Simplicity: smart & stupid

David Gauke has been rightly praised for this:

Rather than recognising the challenges of a fast-changing society require sometimes complex responses, that we live in a world of trade-offs, that easy answers are usually false answers, we have seen the rise of the simplifiers.

This echoes a fine piece by Ian Leslie who says:

The disease of politics today is not populism, so much, as simplism: the oversimplification of complex problems.

I want to quibble, though. There are – to simplify! – two different types of simplicity, the stupid and the smart.

Stupid simplicity is the sort that Gauke and Leslie are rightly decrying – the sort that denies the reality of complexity. This is the sort of simplicity that thinks there is a simple cause of our problems – be it immigrants, Joos or greedy bankers – or that there is a simple answer such as Brexit or higher taxes. Conspiracy theorists are often guilty of this simplicity.  So too are those who fail to see trade-offs, such as between democracy and prosperity: both sides of the Brexit debate are guilty of this.  

There is, however, a smart form of simplicity. It acknowledges that the world is complex and our cognition bounded, and so it looks for simple but empirically-based cognitive short cuts. Gerd Gigerenzer, for example, has shown (pdf) that simple rules of thumb can work well. And Robyn Dawes (pdf) and Paul Meehl (pdf) have both shown that simple statistical models often do better than professional judgments which try to explicitly acknowledge complexity. Sometimes, hedgehogs are better than foxes.

We see this in finance. If you want to know the chances of recession, it’s better to look at a simple indicator such as the yield curve than (pdf) at economists’ forecasts, which almost always fail to see recessions coming. And simple indicators such as the dividend yield, a moving average, foreign buying of US equities or the ratio (pdf) of consumer spending to wealth do a better job or predicting stock market returns than do the professionals.

Sometimes, even, simple policies do better than tricksy ones. Simply buying an index-tracker fund (pdf) means you will often do better than most fund managers who try to incorporate complexity into their stock-picking.

Which poses the question: can the same be said in politics? For me, one of the great virtues of a (high) universal basic income is that it recognizes that people often switch between unemployment, insufficient hours and low-wage work, and that there are huge deadweight costs in trying to assess needs. It’s a simple solution to complexity. In the same vein, simple institutions such as Robert Shiller’s macro markets might do a better job of insuring us against recessions than the impossibly complex job of forecasting them in advance. And part of the case for worker democracy is that dispersed wisdom can cope with complexity better than can top-down management.

My point here is that there are no general rules. Sometimes simple answers work moderately well: in a complex world, the best we can do is satisfice. But sometimes they don’t. All that matters is the evidence. It’s not good enough to pretend that there are simple answers, but equally, telling us that “it’s more complex than that” can be a false sophistication.

Comments

[Christopher Herbert]

Might I recommend a job guarantee policy instead of a universal basic income? The job guarantee serves not only as a strong safety net, but as an inflation stabilizer/hedge. Pgs 304-305 Chapter 19 Full employment Policy, "Macroeconomics" by economists William Mitchell, L. Randall Wray, and Martin Watts. published 2019.

[marku52]

Given the fact that the so-called "experts" have failed in brilliant fashion, a "simple"solution ought to get more popular support.

[Anthony Green]

To act in a Complex Adaptive System it helps to have a framework
https://hbr.org/2007/11/a-leaders-framework-for-decision-making

[Scurra]

@Christopher Herbert: As long as there is an infrastructure in place to ensure that the Job Guarantee is capable of understanding that just because a job is available and a person is available, that doesn't mean they are a good fit.
Alas, we don't even have that *now* so the odds on having it in the future are remote...

[D]

Best to have both lowish UBI and a low paid job guarantee. Combination gives plenty to live on, and keeps incentives to find other jobs reasonably high.

[georgesdelatour]

How would worker democracy work with long, complex supply chains, like those for the iPhone? The design and management of supply chains is considered a high skill job. Companies headhunt individuals who are supposedly good at it. Could worker’s councils in Cupertino, Taipei etc create a “dispersed wisdom” better able to deal with the complexities of supply chains than the people currently paid a lot of money to do it?

[Kester]

I think there is a case for 100% Land Value Tax/Basic Income.

[Blissex]

«economists’ forecasts, which almost always fail to see recessions coming.»

My impression is that most Economists are from the "sell-side" and that they are paid to not predict recessions. JK Galbraith wrote:

“So inaction will be advocated in the present even though it means deep trouble in the future. Here, at least equally with communism, lies the threat to capitalism. It is what causes men who know that things are going quite wrong to say that things are fundamentally sound.”

CAVEAT EMPTOR

[Blissex]

«How would worker democracy work with long, complex supply chains, like those for the iPhone? The design and management of supply chains is considered a high skill job. Companies headhunt individuals who are supposedly good at it.»

That is a really badly chosen example, because "long, complex supply chains" are a very bad business idea, and "optimizing" them makes them even more fragile.

They exist only because they allow bringing forward the benefits of fragility, and pushing back the costs of the same fragility.

The last 1% or even 10% of short-term "efficiency" to be squeezed out of a very fragile supply chain is not even remotely worth it to the business, the employees, the customers.

But it is very much worth it to the CEOs and the investors who want to asset strip their businesses to their own advantage as fast as possible.

Presumably worker councils would take a much longer term, more effectively efficient, view.

[Blissex]

«And part of the case for worker democracy is that dispersed wisdom can cope with complexity better than can top-down management.»

The most important detail is that it has very different incentives from top-down management, among them a bigger incentive for sustainability.

As to narrowly coping with complexity, most real problems cannot be solved, simply or not, because of the prevalence of local maxima, but incrementally improved by looking often in *random* directions; and since the local maxima landscape often changes, adaptivity is necessary.

The point Hayek makes about the wisdom of markets is to some extent correct and completely contradicts "neoclassical" Economics: "tatonnement" needs to be continuous and pervasive in any new or dynamic fields.

[steve]

Current benefit assessment schemes are severely flawed, failing to prevent fraud on the one hand and denying the most vulnerable their rightful benefits on the other. This is not necessarily incompetence. It is simply a difficult problem.
So from this pov UBI is attractive. But: you will still need assessment for additional benefits for children, housing, disability, etc. So how much of the bureaucracy do you save? And what percentage of the most vulnerable will be eligible for the additional benefits, and so get caught up in the same assessment machine? I'd like to see some analysis of these questions before voting for UBI.

[Dagan]

I do not understand why UBI will not be captured by increases in residential rents.

[georgesdelatour]

@Blissex

Thank you for your thoughtful reply which, to be honest, I don’t fully understand yet. I understand that a local restaurant in, say, Cornwall, can promise that all its dishes serve “locally sourced ingredients”, and therefore have a totally local supply chain. It can even charge a premium for this localism. But isn’t an iPhone always going to have planet-spanning supply chains, requiring complex co-ordination? And won’t these co-ordination problems inevitably require hierarchical decision making? Okay, maybe the outer limits of professionalised supply chain design hit diminishing returns, but by that point we’re still ahead aren’t we? I promise I’m not trying to act dumb here.

A more general problem is risk aversion. Apple’s initial decision to make the iPhone was a huge risk which paid off handsomely, transforming the mobile phone market beyond recognition. If the decision to proceed with the new product had needed approval by worker’s councils in Taiwan, the US etc, wouldn’t risk aversion and general human conservatism have frozen the iPhone’s development and manufacture?

[georgesdelatour]

@Blissex

I understand the word “sustainability” in its specific ecological meaning - e.g. we will eventually use up all of the planet’s natural reserves of hydrocarbons, so relying on hydrocarbons for energy is ultimately “unsustainable”.

But I’m less sure about using the word “sustainability” more broadly. It implies we should never do anything unless we can commit to doing it without change for all eternity.

Businesses rarely disappear because they exhaust the reserves of natural resources needed to make their products. Usually it’s because a new product eats the demand for their old product. Sales of Nokia phones dropped sharply not because of natural resource exhaustion, but because would-be customers switched to smartphones. That sort of thing…

[Blissex]

«Current benefit assessment schemes are severely flawed, failing to prevent fraud on the one hand»

According to this Conservative far-right government, benefit fraud (excluding the finance sector of course :->) has reached the outrageous level of 0.5% of amounts paid, proving just how "severely flawed" are assessment schemes in their laxity. :-)

[Blissex]

«But isn’t an iPhone always going to have planet-spanning supply chains, requiring complex co-ordination?»

Making iPhones is always going to be somewhat complex, and according to Hayek top-down coordination is always going to lose to the distributed wisdom of the crowds when dealing with complexity. Therefore even quite complex products are assembler out of discrete parts.

But that does not imply in any way "planet-spanning supply chains", it used to be that products as complex as cars were built in large vertically integrated factories, divided is sections/stages, where each part was built.

The reasons why they are no longer built so, despite the vast extra costs in transport and logistic complexity, are labour arbitrage and the ability to discourage unionisation by dispersing workers across many sites. Lower worker wages then more than justify the extra complexity and higher costs of "planet-spanning supply chains".

If you want an example of something complex that requires fairly high coordination skills, airplanes are better than iPhones, but even those used to be mostly built in vertically integrated plants (also for quality control).

[Blissex]

«If the decision to proceed with the new product had needed approval by worker’s councils in Taiwan, the US etc, wouldn’t risk aversion and general human conservatism have frozen the iPhone’s development and manufacture?»

Well, a decision of this type "Heads all my employees lose their jobs, tails I am ten times more a billionaire than I already am" may not be entirely desirable except from an elitist point of view.

Besides I am tired with the example of the iPhone, where most people's cost of living depends on access to fairly priced housing, transport, education, healthcare, essentials, which are mostly locally produced and consumed.

[Blissex]

«But I’m less sure about using the word “sustainability” more broadly. It implies we should never do anything unless we can commit to doing it without change for all eternity.»

That's only in your imagination, and I am sorry that you burden what you read with an overlay of that imagination. As to the text you were reading it was quite definite and hard to misinterpret:

* “the CEOs and the investors who want to asset strip their businesses to their own advantage as fast as possible.”

* “Presumably worker councils would take a much longer term, more effectively efficient, view.”

* “has very different incentives from top-down management, among them a bigger incentive for sustainability”

Here "sustainability" is in the fairly common sense of "sustainable business model" (plenty of Google hits on that phrase!), that is not one based on asset stripping.

Ordinary workers tend to have a longer time horizon than CEOs, who tend to hold to their post for only a few years, and investors tend to hold on to their security for less than a year.

Sure, worker councils could end up packed with senior workers next to retirement and inclined to also adopt asset-stripping strategies, but that already happens as a rule to management cohorts and to boards of directors.