Signal, Noise and Prediction

I'm now going to turn to Nate Silver's new book, The Signal and the Noise: Why So Many Predictions Fail-but Some Don't. Silver is the well-known political forecaster who parlayed his blog FiveThirtyEight into a prominent spot in the New York Times.

I didn't expect much when I bought it. I thought it would be one of those "my quantitative model explains the universe" books (and investment funds) which are so tiresome and common. The world, and especially the markets, are filled with quants who think all you need is Mathematica and some back issues of Econometrica to explain everything. They are usually overconfident, expert on code rather than decisions,and tend to blow up spectacularly like LTCM given time.

Nothing could be further from the truth in this case. The book massively exceeded expectations and turns out to be a thoughtful, mature and reflective. It is consistent with much of my experience and thinking, but I still learned a lot of things I didn't know. It's also fluent and well-written. I'd recommend it without hesitation, and I'll look at it in some detail.

The crux of the book, from someone known for his number-crunching models, is that there is no such thing as objective data-driven models, at least in human affairs.

The numbers have no way of speaking for themselves. We speak for them. We imbue them with meaning. Like Caesar, we may construe them in self-serving ways that are detached from their objective reality. Data-driven predictions can succeed—and they can fail. It is when we deny our role in the process that the odds of failure rise. Before we demand more of our data, we need to demand more of ourselves.

The more information we have, the more we tend to screen out that which does not match our preconceptions. More information most often makes us narrower rather than wiser.

Alvin Toffler, writing in the book Future Shock in 1970, predicted some of the consequences of what he called “information overload.” He thought our defense mechanism would be to simplify the world in ways that confirmed our biases, even as the world itself was growing more diverse and more complex.

Information is no longer scarce, but much of it is not very useful.

Our biological instincts are not always very well adapted to the information-rich modern world. Unless we work actively to become aware of the biases we introduce, the returns to additional information may be minimal—or diminishing.

This does not mean we should just give up, or adopt lazy relativism. Instead, everything is approximate.

Some of you may be uncomfortable with a premise that I have been hinting at and will now state explicitly: we can never make perfectly objective predictions. They will always be tainted by our subjective point of view. But this book is emphatically against the nihilistic viewpoint that there is no objective truth. It asserts, rather, that a belief in the objective truth—and a commitment to pursuing it—is the first prerequisite of making better predictions. The forecaster’s next commitment is to realize that she perceives it imperfectly.

So what are the causes of failure to predict outcomes?

The most calamitous failures of prediction usually have a lot in common. We focus on those signals that tell a story about the world as we would like it to be, not how it really is. We ignore the risks that are hardest to measure, even when they pose the greatest threats to our well-being. We make approximations and assumptions about the world that are much cruder than we realize. We abhor uncertainty, even when it is an irreducible part of the problem we are trying to solve.

Indeed, experts have a particular tendency to ignore threats to their expertise. The rating agencies, for example, did not think through the possibility that default risk of various CDOs and CDO tranches might not be independent and uncorrelated.

The possibility of a housing bubble, and that it might burst, thus represented a threat to the ratings agencies’ gravy train. Human beings have an extraordinary capacity to ignore risks that threaten their livelihood, as though this will make them go away.

Our expectations about the future are riddled with blind spots, as anyone who has ever really thought about the policy process or had to predict events for a living - and been held accountable for it - knows.

We'll look at some other aspects of the book in more detail.

 

Good intentions and catastrophic outcomes

Here's a great essay by Walter Russel Mead on the troubles of Obama's home state, Illinois, as well as California and New York. These states are the greatest fortresses of Democratic power in the country. But they are becoming fiscal catastrophes, undermining the support for the poor and disadvantaged that politicians had championed.

American liberalism today is in an advanced stage of intellectual decline. Cynical and short sighted interests wrap themselves in the increasingly tattered mantles of sacred ideas. Liberals are right to feel that social justice matters, that the poor should have greater opportunity and that government in a democratic society cannot remain indifferent to the existence of great social evils.

But where liberals in America have the freest hand—in states like New York, California and Illinois—we see incontrovertible evidence that the policies they choose don’t have the consequences they predict. California by now should surely be an educational, environmental and social utopia. New York should be a wonder of glorious liberal governance. Illinois should be known far and wide as the state that works.

Just like Europe, which is also clearly in trouble, massive debt and busted pension promises are the result. Good intentions are not enough.

Incidentally, Gallup polls suggest if Europeans had a vote in the US Presidential election, 96% of Irish and 97% of Germans would support Obama. Extraordinary.

 

Ten thousand years of tension.

I was thinking a little more about Pinker's argument in the post below. The North and Midwest were settled by English farmers, he says, the South and West by Scots-Irish herders. Sheep or cattle are much easier to steal than land, and that produces cultural differences. One view is suited to settled, agricultural, pacified societies, the other for frontier areas where state power is weaker.

It occurred to me this is another version of a much more ancient division - between the nomad and the farmer, which stretches back ten thousand years to the very beginning of agriculture. Nomad groups tend to be more warlike , and can sweep in from the steppe to dominant much larger and more culturally refined civilizations. The Mongols swept in to dominate China. The Huns, Vandals and Goths subdued Rome. The Arabs conquered Mesopotamia, Persia and Egypt. The Turks in their turn conquered Byzantium and (by extension) Northern India.

The division is explicitly there in American history - ranchers versus farmers. "The cowman and the farmer should be friends", goes the song in Oklahoma.

And it is a major theme in a famous Chicano novel set in 1940s New Mexicothat I read recently, Bless Me, Ultima. A young child has to choose between the open range cowboy influence of his father's family and the peasant roots of his mother's family.

They are two very different sensibilities with very deep roots. Red and Blue go back to the beginnings of modern society, perhaps.

 

Why are states so red and blue?

Steven Pinker looks at the evidence.

If this history is right, the American political divide may have arisen not so much from different conceptions of human nature as from differences in how best to tame it. The North and coasts are extensions of Europe and continued the government-driven civilizing process that had been gathering momentum since the Middle Ages. The South and West preserved the culture of honor that emerged in the anarchic territories of the growing country, tempered by their own civilizing forces of churches, families and temperance.

I looked at Pinker's excellent (and very long) last book here.

 

Social Credit

The discussion of Fisher's "Chicago Plan" below reminded me of another monetary reform movement -Social Credit, proposed by British Engineer C. H. Douglas in 1924. It had much traction in the 1920s and 1930s, and lasted into the 1980s in some variants. The ideas proved very attractive in Western Canada, especially Alberta. The British Columbia Social Credit Party, the SocReds, ruled the province for forty years, until it imploded in 1991, although it had abandoned most of the earlier monetary ideas.

Much of the theory is a little hoary, like Douglass's "A+B theorem" which appears to overlook multiplier effects. But some of it appears of contemporary relevance. For example, he emphasizes the stock of knowledge and technique and wealth. From the Wikipedia article:

Douglas disagreed with classical economists who divided the factors of production into only land, labour and capital. While Douglas did not deny these factors in production, he believed the “cultural inheritance of society” was the primary factor. Cultural inheritance is defined as the knowledge, technique and processes that have been handed down to us incrementally from the origins of civilization. Consequently, mankind does not have to keep “reinventing the wheel”. “We are merely the administrators of that cultural inheritance, and to that extent the cultural inheritance is the property of all of us, without exception.”[9] Adam Smith, David Ricardo and Karl Marx claimed that labour creates all value. While Douglas did not deny that all costs are ultimately due to labour charges of some sort (past or present), he denied that the present labour of the world creates all wealth.

And he was correct in focusing on the nature of money:

According to economists, money is a medium of exchange. Douglas argued that this may have once been the case when the majority of wealth was produced by individuals who subsequently exchanged it with each other. But in modern economies, division of labour splits production into multiple processes, and wealth is produced by people working in association with each other. For instance, an automobile worker does not produce any wealth (i.e., the automobile) by himself, but only in conjunction with other auto workers, the producers of roads, gasoline, insurance, etc. In this view, wealth is a pool upon which people can draw, and money becomes a ticketing system. The efficiency gained by individuals cooperating in the productive process was coined by Douglas as the “unearned increment of association” – historic accumulations of which constitute what Douglas called the cultural heritage. The means of drawing upon this pool is money distributed by the banking system.

Douglas believed that money should not be regarded as a commodity but rather as a ticket, a means of distribution of production.

He was also very conscious of the issue of abundance, which I am very interested in. Scarcity is no longer the essential problem of mankind, but distribution (and incentivizing innovation and good behavior).

Douglas also claimed the problem of production, or scarcity, had long been solved. The new problem was one of distribution. However; so long as orthodox economics makes scarcity a value, banks will continue to believe that they are creating value for the money they produce by making it scarce.[20] Douglas criticized the banking system on two counts:

for being a form of government which has been centralizing its power for centuries, and

for claiming ownership of the money they create.

The former Douglas identified as being anti-social in policy.[21] The latter he claimed was equivalent to claiming ownership of the nation.[22] According to Douglas, money is merely an abstract representation of the real credit of the community, which is the ability of the community to deliver goods and services, when and where they are required.

He proposed a "national dividend" to consumers, a form of basic income, to make up for inherent deficiencies in effective demand.

I don't think the solutions he proposes necessarily work out, although I haven't read the original books and papers. But much of the diagnosis is intriguingly correct.

 

Radical Monetary Reform

This is an interesting development - highly radical monetary reform proposals from two staff economists at the IMF. The often excitable Ambrose Evans-Pritchard writes about it in the Telegraph.

The IMF paper, which apparently came out in August, argues for a renewed look at a "Chicago Plan" which Irving Fisher put forward in 1936. It required 100% reserve backing for bank loans, thus eliminating the ability of banks to create credit. And that, IMF authors Benes and Kumhof argue, would eliminate much of the volatility of the business cycle, eliminate the possibility of bank runs, and dramatically reduce both public and private debts. It would not be inflationary, and it would boost output by 10%. They claim they can demonstrate this with a DGSE model in a way which Fisher never could.

This shows just how disillusioned the wider world has become with the banking and financial system. The IMF staff has always had some variety of viewpoints - and of course this is very strictly speaking the view of the authors, not the institution - but it is still hard to see the folks on 19th St NW producing something like this ten years ago.

Would it work? It essentially replaces a largely private money system with a 100% government money system. Fractional reserve banking means that right now a bank only has typically 5-10% of "money" - mostly central bank reserves - underpinning the rest of the asset side of the balance sheet, which is bank-created credit.

In our current system, a bank creates money by issuing a loan, and crediting the borrower with an offsetting deposit at the same time. ( eg if Citibank lends you $10,000, it has a $10,000 loan as its asset, and you have an additional $10,000 in your deposit account to spend). The proposal would stop banks creating money, because they could only re-loan reserves from the central bank.

The problem, of course, is that money systems can be too inflexible, too rigid as well as too volatile. This most often shows up in exchange rate policy. The gold standard had fixed exchange rates (but freer credit). However, it still forced international adjustment by inflation and deflation of the price level, because gold was the fundamental reserve asset, not central bank fiat money. "Ye shall not crucify mankind on a cross of gold", William Jennings Bryan famously said.The euro is a prime contemporary example of the problems monetary inflexibility can cause.

The "Chicago Plan" would require policymakers to carefully calibrate the supply of reserves, and if they did not there would be serious problems. Of course, many economists think rules rather than discretion is better for policy in any case, as the track record for discretionary monetary policy is mixed at best. So making policymakers stick to a rue or money growth might be a good thing.

The plan would not necessarily imply government would allocate loans , or choose specific winners and losers. Banks could still lend to whoever they wanted, but they could not create money at the same time. the lack of leverage would cripple bank profits, however.

There might be more role for equity based venture investors in such a system, offsetting some of the credit supply. And the reduction of government debt looks very attractive in current circumstances.

Patience with banks is wearing thin. Policymakers are clearly very frustrated that they are launching massive balance sheet expansions like QE3, but the transmission mechanism to convey that liquidity to the real economy seems to be blocked in the banking system. The banks generally claim that it is because there is less demand for loans, not their reluctance to lend. But whatever the cause, the Fed, BoE and others are not getting much traction. Cutting out the middleman becomes more attractive, at least in theory.

Clearly the financial system would fight this to the death, as it would remove most of the profits in the industry. But radical as it is, perhaps it deserves serious scrutiny as a plan - if only as a device to hold over the banks. Our system does seem to have an inherent bias and drift towards debt, both private and governmental.

Perhaps we need a better system of money creation calibrated to create flourishing in society rather than simply credit. The legitimacy of the current system has been shaken by the crisis, and much of the intellectual confidence of mainstream economics has been eclipsed. I'll read further reactions to the plan and examination of potential flaws with interest.