May 2, 2024

The mountain of shit theory

Uriel Fanelli's blog in English

Fediverse

The cooked numbers present the bill.

No, I'm not referring to some restaurant that made customers pay for the numbers. I am referring to the fact that today we see "eminent" German guys in an Italian newspaper explaining that the rules of macroeconomics "no longer work". And at that point, having worked with data analytics systems, it makes me laugh.

That classical economic theories were a refined form of astrology had already been demonstrated, with experiments that showed chimpanzees and fortune tellers, but also stock market astrologers, to be more correct than experts.

It is no coincidence that all the great crises of recent times, at least in the West, always originate from the world of finance. So far, supply chains have held up even in the face of very widespread shocks, such as the end of classic globalization, the energy system has resisted the closure of essential gas pipelines, in short, everything is fine where there is real technology, while the world of finance seems like an irreparable mess.

It is therefore not a problem of "theories that no longer work": they have NEVER worked. There is no difference between Otelma and Puglisi, although Puglisi says he used "mathematical rigor": if axioms rain down, you want to do mathematics.

In short, we know well that yet another blow is about to arrive, Lagarde warned us that we are in a "permacrisis", and we still don't quite understand how it is possible to build and manage systems with the complexity of the global telecommunications network, bringing a signal to anyone, wherever they are, but no way has yet been found to bring 1000 euros to a drowning family.


The problem lies in the numbers, and in the possibility of "cooking" them. I'll give you an example: GDP. Or GDP, if you are Anglo-Saxon. We know that GDP is in theory the sum of financial movements considered all positive in a given country.

Let's leave aside the absurdity of considering a fire that destroys a warehouse as positive, but the problem is that calculating this number would require knowing EVERY financial movement in the country.

You will say that with a lot, a lot, a lot of information technology and a system based entirely on digital payments it would be possible, but in reality the problem remains of recognizing the destructive component of financial operations, in the sense that if by buying the $50,000 car I have made +50,000, giving in my old car I didn't necessarily increase the GDP. But on the numbers , I did.

For example, when I buy a can of tuna, the GDP apparently grows, let's say 1.99 Euros, but what happens when I throw away the tin can? It depends on how much GDP the disposal and resale of the aluminum produces (if it happens). It's just not easy to know if it happens.

Ultimately, therefore, GDP is a number calculated in statistical terms, through complex sample estimates. We make statistics and try to assign a number, forgetting that it often has a greater margin of error than the ones we are measuring.

What does it mean? It means that we know that GDP increased by 0.8%, but the sample estimate we use has an error of 1% and a confidence of 95%.

Ultimately, the 1% increase is not really a number, it is not really an increase, it is something that is "cooked" in data analytics. Something that could be true or false, because it produces false evaluations that decision makers will then transform into bad decisions.

No one, however, wonders what disaster moving in a world of false data could cause.


An example is Joe Biden's collapse in popularity.

Based on the carefully cooked numbers, it would seem that the USA has had a large increase in GDP, even higher than China's, that jobs are pouring in and that people's economic problems have almost been solved.

And in a country like the USA, obtaining a number like that is not difficult. But giving statistics without standard deviation and distribution, Trilussa reminds us, is useless and fraudulent.

The problem of the USA is this: only the bottom decile of the population is better off than the fifth decile of the European population.

For those not familiar with deciles and percentiles, it means that only 10% of Americans are better off than the lowest 50% of Europeans.

These are catastrophic numbers, combined with the fact that it is true that many jobs have been created, but they forget to say that the newly created jobs are, on average, 30% less paid than those that died with the pandemic.

Furthermore, there is a real estate crisis due to enormous real estate speculation that has driven prices through the roof, so the number of Americans with jobs who literally live in tents is increasing.

But you, like all of us, have always and only heard about good news from the USA, and the good news from the courier explains to us that yes, in some cities the situation is terrible, but it is only because the mayor is a communist who leaves blacks unpunished. Strangely, things are no better in Republican-controlled cities.

It's not economic theories that no longer work. What no longer works are the calculations of economic indices and the transparency of calculation methodologies.

Why am I talking about transparency? To give you an example, ISTAT was already parceled out during the Berlusconi period, and so on, it has always had a president who was close or very close to the government.

What is the point of dividing up an ISTAT (but things aren't so different abroad?). The meaning is a pressure to cook the numbers, so that they paint more or less happy situations.

For example, Italy was lulled for six or seven months by forecasts that predicted it would grow more than any European country. The purpose of these forecasts was to push investments to help Italy: if a country grows, then it is good to invest in it, so saying that it will grow attracts investments.

Same thing for the USA, which suddenly grows more than China, for the simple reason that it must be said so.


There is only one problem. And the problem is that cooking numbers to say that everything is fine only works in politics as long as people are fine. I mean, if almost everyone has a job they're happy with, and you tell them that you've created a zillion jobs, they'll most likely believe you and vote for you.

If, however, there are really too many unemployed people, and you tell them the same tall tale, they will think that you have forgotten them and only them, and they will hate you.

Cooking the numbers for the better, to boast great victories, only works when a society is truly wealthy. In a society that is becoming impoverished, it has the opposite effect.

And this is what has produced, and produces, the growing distrust both in the political classes and in the experts.

There was a time when numbers, and mathematical sciences, were credible because, precisely, you could see in the numbers a rigor that would be incontrovertible, and therefore when an expert came on TV and gave you numbers, the valley dweller remained silent.

But ever since communication experts understood that numbers can be cooked before feeding them to the masses, a slew of fake indicators have rained down on us (such as the ratio between GDP and debt, which is falsifiable because not even the debt is easy to calculate: the debt of cities and regions is not counted, for example: Rome alone has 9 billion), absurd decisions have been made such as using the ratio between deficit and GDP in the Maastricht Treaty , rather than using for example the ratio between GDP growth and debt growth (obviously weighted, to avoid divisions by zero), or other numbers. The relationship between GDP and deficit is a catastrophic idea, because a very federal country could spread the deficit over the regions or cities, for example.

In general, these fake economic factors were not included just because the macroeconomic theories were fake (and still are), but because they thought they could dominate the numbers, using numbers that were easy to "cook".

Obviously, all this will lead to a moment of rupture, and to a subsequent crisis, with the necessary adaptation to the new reality.

The same productivity parameter, if the measurement method were presented in a physics faculty, would get loud laughs. But listen, SuperMario was sent to explain to the EU how to increase productivity in Europe.

Someone who has never understood anything about industry and production, for the simple reason that he has always worked in a completely different field.

I believe we are very close to another crisis. A VERY strong crisis, with changes that will not be easy to quell and end.


And as usual, the following applies:

Leave a Reply

Your email address will not be published. Required fields are marked *