Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Reelya]

Bills are every 3 months where I live, and Wierd was assuming yearly. That's why I just said billing period.

But it's not really addressing the concept then, if we argue about how long lightbulbs last.

However I still think the thing about the $3 being electricity is weird since the original statement did list that as a separate thing to what happens to the electricity bill.

[smjjames]

I was assuming monthly for the billing since thats how its set up here, but the timespan of the billing period is irrelevant for the thought experiment.

[Reelya]

It's not important to the point however.

I am annoyed however because wierd didnt get what i meant, which is fair enough. However 've repeatedly and consistently clarified the same thing i meant all along. And wierd is just doubling down on not getting it rather than adressing what i clearly did mean in the first place.

And by clearly, i mean the same thing i pointed out a good dozen times now. Sure first post, not getting it is excusable, my fault for not clarifying enough. But if you say "actually i meant xyz" and people double down on the addressing the erroneous interpretation, based on constantly changing questionable reinterpretations of the first post and ignoring all clatifications, then thats just bring obtuse now.

[wierd]

Agreed, billing period is only relevant if it is some really long period, and we intend to move, for the purposes of this thought experiment.

However, it is still a false comparison to have a "licensing fee" in item 1, and "subsidized cost" in item 2. 

[Reelya]

I don't agree with that really.

If all costs and benefits are per unit time then the rational decision doesnt change, however finely or coarsely you slice time. It's not a variable then whatever label you give to the time each billing period represents. That was the reason I initially said "$3 to run" per billing period. It abstracts what a billing period is out of the equation. "Billing period" could be 1 minute or 5 years. It's irrelevant when all costs and gains are defined as being per-time-unit.

And that's why I talk about the "lifetime savings" of a light-bulb when expressing it in terms of the purchase price of the light-bulb. If you're talking about per-month, then the cost of the choice needs to also be expressed per-month, and if you're talking about upfront product price, then the energy savings need to be expressed relative to the lifetime of the product. That's just to ensure that you're comparing apples and apples and that time can be factored out as not relevant to the argument.

And why to that last point? The example with a long-life bulb in your bedroom and shared bills is 100% a real situation that can occur. It's no different to what i meant in the original question mathematically.

The point is: if a cool lightbulb is $30 and saves $40 of power over it's lifespan (which is therefore a time-invariant decision) and bills are shared with even one other person, then there's zero economic incentive for either one person to get one and put it in their room. You spend $30 and get only $20 back. So it fails the "rational choice" test because you chose the option that left you worse off. In fact, it doesn't matter how many light bulbs you replace with better ones: if only you are paying $30 for each of them, while each one saves $40 and the other guy pockets $20 in savings per light bulb replaced, you never in fact "break even" on this deal, no matter if you replace every light bulb in the house by yourself.

This is a classic prisoner's dilemma from Game Theory actually: individual decisions which makes no cost/benefit sense, but if taken by multiple people at once give benefits to everyone. However, even if you get agreement that the lightbulbs should be upgraded, cheaters are still rewarded. e.g. if one person reneges on their share of the upgrades, then they reap more personal benefit than they would by cooperating. It explains part of why people make decisions which look "bad" as a whole: it's because they are actually the optimal personal decision to make, when taken in isolation.

[Maximum Spin]

Now that this whole debate is out of the way, let me let you all in on the dark secret.

If you believe that your roommates roughly share your thought process - that is, if all participants are rational and are known to be rational - then the rational thing to do is buy the widget knowing that everyone else also will and everyone will thus save money.

There is a common misconception that Nash equilibria represent the best possible outcomes of a rational actor. This is false. Nash equilibria represent the best possible outcomes of a rational actor surrounded by idiots; a group of rational actors capable of modelling one another's rational decisions can and will do better. And humans are indeed capable of this, which is why tests of the prisoners' dilemma repeatedly find that humans actually cooperate. The "rationality" Reelya defines here is actually profoundly irrational thinking that I've generally noticed is most characteristic of certain types of autistic, who demonstrate it because they do indeed have deficits in abstractly simulating other human minds.

[Dorsidwarf]

I don't get where all the arguing and bitching about the problem in this thread came from, the problem as clearly stated seems self-evident to me, unless you jump to conclusions and manufacture parts of the problem which aren't there. And as any GCSE student could tell you, reading the question through carefully to make sure you haven't misinterpreted it is really basic stuff.

I have a box into which I feed $3 a month. In return, this theory box saves the apartment $4 of electricity per month. It is illogical of me to buy this box, because I only make any money from it if all 3 of the other residents also buy and continue to run their boxes, a situation highly unlikely to arise considering how communication in this thought experiment isn't permitted.

Edit: in fact with perfectly self-interested actors intent on maximising their own reward, this situation can't be stable, since it is in the interest of all fourtenants to be the only tenant without a device. (I use self interested instead of rational because I don't actually know what the official economic description of a "rational actor" is.)

[Helgoland]

There is a common misconception that Nash equilibria represent the best possible outcomes of a rational actor. This is false. Nash equilibria represent the best possible outcomes of a rational actor surrounded by idiots; a group of rational actors capable of modelling one another's rational decisions can and will do better.

You've implicitly made a shift from one-time games to games over multiple periods. Applying Reels's (intended) scenario there, you could for example buy an LED once, hope the rest of the house notices the reduced consumption, does the math, and switches to efficient bulbs as well. Part of your strategy however should be a plan for when that does not in fact happen, which should result in you not buying more LEDs either.

I think that Reels was originally considering the scenario as a one-time game, so the solution he gives is indeed correct. For a (n infinitely) repeated game it should come down to something like Grim Trigger (the Nash equilibrium, IIRC) and Tit-for-tat (interestingly not a Nash equilibrium, IIRC - there might be some mathematical weirdness about infinite series going on there).

[Reelya]

Helgoland has a good point however, that if you invest in your own energy-saving devices then the other tenants need to notice that you did. Rational actor theory requires the actors to have complete knowledge of the variables involved. With only partial information you can only optimize your own part of it.

I've lived in boarding houses with shared electric bills, trust me none of the other tenants give a fuck that they are consuming more than the others. This is in fact the normal way people interact with people they aren't friends with. It's about trust. You end up doing things such as locking your plates and cutlery in your room because the other tenants cannot be trusted to let you leave your things in the kitchen. One guy was a constant toilet-paper thief so I ended up with a small carry case in which I kept toilet paper and soap and only brought it into the bathroom on an "as needed" basis. If he only used some toilet paper it wouldn't have been a problem, but I'd put a new roll on at 8am and it would be 100% gone by the time I was home around 3pm. Over and over. I came to believe he was just switching it out with an empty roll. Seriously, we're going to talk about how regular people make good decisions relating to how they live with others in a non-friends situation, when I've had the real lived experience of having a weaponized toilet paper Cold War? (edit: actually come to think of it, there was a similar toilet paper situation in an earlier boarding place too).

This has been my *normal* experience of boarding houses where you don't get to pick and choose who is in the other rooms. And I'm not talking autistic geeks here: this is my experience of boarding houses which had "regular joes" living there. The geeky weirdos I've lived with were all more tuned in to their impact on those around them that the "outgoing jocks".

e.g. weren't we saying before that it was wrong to factor in money you'd get from your job for having a car into the car loan / saving up decisions? Now we're supposed to factor in helping the environment and money that other people will be saving into the "energy saving appliance" question? It seems like having too low a bar on the one question and too high on the other.

[smjjames]

Maybe a better example would be a shared apartment or dorm house with multiple people? Not completely getting the foundation of the thought experiment. I mean, I get the thought experiment, but don't know what it was in response to. *goes back and reads pages before I jumped in*

[Reelya]

One thing about the TCO thing or car loans costing more than the up front costs, those aren't necessarily economically irrational or short-sighted choices.

It was in relation to the discussion on what rational choices look like. You see, people were saying that "people don't make rational decisions" because the overall outcomes of those decisions were bad. If that's the case, then "educate the masses" is the simplistic solution to the "bad economic decisions" problem. However, the Prisoner's Dilemma type of thought experiments show that this is flawed thinking. The problem is that individual choices which are optimal given perfect information don't always lead to a global optimal solution.

In fact, it can be quite possible that for each decision the "delta value" of utility for the bad decision is the optimal delta value, thus at every point if you're trying to maximize utility you take that action again. However, paradoxically, each choice adds up to the worst possible outcome for the system as a whole. And this is the reality, more than the "people just aren't educated" scenario. It's not a matter of people not being educated, it's a matter that there are fundamental contradictions / "perverse incentives" in our real-world economic systems.

~~~

the point of the scenario isn't the particulars, drilling down into the specifics is a clear red herring. It's like arguing that the person in Searle's Chinese Room experiment will die before they can respond because the needed book is too large. The objection entirely misses the crux of the matter and takes the debate into a plainly irrelevant direction.

Or it's like being given one of those "two people get on a bus, then at the next stop, 1 person gets off". And then objecting "hold on hold on, is this bus electric or diesel? Is it a schoolbus or a city bus? If it's a school bus then why are there people getting off mid stop? you haven't explained these details and they're IMPORTANT to me understanding the question! Wouldn't they cancel the route if so few people were using it? Asnwer me that ... your stated scenario makes no sense at all".

The point of economic thought experiments / maths word questions isn't to constantly question the specifics of the question, it's to abstract the given word problem into a maths equation, then solve the maths equation.

[smjjames]

Wait, was it an attempt at a 'tragedy of the commons' dilemma? That doesn't work because you can't quantify electricity on the grid, measure, yes, quantify, no. Unless said boardinghouse is off the grid and have a finite amount of energy available from a generator or battery.

If you're on the grid, electricity isn't a finite scarce resource, which is a requirement for the tragedy of the commons dilemma. https://en.wikipedia.org/wiki/Tragedy_of_the_commons

[Reelya]

You're doing the thing where you don't get the problem because of ridiculous arguments again.

I explained it in concrete terms with the lightbulb analogy. That clearly works and is identical to the original question.

~~~

"quantify electricity" was already factored in because I always expressed it as "$4 worth of electricty". How is that not quantifiable? Where is this coming from? you could express that as KWh if you like but it's irrelevant.

~~~

And you have that completely wrong here too. This isn't even making sense from your part. I covered it in the original argument by presenting the concept of bitcoin mining at the household expense:

e.g. - running 1 "unit" of bitcoin mining is defined as activity that consumes $4 worth of electricity, and generates $3 worth of bitcoins for sale. Now, each person has a personal incentive to increase their share of bitcoin mining. Cost to you is $1 for each unit of mininig that you do, whereas the benefit to you is $3. So when asking whether you should add another unit of mining, you do the math and always get the answers "yes". However, as everyone takes this action, then the household bill keeps rising, thus eating in everyone's budget which IS A FINITE RESOURCE.

How can you argue that this doesn't qualify as a tragedy of the commons? I'm not following that argument.

[McTraveller]

The point is: if a cool lightbulb is $30 and saves $40 of power over it's lifespan (which is therefore a time-invariant decision) and bills are shared with even one other person, then there's zero economic incentive for either one person to get one and put it in their room. You spend $30 and get only $20 back. So it fails the "rational choice" test because you chose the option that left you worse off.

Wow you guys are all on a different time zone / posting schedule than I...

But the above is exactly the point I was making: current technology does have efficiency improvements, but its "buy in" cost is still too high for the average (let alone "rational") individual to chose the "efficient" option.  That is - the individual benefit is less than the individual cost.

We will see a real tipping point when the individual benefit exceeds the individual cost.

Climate change has great fodder for examples here - most people say "so if a coast halfway around the world floods, I have no way of knowing how much that is actually going to increase my costs, so me spending money or giving up luxuries has basically unknown benefit to me."

Put another way - the costs are localized, but the benefit is diffuse... it's like the reverse of when we think of negative externalities where the benefits are localized but the cost is diffuse.  That's a very tough sell to the average person.


EDIT: removed spurious apostrophe

[Reelya]

I agree with that, the main problem with electric cars currently is that the pay-off point is longer than most people keep the same car, and that if someone is buying a brand new car, then that's more likely to be someone who buys cars more frequently, e.g. people who don't keep the same car for that many years.

People aren't irrational they're just being cautious and discounting pay-off points which are too far in the future based on their needs. Which is in fact an economically prudent attitude: "take the money now not later" in fact has a lower risk profile.