Reenigne blog

National governments are some of the most powerful entities in the world today. But what are their missions (or what should they be)? What are they optimizing for? Democratic governments at least are supposed to represent the interests of the voting public, but this just means that the voters get to choose between one of several parties, each of which have their own mission.

The mission of parties on the right of the political spectrum seems to be maximization of GDP (or GDP per capita). On the left, things are a bit more complicated. The labour party's ideology is roughly social democracy, but what does that actually mean? Bouncing around on Wikipedia for a bit suggests that this ideology promotes social justice, of which the key tenet seems to be reducing inequality. So if we simplify the right down to "GDP maximization" then the corresponding simplification on the left is "equality maximization". Obviously these are gross caricatures of infinitely more nuanced policy sets, but bear with me.

Suppose we are considering two independent policy changes, A and B. Change A makes the poorest person in the country poorer by £1 and makes the richest person in the country richer by £2 (a net GDP increase of £1). Change B makes the poorest person in the country richer by £1 and makes the richest person in the country richer by £2 (a net GDP increase of £3). Both of these changes increase inequality. All else being equal, it seems that conservative policy would be to support both changes (since they both increase GDP) and labour policy would be to reject both (since they both increase inequality). I think that change A is bad and change B is good, and I think a lot of people feel the same way. I would prefer maximizing a metric which is increased by change B but decreased by change A.

What's an example of such a metric? Well, one example would be "how rich is the poorest person in the country?" It seems like quite a dumb metric at first - how can you judge the performance of a entire country on the outcome for a single person? But the more I've thought about it, the more sense it seems to make. The poorest are almost by definition those who need the most effort expended to help them. It doesn't take very much money to stop that person being the poorest, at which point you switch to the new poorest person. Then you need to raise up the poorest two people to the level of the initially third-poorest and so on.

Where do you get the money to help these poor people? Well, you can take it from the rich - the "poorest person" metric doesn't care about them one way or another to a first approximation (and rightly so - the rich don't need help from the government, they can help themselves). Now, the logical extrapolation of that is that we should take all the money and redistribute it evenly - give everybody N/M where N is the total amount of wealth and M is the number of people. That is a way to make a society that is extremely egalitarian but utterly impoverished. Without any inequality at all, there is practically no incentive for people to work hard to create wealth (if you did you'd only get an infinitesimal 1/M of it). Whenever this kind of thing has been tried in the past, that is the result.

So distributing all the money evenly does not maximize the wealth of the poorest - we can improve things for them even more by allowing some inequality and redistributing some of the wealth (but leaving enough to create incentives). So applying this metric does enforce some compromise between the capitalist and socialist sides.

As written above, the anti-poverty metric is very underspecified. Over what timescale should you maximize the wealth of the poorest? If you choose a very short timescale then that implies that you should just redistribute all the wealth evenly very soon (which is very bad in the longer term) and if you choose a very long timescale then that implies that you should aim to maximize GDP and then redistribute all the wealth evenly at some very far off point in the future (which doesn't actually do anything to help people that are hungry now). Perhaps the difference between left and right politics is that they really want the same things, but just disagree on the timescales that should be involved. As for me, I suspect the ideal timescale would be something between that of a typical political term and that of a typical human lifespan, but the exact length may depend on the specific policy under consideration.

There's more that governments do more than just the redistribution of wealth (or lack thereof). How can the other functions that we'd like governments to do be justified under this metric? Well, we can do so by taking proper care about how "wealth" and "poorest" are defined. There's more to wealth than just money. Somebody who is in need of medical treatment is (all else being equal) "poorer" than somebody who isn't, so governments should provide healthcare. Somebody who is a victim of crime (or who is living in fear of becoming one due to crime going unsolved) is also impoverished. Even a certain amount of military expenditure can be justified on the grounds that getting invaded by another country would be impoverishing. The more things we include in our definition of "wealth", the more government intervention is justified by the metric.

The high cost of housing is a big problem for a lot of people. Younger adults in particular often can't afford to own their own homes like their parents did, despite earning similar wages corrected for (non-housing) inflation. Many of those who do own have huge mortgages that will take the majority of their careers to pay off, and face having to move should they suffer a significant drop in income. Many of those who can't own are priced out of the cities and regions they'd prefer to live in by spiraling rents and may face long commutes to work in those places.

What could be done about this? Well, one idea that keeps bouncing around my head is eliminating a situation which is quite rare and very unfortunate when it does occur - the foreclosure. If mortgage lenders lose the ability to take away the homes of borrowers who stop paying their mortgages, they'll stop lending people money to buy homes. This wouldn't be like a medieval anti-usury law - people would still be able to borrow and lend money with interest, it's just that people would no longer be able to secure those loans with their homes (because they need those homes for living in).

Wait a minute, though, doesn't that sound like it would make homes less affordable rather than more? Well, perhaps at first. But why are house prices so high in the first place? They're as high as there are because that's what people are prepared to pay. And they're only able to pay that much for houses because that's what lenders will lend them. So, in the presence of mortgage lenders (and absent any other restrictions) house prices will trend upwards until the cost of servicing a mortgage is equal to all the money that can be earned by the people living in that house minus what it costs them to live (food and utility bills). Rents will rise similarly, as otherwise it would make more sense for those who would buy property to rent to instead buy mortgage notes.

If this were to come to pass, what would the consequences be? Who would be the winners and who would be the losers? Well, anyone who owns their home or a rental property free and clear (or who owes less on their mortgage than the value of the property would decrease by) would suffer sudden drop in overall wealth, as would anyone who has invested in mortgage notes. However, most property owners with an income would suddenly find that a lot more of it would be disposable (no point making any more mortgage payments!) so there would be a massive influx of cash into the economy. Pretty much everyone but the aforementioned owners and investors would be better off. In other words, it would be a massive redistribution of wealth from the rich to the poor.

What about those who do not own their own property but instead rent? Most such people probably don't have large amounts of savings, so wouldn't be able to buy a house outright even at the new lower prices (if they could they would probably have enough for a mortgage down-payment under the current rules). So they will have to continue to rent. Since there's no money in mortgage notes anymore, buying rental properties becomes a relatively more lucrative investment. So those with the money to do so buy up all the (newly cheap) properties for rental. This (combined with the lack of mortgages), prices would-be homeowners out of the market again and forcing them to continue renting. New builds become less lucrative (since they can't be sold for as much) so the supply of housing goes down, which drives up rents to the limits of what the market can bear. So while getting rid of mortgages would help most current homeowners, everyone else would be screwed in the slightly longer term. Our little "tweak" has actually had the opposite of the desired effect! Not only is housing as expensive as ever, even more people are renting instead of owning.

We can fix this with another small adjustment to the rules - instead of just getting rid of foreclosures, get rid of evictions altogether. Everyone who is renting their home becomes the de-facto owner of that property. Sorry to the rental property owners, you're screwed right along with the mortgage note owners.

How would this work in practice? You can't evict someone from their home, but how do you define what "home" is for the purpose of this law? What if someone has multiple houses? I think my preferred implementation would probably be something like this: each person would be permitted to enter into a particular database a single address, which is the place that they are not allowed to be evicted from (their home address). People who don't like being in government databases don't have to be in this one, but would lose the benefit of having a home that they can't be evicted from. At the point of implementation of the law, you must have some claim on an address (i.e. ownership, rental agreement, or be a dependent of someone with an ownership or rental agreement, or have some reasonable evidence of occupancy) in order to make it your home. There would probably have to be quite a bit of wrangling and untangling to determine who should really have the right to live where. However, once the implementation is complete things become much simpler. From that point on, you'd only be permitted to change your home address to a new one with the permission (by power of attorney if necessary) of all living people who also claim that new address as home. This means that an inhabitable property that doesn't have anybody claiming it as their home could be claimed by anyone - a sort of "squatter's right" for this new regime.

Buying and selling property would be fine, as would any other contracts involving a change of home, with the limitation that at every point in the execution of a contract, every person has a home. So a new rental agreement or mortgage agreement would not be possible because there is a point in the execution of those contracts (i.e. non-payment of rent or mortgage) in which someone ends up without a home. Now in this form there is something very similar to a mortgage contract, namely "if you don't pay your mortgage, you have to change your home to this hovel in the middle of nowhere". It's a de facto eviction if not a de jure one. So, I think contracts involving home changes would have to not extend over time - they would have to specify a single point in time at which point all the home changes happen, and a home for every person involved afterwards. If one of those addresses turns out not to exist, then the entire contract would be void and everything would have to be put back the way it was at the start. If a contract did extend over time, the "hovel in the middle of nowhere" might turn out not to exist, and if too much time has passed then restoring the conditions at the start of the contract may also be impossible.

So in this world the concept of a chain becomes more important when buying and selling property. Every non-circular chain must begin with the splitting of a household or the destruction of an address, and end with the joining of households into a single address or the construction of a new address. "Construction" and "destruction" here do not necessarily mean physical housebuilding and demolition - it could be doing up a previously uninhabitable property or a previously habitable property falling into disrepair - both with the consequent changes to the "home" database.

The existence of "uninhabitable" property not listed in the "home" database presents a potential problem - a "shadow economy" of people living in "uninhabitable" properties with mortgages or rental agreements and all the accompanying baggage, with their "home" addresses (if any) pointing at some undesirable location. Perhaps this could be solved by forbidding people from living in addresses not in the "home" database and/or forcing property that is found to actually be inhabitable into the database. Forcing real estate to be in a database seems less onerous than forcing people who don't want to be in a database to be in one, but people not in the database may find it difficult to find housing if non-database housing is off limits. They'd have to live with somebody who is in the database.

Sometimes houses burn down, or the people living in them have a baby that there isn't room for, or a housemate turns out to be abusive, or any one of many other circumstances in which the supply of housing is insufficient for the people needing it. For these circumstances I think a basic minimum standard of social housing would be a necessary fallback when there's nowhere else to go. As a general rule people would not be expected to live in this high-density, no-frills accommodation indefinitely - it's just a stopgap measure until they can accumulate enough cash to buy their own property. It needs to be sufficient so that people living there can hold down a job (good night's sleep, sufficient facilities to remain clean and fed, no worrying about security) but doesn't need to be fancier than that.

A little care would need to be taken in the areas of hotel rooms, holiday lets and other short-term accommodation. Perhaps these things can be handled by saying something like "if someone stays somewhere for more than 183 nights in a given calendar year, that place is deemed inhabitable and can be claimed as home by that person", but we'd still need some way to allow people to go on holiday without making it possible for property owners to exploit the poor by making them rent and move every 4 months. There are other rough edges as well (what about people moving internationally?) but I'm confident these could be solved.

This all seems so sensible to me. It's too bad there's no apparent way to get from the grossly unjust system we have at the moment to this one - there are just too many powerful people who would stand to lose so much money that they would do everything they could to oppose it.

Several very smart people are worried about artificial intelligences becoming smarter than people, taking over the world, and not necessarily having our best interests at heart. A specific concern is that some form of paperclip maximizer might optimize the world for something that is harmful to us if the optimization in question is taken too far.

I once tried to explain this to my grandma, who replied "if a computer is taking over the world, why can't you just unplug it?" A very sensible plan, if the computer in question isn't already running a lot of critical functions, without which people would die and society would devolve into chaos (and if it doesn't have batteries). Obviously we'd need to put a stop to something like that before it got started.

Except we can't - it's already too late. We already have a "paperclip maximizer" running the world. It's not optimizing for paperclips, though - it's optimizing for profit: GDP (on the level of nations), shareholder value (on the level or public companies) and personal income (on the level of individuals, especially those with bills to pay). No artificial intelligence is required - the "machine" uses the intelligence of the people that comprise it to do its thinking. But all of those people answer to someone else - employees answer to their managers, CEOs answer to their customers (usually) or investors, most investors are just trying to get the best returns so that they can retire as quickly and comfortably as possible, and consumers are just looking for the best performance/price ratios. Elected representatives answer to voters in their constituencies or lobbyists for various special-interest groups, mostly industries. Voters all too often vote for who the media tells them to, or who seems more likely to ensure that they can find work or keep their jobs. There's nobody in charge of the whole thing, putting the brakes on to ensure that profits don't come before the freedom and welfare of individual humans or humanity as a whole. There's no "plug to be taken out" short of a massive revolution which would also dismantle the systems that keep us fed, clothed, warm, clean, secure, healthy and entertained.

Taken to it's logical conclusion, what does a world fully optimized for profit look like? Well, predictions become a whole lot more difficult once super-intelligent AIs are in the picture, but here's what I think it would be if such things turned out not to be possible. As there are things that humans can do that machines can't, most human effort would go into profit-generating activities (in other words, we'd all be wage-slaves). If someone is rich enough not to have to work then they probably wouldn't, so the generated profits can't go to making a lot of people rich - this implies that almost all the generated wealth would be concentrated in the hands of a very small number of individuals. Most workers would probably have quite a lot of debt since if they could have an unchecked accumulation of capital they could get rich enough not to have to work. Social safety-nets for any but those absolutely unable to do any kind of profit-generating work would be dismantled since the more dire the consequences for not working the more people will work. Education of children would focus on those skills needed in the workplace, and pure research would only be funded to the extent that it could eventually lead to more profit. Retirement would probably not be a thing (or if it is, you would not expect to have a lot of quality post-retirement time).

It's not all bad, though. Unemployment would be very low (since if someone could do useful work, there's no value in letting them go idle). Amongst the employed, extreme poverty would be nonexistant as people who starving and/or overly stressed about lack of money aren't able to work as effectively. The standard of healthcare would be good (since getting sick makes you unavailable to work and dying means money spent training your replacement) but may focus more on expensive and continual management of disease rather than cures (since this can be paid for by the worker as another incentive to work). War, crime, and political instability would be non-existent as overall these things destroy wealth rather than creating it. As most of the work that can't be done by machines is intellectual in nature, workers are likely to be well-educated, working conditions are likely to be very good and workers would get as much vacation and leisure time as needed to keep them from burning out and to maximize their overall productivity. Time spent commuting is wasted, so people will tend to live close to their workplaces. Violent revolution would be bad for business, so the general standard of living would be good enough that people would not expect to be able to improve it by revolting. Entrepreneurship would be encouraged (and funded by the wealthy) as long it is profitable overall.

On getting to that point, the amount of wealth being generated will surely far exceed that needed to meet the basic needs of workers and keep them productive. The excess goes to the super-wealthy, but what would they do with it? They would be the few who have the luxury of being motivated by concerns other than profit, so depending on their interests they might invest in the arts, space travel, curing diseases and other ills of the world that are left unaddressed by profitability concerns, or blue skies research. Or perhaps they will just build gigantic monuments to their own egos that will be enjoyed by subseqeuent generations of tourists.

That actually sounds less awful than I thought it would be when I started writing this essay. It's still all rather unjust, though - ideally the excess profits would be distributed far more equitably so that more people have a say in what the priorities of the human race should be. I hope we can find a way to keep the good parts of this scenario while replacing the bad parts. In a future post I'll look more into how this might be achieved.