So if you give all your coders a great test harness and they run more tests because it takes up less of their time, that's not the Jevons paradox. If you give your coders a great test harness and then they go from spending 10% of work hours on testing to spending 20% of work hours on testing because testing has such a good ROI now, that's the Jevons paradox.
Meanwhile real Jevons is a very particular effect that doesn't show up on all road graphs and doesn't simply mean that opening up more lanes will attract more traffic.
People scoff at road construction (or rather, road widening) as a solution to congestion. I don't think anyone is under the impression that kilometers driven would go down as road bandwidth goes up.
>The road capacity led to more cars on the road, but also to more real demand being met.
No, not necessarily. Maybe some people who would have otherwise used public transport opt to drive instead. Making roads wider could literally make them less efficient, in terms of humans moved per hour per meter of width.
It is a solution to congestion because a part of the past congestion was that people were stuck at home and gave up on the trip overall, because it would take so long.
> Maybe some people who would have otherwise used public transport opt to drive instead.
Then presumably comfort increased. Also, add bus lanes.
That's not congestion. If someone opts not to make a trip, or uses a means of transport that doesn't use the road, such as a subway, then they didn't create road traffic. Congestion refers specifically to the failure of traffic to advance at an efficient speed along a thoroughfare.
>Then presumably comfort increased.
Again, "people scoff at road construction as a solution to congestion". Discomfort is not a component of congestion.
As the cost of attention goes down, and the cost of communicating bad ideas goes down, capturing attention with bad ideas paradoxically goes up.
I would argue that is not Jevons paradox but standard supply and demand (and this "reverse Jevons paradox" too). Jevons paradox occurs when a more efficient use of a resource leads to an increase in its use (instead of a decrease as a first order analysis would suggest).
Jevon's paradox is a special case of supply and demand, where people actually end up spending more money because something is cheaper.
It's interesting because consumption then grows in unpredictable ways: it can drive innovation even in cases where markets are constrained by monopolies, for example, where in non-Jevons cases producers would have no incentive to lower prices.
No, this isn't what Jevons says. Jevons isn't concerned about money being spent on something, just total consumption of it. Whether more money or less is spent on it depends on the price elasticity of demand. Inelastic demand will lead to less money spent despite more of the resource/service/whatever being consumed. Elastic demand will lead to more money spent.
What would the food example be? Vanilla ice cream going from a rarity only the rich could afford, to a standard desert for all when the synthetic form was invented?
The example I've heard given is accounting and spreadsheets. It made accountancy cheaper, but people then started asking more questions and analysis became a thing.
Rather than just taking the reduced spend as profit, companies wound up increasing their accountancy spend overall.
Kinda yes. How do you derive total spend from supply-demand curves? Multiply price and quantity at an intersection point. Likewise, you can predict total spend by multiplying p and q on the demand curve.
The difference in total spend is difference between these areas. For the total spend to increase with a drop in price, the the demand must rise faster.
Jevon's paradox implies that the price equilibrium is at the highly elastic portion of the demand curve.
> Jevons paradox occurs when a more efficient use of a resource leads to an increase in its use
While that's mostly true in practical reality in established economies, that does not strictly have to be the case. On the consumer side, especially in manufacturing, there's very little difference between unit price of a good falling and input unit per output units dropping as both lead to decreased COGS. In both cases, market realities might unlock alternative approaches (the classic being robot replacing Robert), leading to increased demand.
I think these are the same, because efficiency is value over cost. In the original formulation of the paradox, a more efficient steam engine lead to a rise in coal consumption. You can look at this as a "money buys coal, coal drives locomotion" system, where the latter part was improved. Modulo practical issues with coal (transport, storage, etc), dropping the price of coal would (probably?) lead to the same effect, since the end result is that locomotion per money is increased. For an outside observer, it doesn't matter if you get more coal per money or more locomotion per coal.
> standard supply and demand
Standard supply and demand doesn't say anything about increase of spend. If food prices drop, I'm not going to buy more food. I might buy better food for the same budget, but there's no reason why my total food spend should increase.
Not for perishables, but for non-perishables.
Imagine yourself a shop keeper, hoping to boost the money coming in at the till. If you increase prices by 10%, you will get more? Right?
That depends on the elasticity of demand. If the elasticity is two, the drop in demand is twice the increase in price. 0.8 times 1.1 is 0.88. Takings fall from $100 to $88.
But if the elasticity is one half, the drop in demand is half the increase in price. 0.95 times 1.1 is 1.045. Takings rise from $100 to $104.5.
When the price goes up the shop always sells less goods, (Law of Demand) but that still leaves it unclear whether more or less money goes in the till. This is first year University economics today.
Back in 1865, it was obvious to every-one that the increased efficiency of steam engines would lead to a reduced demand for coal. Jevons pointed out that increased efficiency makes steam power cheaper. Goodbye water wheel, hello steam engine. More steam engines, greater consumption of steam power, any-one who wants to make a prediction needs to invent the concept of elasticity and try to measure it. Greater than one? Less than one? That is going to decide whether total demand rises or falls.
Well, in that case you're talking about an entirely different phenomenon. Jevons's paradox (as defined by this post) happens when the cost of a resource decreases and the spend on that resource increases. You're talking about what happens to the spend on resource A when the cost of resource B decreases. Whether it rises or falls, it won't be Jevons's paradox.
The law of demand frames demand as a function of price and utility — demand is monotonically non-decreasing with utility (the more useful it is, the more people want it), and monotonically non-increasing with price (the pricier it is, the less people want it), but e.g. Giffen goods and Veblen goods break the "monotonically non-increasing with price" assumption of the law of demand.
You can add efficiency to that equation — demand is a function of price, utility and efficiency, and it is also monotonically non-increasing efficiency (The less of it you need, the less people want it). If you could get twice as much saltiness from table salt, you'd cut down demand by 50%.
The Jevons paradox is about the cases where demand isn't non-increasing with efficiency, because utility is itself a function of efficiency. Increased efficiency directly lowers demand, but, because it increases utility, it also increases demand indirectly.
The paradox is usually framed as more efficiency -> more demand (because of the intermediate "more utility" step), but the author is framing it in the opposite direction, as less efficiency -> less demand (because of the intermediate "less utility"). I would argue it's just that the paradox works both ways, rather than calling it a "reverse", but that's me.
No you are right, it's the same paradox, not some distinct reverse form. If more x can cause more y, then it's logically equivalent that less x can cause less y.
An example not anchored in anything: If public transit costs x, I'll use it every day. If public transit suddenly costs 2x, I'm not gonna use it every other day, I'll rather find an alternative and use 0.
There is a class of changes that take very little time but have a positive impact. If the cost of making a ticket for that change exceeds the cost of the ticket, it's human nature that some people just won't make the change.
This is a bit of a contrived example (e.g. you could bundle multiple small changes into one ticket) but it's still a good example of a policy meant to make things better actually leading to fewer improvements.
Or the changes might be "smuggled through" in an unrelated changeset that has to go through the red tape anyway.
- Dad, dad, have you seen - the prices went up so much! Does this mean you will stop drinking now?
- No son, this means you will eat less.
> If you pay $1.00 to press a button, and pressing the button pays you $0.99, you > will press the button zero times. If you get $1.01 instead, you will press the > button all the time.
One thought: If it only gives you $0.01 profit you'll (have to) push it more than if it gave you $1000 profit. There's a saturation point.
That's called Software Engineering at Google