Sorry, Arnold. I regret your list of GDP pitfalls doesn't include a very important one. You ignore how Government expenditure is misused as a measure of its output. GDP accounting is not only about private output and its market valuation, it's also about government output but not about government redistribution of private income. National accounting pretends to provide good measures of an economy's output, income, and consumption but to a large extent it has failed because of the many government interventions. Unfortunately, public economics textbooks still fail to explain those interventions and their consequences. It's amazing how much time we spend on monetary economics and how little on public economics --yes, this is a good reason to laugh at experts on monetary economics and the army of economists employed by central banks. Although you may prefer to laugh at your government, pay attention to what happened in Argentina yesterday and may happen later today.
"President Biden just happened to be in office when the bonfire was lit."
Total COVID spending dwarves previous ordinary budget deficits by a massive margin. Adding up the entire ordinary deficit of the Trump years wouldn't come close to COVID spending. Had COVID spending not happened, we would not be experiencing this inflation.
If you believe Democrats were the primary drivers of lockdown socialism (fair) and that Biden is nothing more then an empty suit stand in for the Democrats (fair) then I think its fair to blame him for the inflation.
Great essay, Dr. Kling. Nothing in it I can disagree with very much, except for one item at the very end. I can quibble a bit about what isn't included, but that is for another day.
You wrote: "What is the value of the work you do cooking at home?"
Doesn't it show up, though, as spending on other goods and services, or as savings? What you don't spend eating out, gets spent/invested somewhere else. Isn't it a wash?
Finally, you wrote: "People don’t long for: Ford Pintos"
Speak for yourself- my first car in high school was a sky blue Ford Pinto. It had power nothing, no FM radio, but it was my first at age 16, and I would give a lot just get to drive it again.
I have $4.00 but no coffee. The local cafe has coffee but wants $$. We exchange.
This market exchange is evidence of a value add transaction that improves satisfaction (utility, happiness, whatever). I prefer coffee over $, they prefer $ over coffee. There is a deep inequality here -> i have consumer surplus, they have producer surplus.
Measurement of of VOLUNTARY transactions are a (imperfect) measurement of increases in satisfication. GDP sort of measures all transactions, assuming various equalities rather than inequalities (because this is measurable), but that means there is conservativism in the figures, and a few micro assumptions see major sources of inequalities eroded away due to assumed marginal choice effects (eroding consumer surplus of any one decision), competition effects (eroding producer surplus) etc etc.
The big issue though is that a very significant fraction (50%+) of measured economic activity is not actual voluntary but has greater or less coercive elements to it. Taxation is a clear coercive element (partially justified under various democratic and social contract theories that it is actually quasi-volantary hahahahah).....
But more pernicious and invidious is the creep of the regulatory state where "market activity" (i.e transactions that can be measured via tools such GDP) becomes largely driven by the underlying coercive threats (and carrots) from legislation/ regulations, not truly volunatary transactions. With certian interpretations, this could be 80%+ of all measured so-called market activity. By way of example, think of "regulatroy affairs" personnel at large regulated entities. Or entire divisions of people employed at large "private sector" firms whose job activities involve (say) 50-90% of work that is essentially responding to government decree.
All of this massively erodes the theoretical value of GDP as a potential measurement tool of increases in human utility improvement / satisfaction. GDP as a tool potentially started out with some correlations and relevence as measurement of utility, but has become diluted to merely being a measure of money transactions, with almost no relevent to being a measurement of improvement in utility/ satisifcation.
What should GDP measure? Arnold asks, and does answer (in quotes)
Gross Domestic Product - "It is a measure of all of the goods and services that are purchased at prices measured in monetary units. ... explicit compensation in monetary units for providing something, that counts as GDP."
What it IS, today, is also what it SHOULD be measuring.
Tho this does not measure other things that are desired:
"we would like to measure total output of the economy," nor
"We also would like GDP to measure how well the economy provides human satisfaction, or what economists call social welfare" (many social scientists and the UN look for "quality of life")
GDP is the least flawed objective single number measure we have. Much like total SAT as a single number reference for likely college achievement. There's no other better alternative.
Tho it does seem FRED at $24.8 billion of GDP Q2 2022 disagrees with the $19.8 billion number.
"People want to use GDP as a way to “score” economic policy outcomes. I think that is particularly unwise."
They want to, they do, and, like political races being covered like horse races, they will continue to be used thusly.
Is the USA in a recession? Yes, yes it is - but the meaning of "recession" is less significant than GDP. [Recession - when neighbor loses his job.
Depression - when you lose your job.]
At least Arnold provides an alternative.
"the hours-worked number is more reliable than the GDP number, so my bet is that economic activity actually increased between the end of last year and the end of last month."
This might be the case, but after defining 2007 as 100, and calculating % from that so the new high in June 2022 is 112.4, the hours-worked index is not yet used as a key indicator, and there are probably more reasons not yet discussed to not use it.
At least one huge reason is the illegal immigrants, paid in cash to avoid taxes/ deportation/ min wage laws - how they calculated. A long time complaint I have about the 2006 house construction peak that was followed 2 years later by the 2008 financial crisis is the insufficient analysis of the maybe 10 million illegals and how many were working in construction and lost work and even went back to their non-US homes. I don't trust US statistics on them for total hours nor for GDP.
Finally, Arnold claims "The latest numbers mean nothing".
This is totally false.
Biden and his Democratic administration have done a terrible job, and the latest numbers are merely more objective proof of Biden's lousy job and lousy results.
Arnold, somewhat inconsistently as compared to his measured but definite criticisms of Trump, refuses to evaluate Biden's lousy performance.
The GDP numbers, alone, don't show all of the problems, but show that there are some, especially in the Dem opposed energy businesses.
Most folk would usually prefer an economy with 2 jobs at $100,000 in energy rather than 4 jobs in fast food & delivery at $40,000, even with more hours in the later.
I'm not sure why anyone, including you, should buy your list of long-ago antecedent causes for today's inflation. You've said before that there's a huge amount of uncertainty in macro. Surely that uncertainty increases as one goes back a decade or more. Why should we think TARP or the ARRA have anything to do with today's CPI, aside from "butterfly flapping its wings"-style effects that can't be known even with hindsight?
I'd accept GDP has problems, but lets not throw the baby out with the bath water.
Alex Tabarrok has said GDP growth is the best measure of material well-being that we have. It's fine to say it doesn't work, but what's your alternative?
Certainly life is better in the Dominican Republic than in Haiti now, and that was really not the case 100 years ago. Surely its superior GDP growth over these years is not merely coincidence.
Other things being equal, when things are better, people trade more. And when people trade less, I doubt that things improved. So what are those other things? What conditions invalidate GDP comparison? You'd say culture shifts: A Pandemic; The Internet; War; The Sexual Revolution; Dechurching. Fine. But France, Germany, and Italy pretty much went through all that together after WWII, or at least post 1970. Is Italy's GDP stagnation really uninformative?
Overall, the philosophical issue buried in GDP dwarfs most of the technical issues - the idea of a single measure that can be a target is necessary to substantiate a common well-being. If you disaggregate scoring, you break up a team into competing identities. The question of what gets scored is thus a secondary question - whose interests will count when it comes time to make decisions.
"In last week’s post, I was not saying that paid-for child care is any sense “better” or “more right” than parent-provided child care."
In one of your responses in the comments to your post last week, you did seem to make this judgment by implying that paid-for child care is representative of a more "advanced" economy than parent-provided child care. Perhaps you didn't intend that to be a moral judgment, but I think many people would interpret it that way.
Is S&T the alternative to One Factory because in S&T there is no "unused capacity" or "potential" GDP? It's just not clear how One Factory and S&T differ exactly.
I'm having a bit of trouble wrapping my head around what you want to use GDP numbers for if not for comparison to prior periods at least in the short term, even if over the long term there are too many changes in the economic mix to make them comparable. I think the changes in short-term GDP would be more useful for that reason, maybe not as an absolute value but as a directional indicator.
Using your market-based definition of GDP, people moving resources in and out of the market over the short term would be a useful indicator of expectations. When people stop engaging in trade and hoard resources by doing things like cooking meals at home instead of eating in restaurants, the contraction in GDP as measured should be a flag to look for the reason why people feel that need. GDP growth in excess of natural expansion would be flag that inflationary pressures would be building the economy. I'm not really arguing that 'fine-tuning' would be a desirable objective, just more wondering about your rejection of GDP comparison over time given what you think GDP is measuring.
Almost all policy decisions are microeconomic. Is an X billion child tax credit a good idea? It does not depend on GDP. Rather it's a a microeconomic cost benefit analysis that includes the deadweight losses that financing this transfer will cause vs increasing the consumption of low income parents. Would shifting from wage taxation to a VAT for financing retirement and disability benefits, health insurance subsidies, and unemployment insurance? Again, not GDP, but the microeconomies of the deadweight losses of one tax vs the other.
I think the metaphor of GDP as marketed "production" is not bad. The point is, what policies can you get wrong by using it? Surely worrying existentially about whether we ARE or ARE NOT in a recession is not a good use of GDP. (Even apart from preliminary numbers being very unreliable) A NGDP futures market as a guide to monetary policy probably is.
My understanding is aggregates are comparable in the short term. Longer term, technological improvement and culture change results in the hand waving of "hedonic adjustments" and the comparisons are hypothetical.
We intuitively know what a well performing economy looks like: Economic transactions occur with with no obstacles or disruption. Perfect example of this is the economy of cars. How easily are consumers able to purchase a car? How readily are manufacturers able to produce the cars consumers want? The ideal, perfectly efficient economy doesn't exist. There are either product shortages or product sitting unsold. How inefficient is it?
Currently, the American economy for cars and houses and other products is not working well. A number ought to indicate this. And this number ought to show improvement when there is improvement in the production and purchasing of those products.
To extend.... think that all costs of the firm must be passed into prices in equilbrium. Assuming government diktat (regulatory, legislative etc) driven costs are roughly fairly applied into firms of like nature (assuming only limited levels of regulatory artbitrage are possible or temporary in nature only), then all costs loaded onto a firm will be passed to end consumers.
Effectively this is a kind of hidden taxation.
I work within a large firm, and when i do look at both my own job (which notionally has nothing to do with government issues and is the kind of job more regularly accused of the "evils of unfettered captialism") and then vast majority of roles around me..... well after you peel back the onion a few layers, some very very high % of time and work activities is not truly being devoted to satisfaction of a customer want*, but rather satisfaction of some government diktat.
*And even when my work output appears (on the face of it) to be directed towards a customer want.... some significant percentage of that customer want itself is only driven by a government diktat (i work in financial services, serving business/ enterprise type customers).
How does all this get measured? Its all a financial transaction that ends up in GDP statistics in various ways...... and only some % of it is truly related to any kind of improvement in human condition via voluntary trade, satisfaction of true underlying wants and needs
Sorry, Arnold. I regret your list of GDP pitfalls doesn't include a very important one. You ignore how Government expenditure is misused as a measure of its output. GDP accounting is not only about private output and its market valuation, it's also about government output but not about government redistribution of private income. National accounting pretends to provide good measures of an economy's output, income, and consumption but to a large extent it has failed because of the many government interventions. Unfortunately, public economics textbooks still fail to explain those interventions and their consequences. It's amazing how much time we spend on monetary economics and how little on public economics --yes, this is a good reason to laugh at experts on monetary economics and the army of economists employed by central banks. Although you may prefer to laugh at your government, pay attention to what happened in Argentina yesterday and may happen later today.
"President Biden just happened to be in office when the bonfire was lit."
Total COVID spending dwarves previous ordinary budget deficits by a massive margin. Adding up the entire ordinary deficit of the Trump years wouldn't come close to COVID spending. Had COVID spending not happened, we would not be experiencing this inflation.
If you believe Democrats were the primary drivers of lockdown socialism (fair) and that Biden is nothing more then an empty suit stand in for the Democrats (fair) then I think its fair to blame him for the inflation.
Great essay, Dr. Kling. Nothing in it I can disagree with very much, except for one item at the very end. I can quibble a bit about what isn't included, but that is for another day.
You wrote: "What is the value of the work you do cooking at home?"
Doesn't it show up, though, as spending on other goods and services, or as savings? What you don't spend eating out, gets spent/invested somewhere else. Isn't it a wash?
Finally, you wrote: "People don’t long for: Ford Pintos"
Speak for yourself- my first car in high school was a sky blue Ford Pinto. It had power nothing, no FM radio, but it was my first at age 16, and I would give a lot just get to drive it again.
I have $4.00 but no coffee. The local cafe has coffee but wants $$. We exchange.
This market exchange is evidence of a value add transaction that improves satisfaction (utility, happiness, whatever). I prefer coffee over $, they prefer $ over coffee. There is a deep inequality here -> i have consumer surplus, they have producer surplus.
Measurement of of VOLUNTARY transactions are a (imperfect) measurement of increases in satisfication. GDP sort of measures all transactions, assuming various equalities rather than inequalities (because this is measurable), but that means there is conservativism in the figures, and a few micro assumptions see major sources of inequalities eroded away due to assumed marginal choice effects (eroding consumer surplus of any one decision), competition effects (eroding producer surplus) etc etc.
The big issue though is that a very significant fraction (50%+) of measured economic activity is not actual voluntary but has greater or less coercive elements to it. Taxation is a clear coercive element (partially justified under various democratic and social contract theories that it is actually quasi-volantary hahahahah).....
But more pernicious and invidious is the creep of the regulatory state where "market activity" (i.e transactions that can be measured via tools such GDP) becomes largely driven by the underlying coercive threats (and carrots) from legislation/ regulations, not truly volunatary transactions. With certian interpretations, this could be 80%+ of all measured so-called market activity. By way of example, think of "regulatroy affairs" personnel at large regulated entities. Or entire divisions of people employed at large "private sector" firms whose job activities involve (say) 50-90% of work that is essentially responding to government decree.
All of this massively erodes the theoretical value of GDP as a potential measurement tool of increases in human utility improvement / satisfaction. GDP as a tool potentially started out with some correlations and relevence as measurement of utility, but has become diluted to merely being a measure of money transactions, with almost no relevent to being a measurement of improvement in utility/ satisifcation.
GDP star
What should GDP measure? Arnold asks, and does answer (in quotes)
Gross Domestic Product - "It is a measure of all of the goods and services that are purchased at prices measured in monetary units. ... explicit compensation in monetary units for providing something, that counts as GDP."
What it IS, today, is also what it SHOULD be measuring.
Tho this does not measure other things that are desired:
"we would like to measure total output of the economy," nor
"We also would like GDP to measure how well the economy provides human satisfaction, or what economists call social welfare" (many social scientists and the UN look for "quality of life")
GDP is the least flawed objective single number measure we have. Much like total SAT as a single number reference for likely college achievement. There's no other better alternative.
Tho it does seem FRED at $24.8 billion of GDP Q2 2022 disagrees with the $19.8 billion number.
https://fred.stlouisfed.org/series/GDP
"People want to use GDP as a way to “score” economic policy outcomes. I think that is particularly unwise."
They want to, they do, and, like political races being covered like horse races, they will continue to be used thusly.
Is the USA in a recession? Yes, yes it is - but the meaning of "recession" is less significant than GDP. [Recession - when neighbor loses his job.
Depression - when you lose your job.]
At least Arnold provides an alternative.
"the hours-worked number is more reliable than the GDP number, so my bet is that economic activity actually increased between the end of last year and the end of last month."
This might be the case, but after defining 2007 as 100, and calculating % from that so the new high in June 2022 is 112.4, the hours-worked index is not yet used as a key indicator, and there are probably more reasons not yet discussed to not use it.
At least one huge reason is the illegal immigrants, paid in cash to avoid taxes/ deportation/ min wage laws - how they calculated. A long time complaint I have about the 2006 house construction peak that was followed 2 years later by the 2008 financial crisis is the insufficient analysis of the maybe 10 million illegals and how many were working in construction and lost work and even went back to their non-US homes. I don't trust US statistics on them for total hours nor for GDP.
Finally, Arnold claims "The latest numbers mean nothing".
This is totally false.
Biden and his Democratic administration have done a terrible job, and the latest numbers are merely more objective proof of Biden's lousy job and lousy results.
Arnold, somewhat inconsistently as compared to his measured but definite criticisms of Trump, refuses to evaluate Biden's lousy performance.
The GDP numbers, alone, don't show all of the problems, but show that there are some, especially in the Dem opposed energy businesses.
Most folk would usually prefer an economy with 2 jobs at $100,000 in energy rather than 4 jobs in fast food & delivery at $40,000, even with more hours in the later.
I'm not sure why anyone, including you, should buy your list of long-ago antecedent causes for today's inflation. You've said before that there's a huge amount of uncertainty in macro. Surely that uncertainty increases as one goes back a decade or more. Why should we think TARP or the ARRA have anything to do with today's CPI, aside from "butterfly flapping its wings"-style effects that can't be known even with hindsight?
I'd accept GDP has problems, but lets not throw the baby out with the bath water.
Alex Tabarrok has said GDP growth is the best measure of material well-being that we have. It's fine to say it doesn't work, but what's your alternative?
Certainly life is better in the Dominican Republic than in Haiti now, and that was really not the case 100 years ago. Surely its superior GDP growth over these years is not merely coincidence.
Other things being equal, when things are better, people trade more. And when people trade less, I doubt that things improved. So what are those other things? What conditions invalidate GDP comparison? You'd say culture shifts: A Pandemic; The Internet; War; The Sexual Revolution; Dechurching. Fine. But France, Germany, and Italy pretty much went through all that together after WWII, or at least post 1970. Is Italy's GDP stagnation really uninformative?
Overall, the philosophical issue buried in GDP dwarfs most of the technical issues - the idea of a single measure that can be a target is necessary to substantiate a common well-being. If you disaggregate scoring, you break up a team into competing identities. The question of what gets scored is thus a secondary question - whose interests will count when it comes time to make decisions.
"In last week’s post, I was not saying that paid-for child care is any sense “better” or “more right” than parent-provided child care."
In one of your responses in the comments to your post last week, you did seem to make this judgment by implying that paid-for child care is representative of a more "advanced" economy than parent-provided child care. Perhaps you didn't intend that to be a moral judgment, but I think many people would interpret it that way.
Is S&T the alternative to One Factory because in S&T there is no "unused capacity" or "potential" GDP? It's just not clear how One Factory and S&T differ exactly.
I'm having a bit of trouble wrapping my head around what you want to use GDP numbers for if not for comparison to prior periods at least in the short term, even if over the long term there are too many changes in the economic mix to make them comparable. I think the changes in short-term GDP would be more useful for that reason, maybe not as an absolute value but as a directional indicator.
Using your market-based definition of GDP, people moving resources in and out of the market over the short term would be a useful indicator of expectations. When people stop engaging in trade and hoard resources by doing things like cooking meals at home instead of eating in restaurants, the contraction in GDP as measured should be a flag to look for the reason why people feel that need. GDP growth in excess of natural expansion would be flag that inflationary pressures would be building the economy. I'm not really arguing that 'fine-tuning' would be a desirable objective, just more wondering about your rejection of GDP comparison over time given what you think GDP is measuring.
Almost all policy decisions are microeconomic. Is an X billion child tax credit a good idea? It does not depend on GDP. Rather it's a a microeconomic cost benefit analysis that includes the deadweight losses that financing this transfer will cause vs increasing the consumption of low income parents. Would shifting from wage taxation to a VAT for financing retirement and disability benefits, health insurance subsidies, and unemployment insurance? Again, not GDP, but the microeconomies of the deadweight losses of one tax vs the other.
I think the metaphor of GDP as marketed "production" is not bad. The point is, what policies can you get wrong by using it? Surely worrying existentially about whether we ARE or ARE NOT in a recession is not a good use of GDP. (Even apart from preliminary numbers being very unreliable) A NGDP futures market as a guide to monetary policy probably is.
My understanding is aggregates are comparable in the short term. Longer term, technological improvement and culture change results in the hand waving of "hedonic adjustments" and the comparisons are hypothetical.
We intuitively know what a well performing economy looks like: Economic transactions occur with with no obstacles or disruption. Perfect example of this is the economy of cars. How easily are consumers able to purchase a car? How readily are manufacturers able to produce the cars consumers want? The ideal, perfectly efficient economy doesn't exist. There are either product shortages or product sitting unsold. How inefficient is it?
Currently, the American economy for cars and houses and other products is not working well. A number ought to indicate this. And this number ought to show improvement when there is improvement in the production and purchasing of those products.
To extend.... think that all costs of the firm must be passed into prices in equilbrium. Assuming government diktat (regulatory, legislative etc) driven costs are roughly fairly applied into firms of like nature (assuming only limited levels of regulatory artbitrage are possible or temporary in nature only), then all costs loaded onto a firm will be passed to end consumers.
Effectively this is a kind of hidden taxation.
I work within a large firm, and when i do look at both my own job (which notionally has nothing to do with government issues and is the kind of job more regularly accused of the "evils of unfettered captialism") and then vast majority of roles around me..... well after you peel back the onion a few layers, some very very high % of time and work activities is not truly being devoted to satisfaction of a customer want*, but rather satisfaction of some government diktat.
*And even when my work output appears (on the face of it) to be directed towards a customer want.... some significant percentage of that customer want itself is only driven by a government diktat (i work in financial services, serving business/ enterprise type customers).
How does all this get measured? Its all a financial transaction that ends up in GDP statistics in various ways...... and only some % of it is truly related to any kind of improvement in human condition via voluntary trade, satisfaction of true underlying wants and needs