The great financial crisis opened my eyes to the world of economics. It has been an interesting journey. Of course, just because I've walked it a long time doesn't mean my destination is the correct one. Without a correct map or proper road signs one may never reach the correct destination, and there seem to be a lot of fake maps out there. I've followed the ones that seem the most likely to get there.
One of the risks econometrics and trying to centrally control economies (and it applies to economies but it's a common management problem in general) is that controlling to a certain number means that the number eventually stops being indicative of anything other than the degree to which it is controlled.
Inflation is higher than the government pretends, which changes everything. One example of this is housing which has become a dominant part of most people's budgets. Another example is food -- someone compared the cost of fast food in 2024 to fast food in 2019, and it's many multiples higher in cost. Those two things alone make up a large portion of people's budgets so changes to those prices represent mass inflation to people.
If we measured inflation the same way we did in 1990, first of all inflation measurements have been at least double what they have been, and second of all that increased inflation totally changes the story of "the greatest economic expansion on record". Instead it tells a story of 2008s financial crisis never ending.
I've already done some research on inflation numbers where I look at a wide number of recurring costs and compare it to the alleged 2% rate of inflation: https://lotide.fbxl.net/posts/32322
It isn't an anecdote when you can look up data from 20 years ago and see that prices for the same thing have gone up on product category after product category. At that point if you're looking at the prices, it isn't an anecdote anymore. It's based on several actual data points. You could argue my methodology is poor, but ultimately it does represent real data and not anecdotes.
Now if you want to ask me how to measure inflation, I think that's a good start -- look at a basket of goods and the prices people pay for those goods, especially as a proportion of an average family's budget that thing is. Now you might go "Obviously that's what they do", but that's not true. In the early 2000s in particular, they added things like hedonic adjustment where you can say "chicken 20 years ago cost much less than today but it's better chicken so even though you're paying more we're going to say you're paying less". Or substitution, where steak went up but you moved to chicken because you can't afford steak so there's no inflation, and then you moved to organ meats because you can't afford chicken so there's no inflation. In addition, there are situations such as "owner equivalent rent" which asks homeowners how much they think they'd have to pay to rent a home equivalent to the one they own. It's proven that this value is consistently lower than actual rent people pay and so it helps reduce inflation.
Shadowstats proposed using the same inflation calculations used in the 1970s before the calculations were revised to show lower values, so there's an example of data people could use. As the shadowstats shows, if we had continued using that number then inflation would be considered consistently higher than it does under the current methodology. Given that I provided that link, it's disingenuous to suggest I'm suggesting metrics can't possibly be taken.
When I say "They change the inflation number so inflation stays down", it should be obvious I'm not referring to the actual experienced inflation. I'm referring to the measured numbers being lower than are actually reflected in what people actually buy.
It's part of the core problem of measurement and business control, that you can have a value that's important and measured and controlled, people will find ways to make sure the number looks better, even at the cost of not actually improving the actual metric. This isn't limited to economics, it's something management schools such as the Harvard Business School in their Harvard Business Review magazine have written about at length, because good managers (and arguably government consists of managers) need to be cognizant of the effects that measurement and control can have on values separate from the reality we're trying to measure and control.
Economics is the study of incentives, and the incentives are powerful for government to fudge the numbers with respect to inflation. If the government measures inflation low, then they can claim a chunk of prices going up as economic growth. It can claim a chunk of the debt it runs up is just keeping pace with inflation. It gives the central bank an excuse not to fight inflation, since fighting inflation is a really painful thing. None of those incentives represent an omnipotent or impotent organization, but one made up of humans who prefer to do the thing that's incentivized for them. When everyone has the incentive to do the wrong thing, it's likely everyone will go along with doing the wrong thing.
If you'd like an example, when Ronald Reagan was president, the national debt was about 1T dollars. Since then, every president since has roughly doubled the national debt. Reagan more than doubled the debt (according to FRED, the debt went from about 1T to about 3T), and since then every president has taken on some form of his "spend more money cut more taxes" regime, and the democrats and republicans have been in the presidency roughly the same amount of time, and they've often been in congress about half the time, and it doesn't really matter, whoever is president, whoever is in congress, whoever is in the senate, the debt has about doubled since the very late 1970s. (there was a brief reprieve during Clinton's presidency, but it's highly arguable that one of the reasons they could do that was the unprecedented dot com bubble and surrounding economic activity) It's obvious that this will lead to problems down the road, but the incentives are such that it's very difficult to get elected on a platform of raising taxes and cutting spending. It's one of the dangers of democracy Plato warned about.
Of course, that example also shows some of the strong incentives for keeping the inflation number low.
If inflation can be said to be low, then central bank interest rates can be kept low, and that'll help keep overall debt costs low by ensuring there's lots of money in the monetary system to buy that debt, which will help keep debt service costs low, which means borrowing is easier to justify since the cost of borrowing is low. We're seeing that right now, where in spite of only a relatively small increase in nominal debt, the debt service costs have doubled, and they're on track to increase considerably more.
In addition, if inflation is reported as low, then government costs that are indexed with inflation can be kept lower. For example, social security is indexed to inflation, but despite that fact nearly half of baby boomers are considering re-entering the workforce because the cost of living is in fact rising faster than official inflation values. In addition, the government sells inflation indexed bonds, and as much as inflation can be downplayed, every basis point is money in the government's pocket.
If inflation is reported as low, then that also changes the econometrics elected officials can run on, and it changes the required strategy moving forward. If inflation has been 2% from the end of the GFC around 2009 to the beginning of the pandemic in Q1 2024, then it was the longest period of economic expansion in history. If inflation was 6% or more, then it was the longest economic decline on record. If inflation is low, then it justifies continuing availability of debt which feels really good. If inflation is getting higher, then the availability of debt must be restricted which under conventional macroeconomic theory will cause a reduction in economic activity.
Creation of too much money can harm the economy in two ways besides one I've directly discussed.
First, in a previous post I pointed out that most recessions over the past 200 years have been situations where there's too much money, a thing is slightly profitable and so everyone borrows money to invest in it, it turns out to not be astronomically profitable meaning everything is savagely overvalued causing all the money invested in it to effectively disappear, leading to banks failing due to losing the money they lent to customers who invested in the thing. This means that too much money can cause a period of high growth, but it will lead to a period of decline. One of the most famous of these is the 1929 great depression, which was caused in part by the additional money in the system caused by the 1913 creation of the federal reserve bank which led banks to believe they were immune to bank runs, and the profitable thing was farming, combined with a period of improved rainfall in the south which made people believe they could just keep building farms and the rains would come. This led to the period known as the roaring 20s, when everyone was getting rich, but soon the climate changed back to normal and many of those farms turned to dry dust wracked with sand storms in a massive environmental cataclysm called the "dust bowl". When this happened, all those banks were suddenly on the line for massive amounts of money and failed, which was one of the reasons for the massive economic collapse.
Second, a period of increasing inflation eventually stops resulting in additional economic growth, in the same way that someone full of cocaine eventually grows insensitive to the drug and mellows out at the same dose. This leads to stagflation, a situation of high inflation but low or negative economic growth. Most textbooks will simply say "this is a difficult situation to escape from". The 1970s was an example of stagflation due to a combination of monetary policy and growing government spending, combined with a massive economic shock from a huge rise in oil prices. The solution at the time ended up being in part central bank interest rate hikes that make the past few years look like nothing -- mortgages whose interest rates looked like credit cards are an example of the effects on general debt. This was hard for a lot of people, and a lot of people lost their homes, their cars, and more.
Again, none of these incentives require the government to be omnipotent or impotent, they're just incentives that exist and will promote certain behaviors generally within government.
The issues with econometrics have been well understood in other contexts as well. For example, the famines in both Mao's China and the Soviet Union were both caused by perverse incentives surrounding reported data. In both cases, the people in charge were incentivized to make their numbers look good regardless of whether they actually were good or not, and then the leaders looked better than they actually were in the short term at the expense of the people and decisions were made that had negative consequences because they were based on bad data. That case was a lot more direct, but human beings react to incentives, and it doesn't need to be "Mao will have you shot" to modify people's behavior significantly, and it doesn't need people to be actively trying to lie or be malicious to affect reporting or design of econometrics to the detriment of understanding the real world data.
One of the risks econometrics and trying to centrally control economies (and it applies to economies but it's a common management problem in general) is that controlling to a certain number means that the number eventually stops being indicative of anything other than the degree to which it is controlled.
Inflation is higher than the government pretends, which changes everything. One example of this is housing which has become a dominant part of most people's budgets. Another example is food -- someone compared the cost of fast food in 2024 to fast food in 2019, and it's many multiples higher in cost. Those two things alone make up a large portion of people's budgets so changes to those prices represent mass inflation to people.
If we measured inflation the same way we did in 1990, first of all inflation measurements have been at least double what they have been, and second of all that increased inflation totally changes the story of "the greatest economic expansion on record". Instead it tells a story of 2008s financial crisis never ending.
I've already done some research on inflation numbers where I look at a wide number of recurring costs and compare it to the alleged 2% rate of inflation: https://lotide.fbxl.net/posts/32322
It isn't an anecdote when you can look up data from 20 years ago and see that prices for the same thing have gone up on product category after product category. At that point if you're looking at the prices, it isn't an anecdote anymore. It's based on several actual data points. You could argue my methodology is poor, but ultimately it does represent real data and not anecdotes.
Now if you want to ask me how to measure inflation, I think that's a good start -- look at a basket of goods and the prices people pay for those goods, especially as a proportion of an average family's budget that thing is. Now you might go "Obviously that's what they do", but that's not true. In the early 2000s in particular, they added things like hedonic adjustment where you can say "chicken 20 years ago cost much less than today but it's better chicken so even though you're paying more we're going to say you're paying less". Or substitution, where steak went up but you moved to chicken because you can't afford steak so there's no inflation, and then you moved to organ meats because you can't afford chicken so there's no inflation. In addition, there are situations such as "owner equivalent rent" which asks homeowners how much they think they'd have to pay to rent a home equivalent to the one they own. It's proven that this value is consistently lower than actual rent people pay and so it helps reduce inflation.
Shadowstats proposed using the same inflation calculations used in the 1970s before the calculations were revised to show lower values, so there's an example of data people could use. As the shadowstats shows, if we had continued using that number then inflation would be considered consistently higher than it does under the current methodology. Given that I provided that link, it's disingenuous to suggest I'm suggesting metrics can't possibly be taken.
When I say "They change the inflation number so inflation stays down", it should be obvious I'm not referring to the actual experienced inflation. I'm referring to the measured numbers being lower than are actually reflected in what people actually buy.
It's part of the core problem of measurement and business control, that you can have a value that's important and measured and controlled, people will find ways to make sure the number looks better, even at the cost of not actually improving the actual metric. This isn't limited to economics, it's something management schools such as the Harvard Business School in their Harvard Business Review magazine have written about at length, because good managers (and arguably government consists of managers) need to be cognizant of the effects that measurement and control can have on values separate from the reality we're trying to measure and control.
Economics is the study of incentives, and the incentives are powerful for government to fudge the numbers with respect to inflation. If the government measures inflation low, then they can claim a chunk of prices going up as economic growth. It can claim a chunk of the debt it runs up is just keeping pace with inflation. It gives the central bank an excuse not to fight inflation, since fighting inflation is a really painful thing. None of those incentives represent an omnipotent or impotent organization, but one made up of humans who prefer to do the thing that's incentivized for them. When everyone has the incentive to do the wrong thing, it's likely everyone will go along with doing the wrong thing.
If you'd like an example, when Ronald Reagan was president, the national debt was about 1T dollars. Since then, every president since has roughly doubled the national debt. Reagan more than doubled the debt (according to FRED, the debt went from about 1T to about 3T), and since then every president has taken on some form of his "spend more money cut more taxes" regime, and the democrats and republicans have been in the presidency roughly the same amount of time, and they've often been in congress about half the time, and it doesn't really matter, whoever is president, whoever is in congress, whoever is in the senate, the debt has about doubled since the very late 1970s. (there was a brief reprieve during Clinton's presidency, but it's highly arguable that one of the reasons they could do that was the unprecedented dot com bubble and surrounding economic activity) It's obvious that this will lead to problems down the road, but the incentives are such that it's very difficult to get elected on a platform of raising taxes and cutting spending. It's one of the dangers of democracy Plato warned about.
Of course, that example also shows some of the strong incentives for keeping the inflation number low.
If inflation can be said to be low, then central bank interest rates can be kept low, and that'll help keep overall debt costs low by ensuring there's lots of money in the monetary system to buy that debt, which will help keep debt service costs low, which means borrowing is easier to justify since the cost of borrowing is low. We're seeing that right now, where in spite of only a relatively small increase in nominal debt, the debt service costs have doubled, and they're on track to increase considerably more.
In addition, if inflation is reported as low, then government costs that are indexed with inflation can be kept lower. For example, social security is indexed to inflation, but despite that fact nearly half of baby boomers are considering re-entering the workforce because the cost of living is in fact rising faster than official inflation values. In addition, the government sells inflation indexed bonds, and as much as inflation can be downplayed, every basis point is money in the government's pocket.
If inflation is reported as low, then that also changes the econometrics elected officials can run on, and it changes the required strategy moving forward. If inflation has been 2% from the end of the GFC around 2009 to the beginning of the pandemic in Q1 2024, then it was the longest period of economic expansion in history. If inflation was 6% or more, then it was the longest economic decline on record. If inflation is low, then it justifies continuing availability of debt which feels really good. If inflation is getting higher, then the availability of debt must be restricted which under conventional macroeconomic theory will cause a reduction in economic activity.
Creation of too much money can harm the economy in two ways besides one I've directly discussed.
First, in a previous post I pointed out that most recessions over the past 200 years have been situations where there's too much money, a thing is slightly profitable and so everyone borrows money to invest in it, it turns out to not be astronomically profitable meaning everything is savagely overvalued causing all the money invested in it to effectively disappear, leading to banks failing due to losing the money they lent to customers who invested in the thing. This means that too much money can cause a period of high growth, but it will lead to a period of decline. One of the most famous of these is the 1929 great depression, which was caused in part by the additional money in the system caused by the 1913 creation of the federal reserve bank which led banks to believe they were immune to bank runs, and the profitable thing was farming, combined with a period of improved rainfall in the south which made people believe they could just keep building farms and the rains would come. This led to the period known as the roaring 20s, when everyone was getting rich, but soon the climate changed back to normal and many of those farms turned to dry dust wracked with sand storms in a massive environmental cataclysm called the "dust bowl". When this happened, all those banks were suddenly on the line for massive amounts of money and failed, which was one of the reasons for the massive economic collapse.
Second, a period of increasing inflation eventually stops resulting in additional economic growth, in the same way that someone full of cocaine eventually grows insensitive to the drug and mellows out at the same dose. This leads to stagflation, a situation of high inflation but low or negative economic growth. Most textbooks will simply say "this is a difficult situation to escape from". The 1970s was an example of stagflation due to a combination of monetary policy and growing government spending, combined with a massive economic shock from a huge rise in oil prices. The solution at the time ended up being in part central bank interest rate hikes that make the past few years look like nothing -- mortgages whose interest rates looked like credit cards are an example of the effects on general debt. This was hard for a lot of people, and a lot of people lost their homes, their cars, and more.
Again, none of these incentives require the government to be omnipotent or impotent, they're just incentives that exist and will promote certain behaviors generally within government.
The issues with econometrics have been well understood in other contexts as well. For example, the famines in both Mao's China and the Soviet Union were both caused by perverse incentives surrounding reported data. In both cases, the people in charge were incentivized to make their numbers look good regardless of whether they actually were good or not, and then the leaders looked better than they actually were in the short term at the expense of the people and decisions were made that had negative consequences because they were based on bad data. That case was a lot more direct, but human beings react to incentives, and it doesn't need to be "Mao will have you shot" to modify people's behavior significantly, and it doesn't need people to be actively trying to lie or be malicious to affect reporting or design of econometrics to the detriment of understanding the real world data.
It could be. The post on fbxl social is about 11k and ends with a paragraph containing the text "Mao will have you shot"
I think a lot of sites only federate so many characters.
I think a lot of sites only federate so many characters.
Ah, not that much left then. Here's the remainder:
First, in a previous post I pointed out that most recessions over the past 200 years have been situations where there's too much money, a thing is slightly profitable and so everyone borrows money to invest in it, it turns out to not be astronomically profitable meaning everything is savagely overvalued causing all the money invested in it to effectively disappear, leading to banks failing due to losing the money they lent to customers who invested in the thing. This means that too much money can cause a period of high growth, but it will lead to a period of decline. One of the most famous of these is the 1929 great depression, which was caused in part by the additional money in the system caused by the 1913 creation of the federal reserve bank which led banks to believe they were immune to bank runs, and the profitable thing was farming, combined with a period of improved rainfall in the south which made people believe they could just keep building farms and the rains would come. This led to the period known as the roaring 20s, when everyone was getting rich, but soon the climate changed back to normal and many of those farms turned to dry dust wracked with sand storms in a massive environmental cataclysm called the "dust bowl". When this happened, all those banks were suddenly on the line for massive amounts of money and failed, which was one of the reasons for the massive economic collapse.
Second, a period of increasing inflation eventually stops resulting in additional economic growth, in the same way that someone full of cocaine eventually grows insensitive to the drug and mellows out at the same dose. This leads to stagflation, a situation of high inflation but low or negative economic growth. Most textbooks will simply say "this is a difficult situation to escape from". The 1970s was an example of stagflation due to a combination of monetary policy and growing government spending, combined with a massive economic shock from a huge rise in oil prices. The solution at the time ended up being in part central bank interest rate hikes that make the past few years look like nothing -- mortgages whose interest rates looked like credit cards are an example of the effects on general debt. This was hard for a lot of people, and a lot of people lost their homes, their cars, and more.
Again, none of these incentives require the government to be omnipotent or impotent, they're just incentives that exist and will promote certain behaviors generally within government.
The issues with econometrics have been well understood in other contexts as well. For example, the famines in both Mao's China and the Soviet Union were both caused by perverse incentives surrounding reported data. In both cases, the people in charge were incentivized to make their numbers look good regardless of whether they actually were good or not, and then the leaders looked better than they actually were in the short term at the expense of the people and decisions were made that had negative consequences because they were based on bad data. That case was a lot more direct, but human beings react to incentives, and it doesn't need to be "Mao will have you shot" to modify people's behavior significantly, and it doesn't need people to be actively trying to lie or be malicious to affect reporting or design of econometrics to the detriment of understanding the real world data.
First, in a previous post I pointed out that most recessions over the past 200 years have been situations where there's too much money, a thing is slightly profitable and so everyone borrows money to invest in it, it turns out to not be astronomically profitable meaning everything is savagely overvalued causing all the money invested in it to effectively disappear, leading to banks failing due to losing the money they lent to customers who invested in the thing. This means that too much money can cause a period of high growth, but it will lead to a period of decline. One of the most famous of these is the 1929 great depression, which was caused in part by the additional money in the system caused by the 1913 creation of the federal reserve bank which led banks to believe they were immune to bank runs, and the profitable thing was farming, combined with a period of improved rainfall in the south which made people believe they could just keep building farms and the rains would come. This led to the period known as the roaring 20s, when everyone was getting rich, but soon the climate changed back to normal and many of those farms turned to dry dust wracked with sand storms in a massive environmental cataclysm called the "dust bowl". When this happened, all those banks were suddenly on the line for massive amounts of money and failed, which was one of the reasons for the massive economic collapse.
Second, a period of increasing inflation eventually stops resulting in additional economic growth, in the same way that someone full of cocaine eventually grows insensitive to the drug and mellows out at the same dose. This leads to stagflation, a situation of high inflation but low or negative economic growth. Most textbooks will simply say "this is a difficult situation to escape from". The 1970s was an example of stagflation due to a combination of monetary policy and growing government spending, combined with a massive economic shock from a huge rise in oil prices. The solution at the time ended up being in part central bank interest rate hikes that make the past few years look like nothing -- mortgages whose interest rates looked like credit cards are an example of the effects on general debt. This was hard for a lot of people, and a lot of people lost their homes, their cars, and more.
Again, none of these incentives require the government to be omnipotent or impotent, they're just incentives that exist and will promote certain behaviors generally within government.
The issues with econometrics have been well understood in other contexts as well. For example, the famines in both Mao's China and the Soviet Union were both caused by perverse incentives surrounding reported data. In both cases, the people in charge were incentivized to make their numbers look good regardless of whether they actually were good or not, and then the leaders looked better than they actually were in the short term at the expense of the people and decisions were made that had negative consequences because they were based on bad data. That case was a lot more direct, but human beings react to incentives, and it doesn't need to be "Mao will have you shot" to modify people's behavior significantly, and it doesn't need people to be actively trying to lie or be malicious to affect reporting or design of econometrics to the detriment of understanding the real world data.
- replies
- 0
- announces
- 0
- likes
- 0
That's a really insightful thing to say. Makes sense given my personal experience of bureaucracies.
And we have many bureaucrats as leaders, but leaders need not be bureaucrats and outside of a very small number of situations you probably don't want them to be either because just as you say, they're busy trying to get the results they know they're supposed to get and show the people who need to see it that they got that result. Contrast another kind of leader which may have a plan or a vision to go beyond benchmarks, or yet another kind of leader who can help get the best out of their people in ways you might not be able to measure using a microcaliper.
And we have many bureaucrats as leaders, but leaders need not be bureaucrats and outside of a very small number of situations you probably don't want them to be either because just as you say, they're busy trying to get the results they know they're supposed to get and show the people who need to see it that they got that result. Contrast another kind of leader which may have a plan or a vision to go beyond benchmarks, or yet another kind of leader who can help get the best out of their people in ways you might not be able to measure using a microcaliper.