Strong links vs weak links

The general election should rid us of one of the most vicious, corrupt and incompetent governments in our history, but it will probably not install a government able or willing to meet the huge challenges of a failing economy, wrecked public services and climate change.

With players of the quality of Saka, Foden, Rice and Bellingham, England have a great chance of winning the Euros this year, but with players like Maguire or Dunk they do not.

These two paragraphs seem very different. But they have something important in common. They illustrate the distinction between strong-link problems and weak-link ones.

A weak-link problem is where success depends upon the quality of the worst component, whereas a strong-link problem is where it depends upon the quality of the best. So, if you're looking at Foden and Saka, you're thinking in strong-link terms whereas if you're looking at Maguire and Dunk, you're seeing through a weak-link lens. Similarly, the election should solve the weak-link problem of ridding us of a terrible government, but it will not solve the strong-link one: it'll not give us the best possible government.

This distinction is ubiquitous.

If you're running an airline, oil rig or nuclear plant your focus must be on eliminating the weak-links that can cause catastrophes. Other businesses, though, have strong-link problems. It doesn't much matter if there's a lot of rubbish on Netflix as long as there's sufficient good stuff to attract customers. It has the strong-link problem of how to find good shows and can ignore the weak-link one of what to do about rubbish*.

Many growing companies face strong-link problems - how to stand out sufficiently to win new customers. Mature ones, by contrast, have weak-link problems: they must avoid doing anything which alienates existing customers.

The distinction also matters in investing. Many retired people have the weak-link problem of how not to lose money - a problem which is surprisingly easy to manage for sterling-based investors. Venture capitalists, however, have the strong-link problem: they must find one or two companies with massive payoffs to cover the inevitable losses of failing companies. This is why a state investment bank is a good idea economically speaking - because the state can better throw money around than the private sector - but a bad idea politically because the shitmedia will brand the inevitable failures as a waste of taxpayers' money.

By contrast, economic development can be a weak-link problem. Poor countries are poor often because important things are lacking - reliable electricity, roads, ports, property rights and so on. They can achieve growth therefore by improving these weaknesses. As Charles Jones puts it (pdf):

In any production process, there are many things that can go wrong that will sharply reduce the value of production. In rich countries, there are enough substitution possibilities that these things do not often go wrong. In poor countries, on the other hand, any one of several problems can doom a project. Obtaining the instruction manual (the “knowledge”) for how to produce socks is not especially useful if the import of knitting equipment is restricted, if replacement parts are not readily available, if the electricity supply is erratic, if cotton and polyester threads cannot be obtained, if legal and regulatory requirements cannot be met, if property rights are not secure, or if the market to which these socks will be sold is unknown.

It's not always the case, however, that rich countries do have enough substitution possibilities to prevent trouble when one element fails. As we saw in 2008, the collapse of banks led to big falls in GDP. The crisis taught us that when thinking about banks we have a weak-link problem: the priority is to avoid disaster.

What's true of one industry, however, isn't true of others. Whilst the collapse of NatWest triggered a crisis that of Wilko, Debenhams, Woolworths or Cazoo did not. Which tells us that in well-functioning markets we have strong-link problems: we needn't worry about the existence of bad companies because the market will - eventually - select for good ones.

This poses the question: is the impending collapse of Thames Water a weak-link problem like NatWest or a strong-link one like Wilko? I suspect the latter: it doesn't matter if Thames Water can't supply water as long as somebody can; the logo doesn't matter as long as the taps still run.

Yet another weak-link problem is monetary policy. For Milton Friedman good policy (pdf) was a weak-link problem: “a negative proposition: avoid major mistakes.” The strong-link problem - boosting economic growth - requires other policies.

By contrast, science and the arts are strong-link problems. It doesn't much matter if there's a lot of dross as long as people aren't daft enough to act upon bad research! What matters is that quality research is also done, as this is where progress occurs.

Strong-link and weak-link problems have very different solutions**. With weak links, we need risk-aversion and a focus on high minimum standards and safety. With strong-link problems, however, we need to embrace risk, diversity and competition.

This means that sometimes, solving a weak-link problem might exacerbate a strong-link problem. This is perhaps true of the US constitution. It has done a good job in solving the weak-link problem of protecting Americans from tyranny, but not perhaps so good in delivering effective active government. It might also be true in academia: managerialism has weeded out weak-links - academics who are idle drunks and sex pests - but has been less good at fostering strong-links; pressure to publish has given us shoddy research as academics have less time to focus on good work. Many free marketeers, such as Ryan Bourne in Economics in One Virus, add that regulation protects us from the weak-links of dangerous medicines (except OxyContin) but at the expense of slowing down the development of effective ones.

There are trade-offs here, and we must be aware of them.

Often, though, people are not. Sometimes, they have strong-link thinking where we need weak-link thinking.

Boris Groysberg has given some examples (pdf) of this. He's shown that when a good manager of growth companies takes over a mature business (or vice versa) the result is often poor performance; a growth manager has a strong-link mindset whereas mature businesses need weak-link thinking. What matters is not merely an individual's skills, but the match between them and the job requirement.

I saw other examples back in the day job: retail investors who claimed to be risk averse sometimes piled into (correlated!) spivvy Aim stocks.

We see it too in recruiting. Michael Housman and Dylan Minor have shown that companies spend too much time looking for "talent" when they should instead be screening out toxic workers such as the rogue traders who might lose billions, the sexual harrasser who attracts expensive compensation claims or the over-mighty CEO such as Fred Goodwin who could bring down the whole firm.

We also see it in academic publishing, where peer review selects for work that corroborates reviewers' prejudices and selects against heterodox ideas, thereby restricting the diversity upon which the development of strong links depends. Even papers that helped their authors win Nobel prizes have been rejected.

Politics gives us other examples. Sunak has recently spoken of how AI might quickly double productivity whilst Hunt has called for the creation of a "British Microsoft". These are examples of strong-link thinking. But in fact we can better improve our economic performance with weak-link thinking - by simply not doing stupid things such as having trade barriers with the EU or excessively complex taxes and planning rules. Governments should stop talking about being world leaders and instead set themselves the more challenging aspiration of rising to dull mediocrity.

Herein, however, lies a problem - incentives. Sometimes, bosses are incentivized to be strong-link thinkers when they should be weak-link ones. The more naive among you might think that they should ensure that their organizations solve a weak-link problem and eliminate egregious wrong-doing. But this did not happen with the Post Office. As long as the bonuses kept coming, Paula Vennells had no incentive to stop driving sub-postmasters to suicide. Similarly, you might think that police recruitment is a weak-link problem: because bad coppers can do huge harm it's important to weed them out. But because the costs of a Wayne Couzens or Cliff Mitchell and many others are borne by their victims much more than by Home Secretaries or Met Commissioners, such screening is inadequate.

We had the same problem in the run-up to the financial crisis. Banking should be a weak-link problem - stopping them cratering the economy by going bust. But bankers were incentivized to increase returns and take on risk - that is to chase strong links.

My story here is about selection mechanisms. All institutions - be they companies, markets, government departments, elections or whatever - are selection mechanisms: they select for some behaviours and against others. The questions are: what do we want our institutions to select for or against; how well do they do such selections? and; how might we improve them? Such questions are unlikely to be much discussed during this election campaign.

* This is why Disney's decision to withdraw some films from its platforms was so odd, and costly.

** Insofar as they have solutions at all which sometimes they don't: Gareth Southgate can't create top-quality central defenders.

Comments

[Kester Pembroke]

"Yet another weak-link problem is monetary policy. For Milton Friedman good policy (pdf) was a weak-link problem: “a negative proposition: avoid major mistakes.” The strong-link problem - boosting economic growth - requires other policies."

You're forgetting the other automatic stabiliser - the currency exchange rate.

What everybody misses in MMT, largely because they are usually American, is Warren's analysis of the floating FX system. What a largely static tax base (land value tax) does is anchor the FX exchange by providing a constant demand for the currency.

You don't want the level of tax to auto stabilise that much and you certainly don't want it to start hard and then ease off as behaviour changes as a net effect. You also don't want to end up creating a perverse effect as we see now with fuel duty where certain characters start talking about 'black holes' in the budget if we move to EVs.

"Herein, however, lies a problem - incentives."

With a levy/subsidy hypothecated pairing nobody can play the 'black hole' game. It's crystal clear what spending drops as the tax raised disappears.

Tax does fund discretionary spending in MMT, just at physical level not money level. Everything is still tax and spend once the JG auto stabiliser is in place.

[Mike]

"the sh*tmedia will brand the inevitable failures as a waste of taxpayers' money"

spot on, Britain can never have an industrial strategy because it once subsidised a crap car in the 70s

[rsm]

"But bankers were incentivized to increase returns and take on risk - that is to chase strong links."

What has changed? And didn't CDS cover the weak link risk, except rumors and a spreading panic about a default wave that never materialized led irrational traders to devalue even the safe assets that the insurance held as collateral? So what is wrong with relying on the Fed to act as a value buyer in spreading, emotional panics to prevent contagion to the real economy?

[Blissex]

«A weak-link problem is where success depends upon the quality of the worst component, whereas a strong-link problem is where it depends upon the quality of the best.»

These used to be called differently:

MINMAX: minimize the maximum loss.
MAXMAX: maximize the maximum win.

There are other similar concepts, for example:

MAXMIN: maximize the minimum win.

They can be combined as primary and secondary goals, even if often they are conflicting.

[Blissex]

«Britain can never have an industrial strategy because it once subsidised a crap car in the 70s»

That myth of "no industrial strategy" is about as baseless as the myth of "austerity" because the governments of the past 40 years have all intervened massively in the markets to pick winners, that is finance, property, higher education, and secondarily pharma and defense, with both colossal amounts of public money and with ultra-favourable regulation and policy, for example:

Osborne: «It was up to the Bank of England, he added, to support demand over the coming months. “A credible fiscal plan allows you to have a looser monetary policy than would otherwise be the case. My approach is to be fiscally conservative but monetarily active.”»

Osborne: «Hopefully we will get a little housing boom and everyone will be happy as property values go up»

Cameron: «It is hard to overstate the fundamental importance of low interest rates for an economy as indebted as ours… …and the unthinkable damage that a sharp rise in interest rates would do. When you’ve got a mountain of private sector debt, built up during the boom… …low interest rates mean indebted businesses and families don’t have to spend every spare pound just paying their interest bills.»

In particular the governments of the past 40 years have spent many many times more on subsidising finance and property than the "socialist" governments of the past spent keeping BL etc. going. Just one example (the balance sheet of the BoE shows how much bigger state generosity has been to the failed finance and property industries):

http://www.coppolacomment.com/2018/01/the-carillion-whitewash.html
«RBS was deeply insolvent. Rescuing it cost the U.K. Government £45bn, and RBS has lost a further £58bn since. Nearly ten years after the crisis, it is still in majority public ownership.»

The point that ultra-neoliberal "The Economist" loudly called for huge state spending to save "jobs for the precious scions of the elite" in the City:

http://www.economist.com/node/21542417/
“Britain will one day wake up to discover that it has lost one of the world's most successful business clusters, and the best hope the next generation has of earning a decent living.”

[Blissex]

«The questions are: what do we want our institutions to select for or against;»

I think this used to be called "politics" and used to be about big conflicts of interests among classes, not a generic "we", but them vs. us. The technical questions that follow:

«how well do they do such selections? and; how might we improve them?»

Seem also rather minor compared to "politics", answers are easy to find and try. However given that "we are all thatcherites now" (Mandelson) and "we are all in the same boat" (Cameron) I guess that we are at "end the of history" (Hegel, Fukuyama) and "politics" is over. :-)

[rsm]

《colossal amounts of public money》

Is it only politics based on faulty, sycophantic economists, that say that printed money has to somehow be "paid back" to satisfy the Zero-Sum Gods?

[Jan Wiklund]

It's not entirely clear that poor countries only need to think of weak-link matters. Alice Amsden has pointed out that those poor countries that have advanced have focussed on upgrading, not roads. They have discriminated against low-technology and tried to break into high-income countries' monopolies – like when South Korea's military dictator threatened then-textile producer Hyundai's CEO with prison if he didn't develop it into a shipyard instead.

This was what the US and Germany did in the 19th century, Japan and South Korea did in the 20th and what China did from the 1970s. Getting rich, on a national level, is all about breaking into rich countries' monopolies.

[Blissex]

«Getting rich, on a national level, is all about breaking into rich countries' monopolies.»

That is a good insight but that applies to *relative* riches. To become rich in an absolute sense there is only one way, and it is to substitute vegetable powered (animal and human) labor with oil powered (engine) labor.

"Energy and the english industrial revolution (2013): “A coal miner who consumes in his own body about 3,500 Calories a day will, if he mines 500 pounds of coal, produce coal with a heat value 500 times the heat value of the food which he consumed while mining it. At 20 per cent efficiency he expends about 1 horsepower-hour of mechanical energy to get the coal. Now, if the coal he mines is burned in a steam engine of even 1 per cent efficiency it will yield about 27 horsepower-hours of mechanical energy. The surplus of mechanical energy gained would thus be 26 horsepower-hours, or the equivalent of 26 man-days per man-day.”

[Blissex]

“Getting rich, on a national level, is all about breaking into rich countries' monopolies.”

One crude way that political economists work use is the average price per kg of a country's exports (just goods or both goods and services).

«substitute vegetable powered (animal and human) labor with oil powered (engine) labor.»

Indeed related to the above it matters quite a lot if the productivity of oil is used to make flip-flops (low price per pkg) or telecom switches (high price per kg).

David Landes in his excellent "The wealth and poverty of nations" makes this important point:

D Landes this is the most important one:

“The gains from trade are unequal. As history has shown, some countries will do much better than others. The primary reason is that comparative advantage is not the same for all, and that some activities are more lucrative and productive and than others. (A dollar is not a dollar is not a dollar.) They require and yield greater gains in knowledge and know-how, within and without. [...] Comparative advantage is not fixed and it can move for or against.”

[Blissex]

«those poor countries that have advanced have focussed on upgrading, not roads. They have discriminated against low-technology and tried to break into high-income countries' monopolies»

I think that according to our blogger poor countries become rich by making their best qualified and most active people mass-emigrate to rich countries, just like the USA, Japan, Taiwan, South-Korea, Singapore did :-).

According to Joe Biden poor countries become rich when they take in lots of foreign immigrants:

https://www.nbcnews.com/news/world/biden-japan-xenophobic-rcna150332
«Biden said the U.S. economy was growing in part “because we welcome immigrants. Think about it. Why is China stalling so bad economically?” he said. “Why is Japan having trouble? Why is India? Because they’re xenophobic. They don’t want immigrants.”»

https://www.economist.com/leaders/2024/04/25/how-strong-is-indias-economy
«Yet despite its slick tech campuses, India is still a semirural society [...] Out of a working-age population of 1bn, only 100m or so have a formal job. Most of the rest are stuck in casual work or joblessness.»

[rsm]

"One crude way that political economists work use is the average price per kg of a country's exports (just goods or both goods and services)."

Why myopically ignore the far greater trade volume in virtual, financial goods? In other words, if worldwide household expenditures are around $90 trillion (see OECD (2024), Household spending (indicator). doi: 10.1787/b5f46047-en (Accessed on 28 May 2024) per year, how much more of an influence on prices does financial trading have, where BIS statistics (see https://data.bis.org/topics/OTC_DER/BIS,WS_OTC_DERIV2,1.0/H.A.A.A.5J.A.5J.A.TO1.TO1.A.A.3.C ) indicates a volume of $600 trillion per year?

[Jan Wiklund]

Blissex | May 28, 2024 at 08:49 PM:

According to modern development economists, such as Amsden, Ha-Joon Chang, Erik Reinert et al - and also Wallerstein, by the way – the more advanced, and therefore more difficult to produce, the more you can charge for the produce.

If course trade is a relative thing. You may earn a lot if you are advanced/heavily capitalized, or you can earn very little, if at all, if you are poor and only able to cut sugarcane. But that's the thing: it is salutary that poor countries get into the niches of the rich ones so that advantages are evened out on the global level.

It's a hard way, and so far few have succeeded. But those who have succeeded have behaved in much the same way, and I put in this reservation just to show that it is not all about weak-link.

[Jan Wiklund]

And, to continue:

The reason why a por country should get out of for example cutting sugarcane or exporting hair for wigs as South Korea did before industrialization, is that wealth is a product of economies of scale, and there is no economies of scale in produces like the above. Only industry has.

And only those industries where competition is weak bring in the big cash.

In a way a country is in the same situation as a growth company. Or at least partially: there must be at least some strong links.

If you don't believe me, ask the Chinese.

[Blissex]

"According to modern development economists, such as Amsden, Ha-Joon Chang, Erik Reinert et al - and also Wallerstein, by the way – the more advanced, and therefore more difficult to produce, the more you can charge for the produce."

That is at the same time *obvious* and highly misleading propaganda: it is a derivative of the Ayn Rand/Robert Gordon propaganda that "creative geniuses" create all wealth by innovating business practices or inventing new technologies.

Now imagine that using oil fueled machines some factory produces 200 woven baskets a day, and another produces 200 jet engine a year: obviously the one producing jet engines a year will be able to command a much higher margin.

But without oil the first factory would be able to produce 10 woven baskets a day, and the second 5 jet engines a year, and of course the second one would still be able to command much higher margins, but most of the wealth would have disappeared in both cases.

So again the crude "price per kg" influences *relative* incomes, but *absolute* incomes depend almost entirely on the widespread and efficient use of oil to feed machines.

"wealth is a product of economies of scale"

Economies of scale are vastly overrated, again because of propaganda (CEO pay depends mostly on the size of their business, being mostly a rent).

Consider the case of an oil-powered big press that can stamp a car door out of a steel ingot:
it is enormously more productive than a small press never mind some workers with anvils and hammers.

But very quickly it is pointless to build an even bigger one, and it is actually counterproductive, so many factories have several such presses. Same for smelters etc.

In particular there are some papers showing that economies of scale are very difficult to achieve beyond a fairly small size in banking and other services, and 80-90% of modern economies are in services, some of them pointless (which may explain quite a bit).

[Blissex]

«But very quickly it is pointless to build an even bigger one, and it is actually counterproductive, so many factories have several such presses»

The point at which instead of bigger-better instances of a machines/complex you put in several means economies of scale have nearly disappeared. There are still probably *some* in putting several identical machines in the same location (e.g. multiple smaller nuclear generators at the same site), but they are second- and third-order compared to the big deal.