wouldn't be suspiciously perfect, etc.
It's all terribly amusing, but also terrifying. One of my big worries about our algorithmic overlords is not that the machines are coming for us, but the complexity of the algorithms means only a few people are capable of understanding what is actually happening, and detecting errors. Our algorithmic overlords are not necessarily the algorithms, but the few humans who understand them.
In the meantime, the machines that are taking over are not necessarily the ones you might be thinking of. Driverless tractors are way ahead of driverless cars.
Graphic of the Week
There are all kinds of things wrong with this map of how land in the US is used in terms of data visualization, if you take it too seriously. But as a way of understanding how land is actually used, it's pretty interesting. There are a bunch of other ways of looking at US land use at the source.
