Does anyone know of any efforts to measure total consumers’ surplus in the U.S. economy? What do you think it would be relative to GDP?
a) a fraction of GDP
b) about equal to GDP
c) a multiple of GDP
Justify your answer.
16 thoughts on “Total Consumers’ Surplus”
The question doesn’t quite work because “consumer surplus” is a partial equilibrium concept. Calculated consumer surplus is highly sensitive to elasticity and therefore market definition. The sum of consumer surplus for a bunch of narrowly-defined markets is less than the consumer surplus for a broadly-defined market that encompasses only those items (usually), because of substitution effects. Suppose a market has an always-inelastic demand curve such as Q=P^-0.5; then consumer surplus is always infinite no matter the price. You could interpret this to mean that such a demand curve does not exist, or you could say that in some sense “demand for the economy” approaches “always inelastic” since “the economy” is just a very broadly-defined market. If you are comfortable with the latter then the answer is (c), but there are a lot of holes in this method of reasoning.
Maybe out of topic but I recall vaguely that under the assumption of Cobb-Douglas utility functions the compensating variation is equal to change in real GDP, ie the change in real total income.
I agree with Eli, though I’m not sure I understand his reasoning. As I see it, Consumer Surplus can only be defined relative to a counterfactual.
I turn on my sink and drink a glass of water. What’s my consumer surplus? I’ve paid essentially $0.00, but how do I measure the utility? Is the counterfactual civic water vs. well pumped? indoor plumbing vs a private/communal well? In the limit where I have no alternate source of water, the consumer surplus is nearly infinite- I’ll pay anything rather than die of dehydration.
Suppose that we take as the baseline “the U.S. economy in 1970.” That is, the counterfactual is that the economy had kept the same GDP and consumers’ surplus in 1970. Now, we take the actual GDP and consumers’ surplus in 2014. Has the increase in consumers’ surplus been (a), (b), or (c)?
One microcosm of this is to ask- what’s the consumer surplus from that most modern of technologies, cell phones?
As a software engineer, my consumer surplus is from smartphones is enormous. I pay about $100/month for this service. I probably value my data+telephony service at upwards of $400/month, for a consumer surplus of $3600/year.
But that’s me. Clearly not everyone enjoys the same surplus. 10% of US adults still do not have cell phones, implying they value it at less than $100/month (and in fact less than $20/month which is ~ the lowest entry price point).
The adoption/price curve over time [I can’t find any good online references, but if memory serves, widespread cell phone adoption came around the time carriers started offering prices around $0.10/min which is not that much higher than current prices]. This suggests that consumer surplus is rather low, since it took large price declines to spur adoption. This implies that the utility is rather close to the selling price.
There are a couple reasons why the adoption curve might not be indicative of the demand curve:
1) Non-price factors might drive adoption (not everyone knows how great cellphones are).
2) There are network effects which increase the per-capita consumer surplus at the same price as more consumers adopt the technology.
3) Consumers were mistaken is assessing the pre-purchase value (i.e. they didn’t think it was worth $50 but having purchased found it was actually worth $500).
As far as I know, essentially no one, having bought a cell phone, decides it’s not worth it and cancels their service. Further, no one, having paid for a data plan, decides it’s not worth it and cancels. This seems to hold true even through adverse wealth shocks (2007-2009 seems to have had little impact on adoption) or personal wealth shocks (do the recently unemployed cancel their cell phones or data plans? My sense is no one does this, but perhaps that’s my social bubble).
Since it’s a repeated game- users re-up each month, year, or at worst bi-yearly, and no one defects, this implies a large consumer surplus which is at odds with the simple facts of the adoption/price curve. So I tend to explanations #2 and #3, meaning that consumer surplus for this product is large (>2-3x measured price/GDP).
How relevant is this to the larger economy? I’m not sure, but I suspect that this pattern is fairly common. Consumers are reluctant to adopt technologies, but having adopted them, find the consumer surplus quite large.
I’m going to go with (a) for two reasons.
1. Producer Sophistication has increased: There is a huge incentive for suppliers to market-segment and price-discriminate, and they are arguable better than ever at identifying individualized demand curves and adaptively and iteratively solving for ideal strategies. Thing of prices for flights, or cell phone plans, etc. They should know more about how to extract more from those willing to pay more.
2. The largest component of personal consumption expenditures is housing which is sui generis. There is no giant ‘shelter-supply’ producing company. Instead, there are many millions of individuals owners that constitute the buyers and sellers in a fairly efficient auction-style market. The winning bids in efficient auctions should have approximately zero consumer surplus by definition, so long as the housing market and neighborhoods remain highly differentiated with regards to ‘quality’ – which they most definitely are.
Yes, producers gain from market segmentation.
Consumer pay a 1.8% in inflation each year, more or less. Their surplus is negative 1.8%, generally.
Since the seventies I have the consumer deficit, inflation, decreasing. It likely decreased by an amount equal to GDP, and that is the amount of productivity gain. Do I have a geater diversity of goods available? Sure, but I still consume the same amount of goods, and pay with my inflation losses which are reduced. But it is still inflation, a reduction in the planned flow of goods due to unplanned losses.
The mistake to equate the larger diversity of consumers and the larger diversity of goods. I still listen to the same amount of music, and my tastes are just as specialized. It is just that my consumer search is shortened since the seventies.
Every American would pay his entire income for the necessities of life. Suppose, for the sake of argument, that the average person gets that for 10% of actual income. That means the consumer surplus is 90%.
But if that’s the case, when I buy a coffee for $1 for which I would have paid $3, where does that “extra” $2 in consumer surplus come from? Shouldn’t it already be in the 90%?
What’s the solution? Does consumer surplus only make sense at the margin, holding everything else constant? In that case, you can’t really add it up, right?
Still thinking about this …
My answer is C, a multiple of GDP. This is the case as well if considering the 1970 counter factual. Here is why:
The expansions of scope, scale, and quality of goods and services has vastly expanded and done so faster than the expansion of producer revenues. Producers abilities to capture these gains look poor when we consider how much more leisure consumers are now able to afford. Producer profitability has been as much driven by bending cost curves down as it has price discrimination. Therefore, profit growth is multicausal and misleading.
I think it’s a hopeless concept to measure. Eli’s point is an excellent one. In general though, I think of it this way:
1) GDP is (whenever possible) valued at a market price and Market price = marginal utility < total utility (diminishing marginal utility). So total utility is larger than GDP for all things that are purchased in well functioning markets. This points to C.
2) Leisure isn't included and that is going up. So is all kinds of essentially free entertainment on the web. Youtube is also making home production more valuable. This also points to C.
3) There is a growing fraction of stuff in GDP that is valued at "prices" from poorly functioning markets (education, health care, and much of government spending). This points to A.
On the balance I'm fairly confident that it's C because I can think of many more things that are a great value than things that are a terrible value.
Wonderful question, Arnold!
My answer is C. And I would add that the “multiple” of GDP steadily increases.
Justification:
Much of what I own continues to provide me with Consumer Surplus long, long after its purchase price registers in national income accountancy (GDP).
For example, my home heating system purchase registered in 1973 GDP. In the intervening 40 years, some additional 20% of the original purchase price has been spent on maintenance, registering in national income accounting in those years, as of course did my purchases of fuel/energy. Two weeks ago, my geographic location in the U.S. experienced a full week of sustained dead-air temperatures well below zero – my heating system, registering in national income accounting in 1973, provided me with ENORMOUS Consumer Surplus in February 2014. And that Consumer Surplus, realized (at least) in February 2014, will not be included in 2014 national income accounting.
Am I misrepresenting the concept of Consumer Surplus here in any way?
One of the most concise, powerful and important statements I’ve seen you post over the years Arnold, is that, “The U.S. economy is not just a GDP factory.” I’d like to hope that more “economists” would realize that as being true.
I think PJ’s comment is especially a propos:
“Leisure isn’t included and that is going up.”
Since more stuff/less work is really what we’re after, right?
What if we added $X/hour for every waking hour not worked? Should we add it to both consumer surplus and GDP (call it Gross Domestic Prosperity?), or just one or the other? How would the relative measures change?
Which raises the question: why are conservatives so desperate for people to work more?
Here is a related question that I’ve been pondering.
As you are probably aware, Paul Newman has a food line, “Newman’s own”, which is quite popular and all profits are donated to charity.
Economically speaking, which provides a greater benefit to humanity?
a) The charitable causes that are paid for with the profits? or
b) The consumer surplus generated via the sale of the goods?
Professor:
My answer is (C) based strictly on necessities of life: food, clothing, and shelter.
In America, we are probably in the elastic part of the demand curve for food. But there is a very inelastic part where the demand curve is far above the price, resulting in huge consumer surplus.
Same with clothes.
Housing might be considered different, especially single-family homes. But half the population lives in apartments where the auction process does not result in elimination of consumer surplus. I would guess that there is a pretty big consumer surplus there, too.
Looking at this question from the perspective of economic history is useful, because until the 1700’s, most people spent nearly 100% of their income on food, clothes, and housing. The world became wealthy mainly by making two of those three things very, very cheap.
The question doesn’t quite work because “consumer surplus” is a partial equilibrium concept. Calculated consumer surplus is highly sensitive to elasticity and therefore market definition. The sum of consumer surplus for a bunch of narrowly-defined markets is less than the consumer surplus for a broadly-defined market that encompasses only those items (usually), because of substitution effects. Suppose a market has an always-inelastic demand curve such as Q=P^-0.5; then consumer surplus is always infinite no matter the price. You could interpret this to mean that such a demand curve does not exist, or you could say that in some sense “demand for the economy” approaches “always inelastic” since “the economy” is just a very broadly-defined market. If you are comfortable with the latter then the answer is (c), but there are a lot of holes in this method of reasoning.
Maybe out of topic but I recall vaguely that under the assumption of Cobb-Douglas utility functions the compensating variation is equal to change in real GDP, ie the change in real total income.
I agree with Eli, though I’m not sure I understand his reasoning. As I see it, Consumer Surplus can only be defined relative to a counterfactual.
I turn on my sink and drink a glass of water. What’s my consumer surplus? I’ve paid essentially $0.00, but how do I measure the utility? Is the counterfactual civic water vs. well pumped? indoor plumbing vs a private/communal well? In the limit where I have no alternate source of water, the consumer surplus is nearly infinite- I’ll pay anything rather than die of dehydration.
Suppose that we take as the baseline “the U.S. economy in 1970.” That is, the counterfactual is that the economy had kept the same GDP and consumers’ surplus in 1970. Now, we take the actual GDP and consumers’ surplus in 2014. Has the increase in consumers’ surplus been (a), (b), or (c)?
One microcosm of this is to ask- what’s the consumer surplus from that most modern of technologies, cell phones?
As a software engineer, my consumer surplus is from smartphones is enormous. I pay about $100/month for this service. I probably value my data+telephony service at upwards of $400/month, for a consumer surplus of $3600/year.
But that’s me. Clearly not everyone enjoys the same surplus. 10% of US adults still do not have cell phones, implying they value it at less than $100/month (and in fact less than $20/month which is ~ the lowest entry price point).
The adoption/price curve over time [I can’t find any good online references, but if memory serves, widespread cell phone adoption came around the time carriers started offering prices around $0.10/min which is not that much higher than current prices]. This suggests that consumer surplus is rather low, since it took large price declines to spur adoption. This implies that the utility is rather close to the selling price.
There are a couple reasons why the adoption curve might not be indicative of the demand curve:
1) Non-price factors might drive adoption (not everyone knows how great cellphones are).
2) There are network effects which increase the per-capita consumer surplus at the same price as more consumers adopt the technology.
3) Consumers were mistaken is assessing the pre-purchase value (i.e. they didn’t think it was worth $50 but having purchased found it was actually worth $500).
As far as I know, essentially no one, having bought a cell phone, decides it’s not worth it and cancels their service. Further, no one, having paid for a data plan, decides it’s not worth it and cancels. This seems to hold true even through adverse wealth shocks (2007-2009 seems to have had little impact on adoption) or personal wealth shocks (do the recently unemployed cancel their cell phones or data plans? My sense is no one does this, but perhaps that’s my social bubble).
Since it’s a repeated game- users re-up each month, year, or at worst bi-yearly, and no one defects, this implies a large consumer surplus which is at odds with the simple facts of the adoption/price curve. So I tend to explanations #2 and #3, meaning that consumer surplus for this product is large (>2-3x measured price/GDP).
How relevant is this to the larger economy? I’m not sure, but I suspect that this pattern is fairly common. Consumers are reluctant to adopt technologies, but having adopted them, find the consumer surplus quite large.
I’m going to go with (a) for two reasons.
1. Producer Sophistication has increased: There is a huge incentive for suppliers to market-segment and price-discriminate, and they are arguable better than ever at identifying individualized demand curves and adaptively and iteratively solving for ideal strategies. Thing of prices for flights, or cell phone plans, etc. They should know more about how to extract more from those willing to pay more.
2. The largest component of personal consumption expenditures is housing which is sui generis. There is no giant ‘shelter-supply’ producing company. Instead, there are many millions of individuals owners that constitute the buyers and sellers in a fairly efficient auction-style market. The winning bids in efficient auctions should have approximately zero consumer surplus by definition, so long as the housing market and neighborhoods remain highly differentiated with regards to ‘quality’ – which they most definitely are.
Yes, producers gain from market segmentation.
Consumer pay a 1.8% in inflation each year, more or less. Their surplus is negative 1.8%, generally.
Since the seventies I have the consumer deficit, inflation, decreasing. It likely decreased by an amount equal to GDP, and that is the amount of productivity gain. Do I have a geater diversity of goods available? Sure, but I still consume the same amount of goods, and pay with my inflation losses which are reduced. But it is still inflation, a reduction in the planned flow of goods due to unplanned losses.
The mistake to equate the larger diversity of consumers and the larger diversity of goods. I still listen to the same amount of music, and my tastes are just as specialized. It is just that my consumer search is shortened since the seventies.
Every American would pay his entire income for the necessities of life. Suppose, for the sake of argument, that the average person gets that for 10% of actual income. That means the consumer surplus is 90%.
But if that’s the case, when I buy a coffee for $1 for which I would have paid $3, where does that “extra” $2 in consumer surplus come from? Shouldn’t it already be in the 90%?
What’s the solution? Does consumer surplus only make sense at the margin, holding everything else constant? In that case, you can’t really add it up, right?
Still thinking about this …
My answer is C, a multiple of GDP. This is the case as well if considering the 1970 counter factual. Here is why:
The expansions of scope, scale, and quality of goods and services has vastly expanded and done so faster than the expansion of producer revenues. Producers abilities to capture these gains look poor when we consider how much more leisure consumers are now able to afford. Producer profitability has been as much driven by bending cost curves down as it has price discrimination. Therefore, profit growth is multicausal and misleading.
I think it’s a hopeless concept to measure. Eli’s point is an excellent one. In general though, I think of it this way:
1) GDP is (whenever possible) valued at a market price and Market price = marginal utility < total utility (diminishing marginal utility). So total utility is larger than GDP for all things that are purchased in well functioning markets. This points to C.
2) Leisure isn't included and that is going up. So is all kinds of essentially free entertainment on the web. Youtube is also making home production more valuable. This also points to C.
3) There is a growing fraction of stuff in GDP that is valued at "prices" from poorly functioning markets (education, health care, and much of government spending). This points to A.
On the balance I'm fairly confident that it's C because I can think of many more things that are a great value than things that are a terrible value.
Wonderful question, Arnold!
My answer is C. And I would add that the “multiple” of GDP steadily increases.
Justification:
Much of what I own continues to provide me with Consumer Surplus long, long after its purchase price registers in national income accountancy (GDP).
For example, my home heating system purchase registered in 1973 GDP. In the intervening 40 years, some additional 20% of the original purchase price has been spent on maintenance, registering in national income accounting in those years, as of course did my purchases of fuel/energy. Two weeks ago, my geographic location in the U.S. experienced a full week of sustained dead-air temperatures well below zero – my heating system, registering in national income accounting in 1973, provided me with ENORMOUS Consumer Surplus in February 2014. And that Consumer Surplus, realized (at least) in February 2014, will not be included in 2014 national income accounting.
Am I misrepresenting the concept of Consumer Surplus here in any way?
One of the most concise, powerful and important statements I’ve seen you post over the years Arnold, is that, “The U.S. economy is not just a GDP factory.” I’d like to hope that more “economists” would realize that as being true.
I think PJ’s comment is especially a propos:
“Leisure isn’t included and that is going up.”
Since more stuff/less work is really what we’re after, right?
What if we added $X/hour for every waking hour not worked? Should we add it to both consumer surplus and GDP (call it Gross Domestic Prosperity?), or just one or the other? How would the relative measures change?
Which raises the question: why are conservatives so desperate for people to work more?
Here is a related question that I’ve been pondering.
As you are probably aware, Paul Newman has a food line, “Newman’s own”, which is quite popular and all profits are donated to charity.
Economically speaking, which provides a greater benefit to humanity?
a) The charitable causes that are paid for with the profits? or
b) The consumer surplus generated via the sale of the goods?
Professor:
My answer is (C) based strictly on necessities of life: food, clothing, and shelter.
In America, we are probably in the elastic part of the demand curve for food. But there is a very inelastic part where the demand curve is far above the price, resulting in huge consumer surplus.
Same with clothes.
Housing might be considered different, especially single-family homes. But half the population lives in apartments where the auction process does not result in elimination of consumer surplus. I would guess that there is a pretty big consumer surplus there, too.
Looking at this question from the perspective of economic history is useful, because until the 1700’s, most people spent nearly 100% of their income on food, clothes, and housing. The world became wealthy mainly by making two of those three things very, very cheap.
Max