Hillary Clinton proposed Thursday that Americans be automatically registered to vote when they turn 18, unless they opt out, one of a series of voting-law changes she said would expand access to the ballot box.So, if you're a man -- and not a male-leaning female-born whatever -- you have to register for the Selective Service at 18. They'd love to automate this process, since the concept is zero-noncompliance. But you actually have to go down to the Post Office and send them a card, or at least I did at 18. I assume there are other options now, but they all require you to do something. Because they assume you won't do it for free, there are significant legal penalties for failing to do it.
Most Americans like to drive. So states everywhere have passed "Motor-Voter" laws, requiring the DMV or MVA or whatever they have in their state to ask you if you want to register to vote while you're obtaining a license. Clearly, this doesn't get everyone because we're still talking about it.
Public libraries serve lots of the folks we are hoping to reach, at least we hope they do. It'd be great if they were taking advantage of free opportunities to educate themselves. So when you apply for a library card in many states, they automatically ask you if you'd like to register to vote. But there's no guarantee you go to the library, and if you do, you only need a card to check out books -- not to use the free computers.
We've chased this concept a long way already, and as far as I know we don't have a good idea of how to do what she's talking about. Maybe someone will ask her how she intends to accomplish what Selective Service, Motor-Voter and the Public Library plans haven't managed to do. Assuming, that is, that anyone is allowed to ask her a question.
55 comments:
The people she wants to vote will be the last ones in the country with any use for a library......
The goal of course is to get the potential bread and circus voter. This brings up a question I have had on the upper shelf for a long time. Some people bring an economic value to the country, expressed in taxes paid. No matter what the form of government, a tax is a tax and adds value to a person or persons,from the gov. perspective. If you do not pay taxes, and your only value to the gov is as a vote, what happens if the gov decides it no longer needs votes, even as a charade?
Maybe it needs labor. A lot of the folks you're talking about are tax-eaters (although they do pay taxes, at least sales taxes when purchasing goods, and possibly other taxes as well -- but the benefits they draw out vastly outweigh the taxes they put in). But they provide cheap labor to people who end up paying significant taxes.
So if they provide cheap labor that holds down costs, the company can afford to pay a bit more in taxes. If they'll vote the right way too, they're very useful people to have around -- until the automation wave gets far enough along that we don't need labor, either.
After which we'll be in a very different world. Very different, but coming fast.
On the how, I suppose you could simply register everyone who shows up at a voting station. It seems like you'd need them to show some form of ID, though, which as I understand it is to be opposed at all costs where voting is concerned.
Let's see: automatic voter registration, gotta opt out. It's really quite straightforward.
A close enough to 100% for government work per centage of live births in the US are registered in some way within a few days of birth, usually on the day of or the next day. Government just collects that datum along with all the other data it's already collecting from the family and/or the relevant "health" care provider.
Government then sets a marker in its calendar for 18 years later and notes the person as registered to vote. No need, even, to send a reminder of that fact to the individual involved (spending cut! no postage for the task!). It's on the individual to opt out. With suitable numbers of suitably complex forms to fill out, repeatedly, to justify to IRS' and NLRB's satisfaction (bonus perk: union registration) that your opt out is legitimate.
There, too, is your labor solution.
Eric Hines
Yeah, but what percentage of people born in the USA are in the same state 18 years later? Registration is done at the state level.
Indeed, it's really important to know what Congressional district you are in -- and state Senate district, or state Rep district, etc. So we need something more current than 18 years prior, right?
All that follows along with the citizen. Utility bills (the classic proof of location), licenses for this or that, mailing addresses from tax forms, welfare checks, and the like. All automated into the central database.
Eric Hines
Did Clinton say anything about "citizens"?
It's simplest if we assume that everyone voted for Hillary Clinton unless they manage to register and cast a vote otherwise, and even then maybe still credit her with the vote, in case there might have been undue coercion.
As for people who have moved or died since they hit the government radar at birth, I hope you're not suggesting that their franchise should be impaired? We've been tracking their location by drone and/or iPhone all this time. When in doubt, count their vote in all the counties and states they've ever been in. If they're dead, assume you have their proxy for any contest(s) that may be convenient.
That really would save a lot of time. Then we could cut back on polling stations, since only a few people would want to register their objection -- maybe a single central polling station per state.
Okay, so I'm guilty of veering away from the post's topic- sorry.
"...until the automation wave gets far enough along that we don't need labor, either.
After which we'll be in a very different world. Very different, but coming fast."
You'd think that would have been the view at the dawn of the industrial revolution- we may never see an equivalent concentration of labor reduction ever. But the thing is, the labor we "need" changes- what were once not even "wants" but mere fantasies are now "needs"- cell phones for instance. We also have an unending appetite for luxuries and frivolities of all kinds. The potential market never ends, so the real need for production (of an item whether it's actually 'needed' or not) never ends. What you're positing will never happen.
My concern is more along the lines of us needing to find a way to produce enough of what we need after some major calamity, given the trajectory of the world right now, and hints like this.
Agreed. We've been automating since the dawn of time. We sometimes sideline whole groups of people from useful economic activity, but as far as I can tell it's not because of automation.
We used to spend nearly every waking moment producing the bare minimum in material goods required to survive and reproduce. Now we have lots of excess. But until we all become male lions, we're going to want and need things from each other--whether bare necessities or luxuries, whether material goods or personal services--and that means trade and employment.
We'll see. Just because it went that way before doesn't mean it'll got that way again. It seems to me that we went from "You can make this" to "you can make that." But now every other 'that' we could make seems fit for automation. What can't the machines make, now? Increasingly, they can even make and maintain themselves.
Everyone keeps telling me 'services' are the answer. But a lot of services -- travel agent? Expedia? -- have been automated too.
So maybe, as you assert, we'll never run out of things for people to do, that pay well enough that people won't die working full time for what people are willing to pay them to do those things.
But I doubt it.
Bottomless pits, after all, usually prove to have a bottom.
You tend to think of the economy (if I understand you correctly) as a device that ensures delivery of critical needs to every single citizen. I, of course, think of it as the way people signal to each other the relative value of all the things in limited supply (and with alternative uses) that they are prepared to produce and exchange with each other.
It's a mistake, I think, to imagine that we understand today everything that future people will want other people to do for them--which, after all, is all that compensated labor is about, and all that an economy keeps track of. When everyone's cold and starving, they want food and shelter from each other. As soon as they're fed and warm, they want other things. No matter how much stuff robots give us, are people going to live alone? Are they going to drift around each other without mutual wants? If they're happy that way, and robots are standing by to provide them everything they could possibly desire as if in Eden, then I guess there won't be an economy--but then, no one will miss having a job, either. They'll all just compose poetry or something, and recite it to themselves in the shower.
But it doesn't square with what I know of human nature. People always want more things, more tasks done, more entertainment, more art, more pampering. They always find that some are better than other at doing each of these things, and start trading. If they didn't, everyone would live alone in the woods providing himself with all the necessities. Should we really think that in the luxurious future they'll live alone in comfy towers? Or that everyone will do things for each other gratis, out of sheer repleteness and spontaneous generosity? Will no one have to keep the peace or solve problems of any kind? Will the robots build themselves, and find all the resources needed to keep up the construction, and invent the new cool things? Will no one ever run out of room or resources?
When I was a kid, my father tried to explain to me how it was possible to grow up and invent things for a living. I protested that everything already had been invented; the age in which this stuff hadn't been thought of was irretrievably in the past. i couldn't imagine that people would want new stuff. Wasn't everything already getting done?
People may always want more things, but do they want them from people? I very much prefer Expedia to having to call a travel agent and try to express my needs or wants over the phone. I always use the self-checkout lane at the grocery store in preference to having to maintain a fake conversation while checking out. There's been some talk about replacing $15-an-hour fast food cashiers with automated order machines. Taco Bell installed such machines in Atlanta for the 1996 Olympics, nearly 20 years ago. They worked way better than human cashiers. I was very sorry when they took them out.
Now, it could be that I'm given to misandry, and most people long for human contact. But I don't think so: I think I like people for the most part, but also like being left alone, and these are ordinary human feelings.
So what services do I want? Massages? Robots could do that without intrusion. Someone to mow my pastures? Robots could do that too. I can't think of anything for which I really would want a stranger to come to my house if I could buy or lease a cheap machine to do the job as well. Can you?
It seems as if we're approaching a moment in which it will be possible to build machines for almost all ordinary tasks, and many that are now the province of skilled labor. There are robots who are increasingly capable of replacing anesthetists. And good riddance, assuming they do it well. But of course, someone must buy these things if there is to be an economy of making and selling them. The market for roofer robots -- a new roof would be nice, especially if it were cheaper and didn't involve having strangers prowling around my home for days on end -- can't be limited to the very top, even if they all own several homes, or there will be no profit in it.
So what I think we'll have to do is create a kind of artificial demand. The usual way of doing that, since Keynes, is government spending. That seems inefficient to me, as well as an undesirable concentration of power in the government. What we want is some way to empower individuals to control their own means of production. But if they aren't wanted for anything, how to do that?
You tend to think of the economy (if I understand you correctly) as a device that ensures delivery of critical needs to every single citizen.
I don't think that's quite how I'd put it. I think of the economy as a analogically like a food web, in which thing 'eats' that thing, which produces this other thing, and so forth. I do think it has a goal, though, as does the food web. We could talk about a food web as having fallen out of order if it was no longer sustaining its order but was imbalanced to the point of extinction of its members. The weaker organisms may go extinct first, but eventually the imbalance will disrupt the whole system and destroy even the strongest organisms that are a part of it.
That's my analysis of this process of cheapening labor via globalization and immigration, coupled with automation. At some point, the whole system will collapse. It'll destroy the strong as well as the weak if we don't find a way to shore it up.
If everyone is getting what he wants and needs from robots, why would we create artificial demand? What would jobs be for? What's the "system" that's supposed to collapse in this scenario? What's the destructive force that will come into play that both the weak and strong must fear?
I'm assuming that robots can generate themselves endlessly, of course, and that no one will need people to stay involved in that process. If I'm wrong about that, presto, jobs.
Or maybe some people will want stuff from other people, and other people will be content in isolation. Either way, what's the problem?
"I think of the economy as a analogically like a food web, in which thing 'eats' that thing, which produces this other thing, and so forth." Fascinating: it's all about people preying on each other? Why? Why not a symbiotic model? The "food web" model is something I'd apply to the welfare state, perhaps; you can imagine a parasite getting out of control and killing its host. A symbiotic model just grows contentedly, as long as the parts find ways to make each other happier together than they'd have been in isolation.
So no analogy is perfect, and the predatory aspect of the food web (which isn't universal to it -- plant life generates without predation) is one place the analogy isn't perfect. It obviously isn't necessary (or desirable) for me to kill my employer in order to get my living from him.
If you can abstract that idea away, though, you can perhaps see why the analogy works. There's a vast array of entities getting their living from each other. I may get my living from an employer, and then give partial livings to several companies (the grocery store gets a lot of money from me, the hardware store somewhat, the local junkyard -- scavengers! -- a certain amount, etc). Each of them 'gets' from many sources, and then 'gives' to many others.
So what happens when a part of the food web starts being unable to produce its living proportionately to how it is being 'consumed' by the other parts? That part starts to vanish. And that ends up damaging the whole, and eventually all of the parts.
What would jobs be for? What's the "system" that's supposed to collapse in this scenario?
So imagine that I'm right, and there just cease to be manufacturing jobs to a large degree. It's not necessary that every job ceases to exist, but that mostly they do. So people become service workers, but service jobs can also be done more cheaply by computers or automated devices to an increasing degree. So at some point you have vast numbers of people for whom the economy has no work, or work that no one values enough to pay them enough to survive doing it.
People who own shares in robot manufacturing companies don't really need jobs, at first, because the robot-replacement will make a ton of money. Yet all those people out of work are not able to contribute to the economy -- they're like a set of entities in the analogical food web that cease to be sources of 'food.' So the things that got their living from them start to die off too: cheaper stores, gas stations that sold gas to them when they were commuting to jobs, etc. Now those things also bought products, such as groceries and gas, which means that the producers of those things start to starve a bit. Of course, at first they can sustain themselves a little longer by this very process of automation, lowering their labor costs, but that only feeds into the bigger cycle. Every worker they lay off ceases to be part of the 'food web.'
At some point even the strongest and best-placed entities cease to flourish. But you can judge the health of the system as a system well before that.
Or think of technology changes as analogical to the introduction of new competing species. So your buggy-whip manufacturers might be the native red squirrel, and automobile manufacturing workers are the invasive gray squirrel. The one kind of worker 'dies off,' but just because of competition from a replacement 'species.' The system as a whole remains healthy.
What I'm talking about differs from that model because the worker 'dying off' isn't being replaced by a worker of another kind. It's analogous to a set of species dying off without being replaced. The web can adapt somewhat, but it's going to suffer. If this process is repeated in many different sorts of occupations, many kinds of species 'die off' without being replaced. At some point there's a threat to the health of the system itself, so that not just individual species but all the species in the web are suffering or dying.
The idea is that robots will produce everything worth having, but only some people will control the robots and therefore will have exclusive access to the goodies? But then what do you think all the people who don't have access to the goodies will do? Won't they get together and start producing things and trading them with each other? Now suddenly there's an economy and jobs again. They may not have as nice things as the people hogging all the robots, but they're not just going to sit around and starve.
It looks to me like either everyone gets the benefit of the automated production, in which case no one needs a job anymore, or only some people get the benefit of the automated production, in which case the economy continues in effect with everyone else. In order to get into a serious jam, the people who control the robots would also have to lock up all access to all the resources, too, so it's either "robots"or "starve on the desert island that the robot culture chased you onto." In that case, the problem is not economic, it's one of conquest that has to be met with force--or off-planet migration.
Suppose I have a farm, but the only other human nearby is Mountain Man, who lives in glorious self-sufficiency in his cabin, eating squirrels. My beef is that he doesn't need my vegetables and therefore can't be lured into a deal to help me plow. OK, then, I have to do my own plowing and eat my own vegetables; he shoots his own squirrels and eats them. Or, maybe we talk to each other for a while and I convince him he's really missing out on the vegetable front, and we start trading. Is it an economic crisis that he doesn't need anything from me? Only if I'm helpless to take care of myself without his help, and he's genuinely indifferent to anything I have to offer. Should I be aggrieved because automation has put Mountain Man in possession of tools by which he can feed himself on squirrels and therefore turn his nose up at my potatoes? Shouldn't I just be using more imagination about what I might tempt him with? Maybe potatoes aren't the only conceivable trade good. Maybe he'll turn out to be wild for square-dancing. If I really have nothing to offer him, then I'd better be making a case to him for why he should take me on as a charity.
In this hypothetical, it's possible that MM really is happier in isolation in his cabin, and we won't get an economy going. Maybe my potato crop will fail and he will decline to rescue me. Then we can be like two distant countries with very effective trade barriers instead of like neighbors who might build an economy together. But in a real-world scenario, there are lots of different people, and far less chance that none or them will need or want anything from each other. When people interact, they find things they want from each other. The big question usually is whether they'll try to take them by force, or will feel obligated to strike a bargain.
The idea is that robots will produce everything worth having, but only some people will control the robots and therefore will have exclusive access to the goodies?
No, the idea is that robots can produce more, and more cheaply, so that they will eventually replace most jobs. But who, then, can buy the goods they produce however cheaply? There needs to be a large number of people with wealth to consume the goods being produced, but most people have no sources of wealth because there are no longer jobs for them. So even the robot-owners only do well at first. In the end, they can produce their products very cheaply, but have fewer and fewer customers to buy the cheaper and cheaper products.
It ends up being destructive for everyone. It's not a rich versus poor scenario. It's a scenario in which the whole system collapses. And it does so not because it can't produce enough: it could be an economy that is even more robustly productive and efficient than ours. Yet it can't work the way ours works. However much people want or even need goods, they won't be able to produce wealth for themselves. Without access to wealth, you can't convert 'want/need' into 'demand.' Demand in economic terms entails ability to pay.
Wait a minute, are the robots demanding prices for their production? They're not just flooding the market with free wonderful stuff? If not, why is there a problem with people needing to come up with the means to buy all this great free stuff? The cheaper it is, the less effort they'll need to put into producing enough to make up the price.
You don't need something abstract like "wealth," you need the stuff that wealth will buy you. If it all grows on trees, the wealth is irrelevant. There's no need for either jobs or an economy if everything you need is infinitely available, only if stuff you need is scarce and people have better things to do with their time than to make it available to you.
The robots aren't demanding anything, but presumably their owners are. After all, in the economy this new economy is evolving out of, that's how they got the wealth they needed to 'demand' the things they need (and some things they may want as well).
But as fewer and fewer of their fellows have jobs that allow them to express demand, demand for robot-produced products will drop. They'll drop not because there aren't people who want or need those things, but because there aren't consumers with wealth to demand those things.
So we need a way of addressing this demand issue. That's my analysis. Having the government give them money is the usual Keynesian answer. I'd prefer a different answer.
So we're back to the idea that only a few people have access to the robot manna. So why don't all the other people get together and trade goods and skills with each other the old-fashioned way? You can't glut a market with stuff that you withhold from people. The robot manna is irrelevant to their economy if they can't get their hands on it, and scarcity is irrelevant to their problems if they can't get their hands on it. Which is it?
As for demand, all these people who are locked out of the robot manna have plenty of demand. They also have access to resources. Why don't they provide things to each other, the old-fashioned way? I don't see how the problem is one of addressing a demand issue. I do agree that having the government give them money is a stupid answer, though I suppose if the government has infinite powers to tax all the robot manna, it might work for a change.
Or to look at it another way: We don't have a market in air to speak of, because nearly everyone can get it most easily by breathing on his own. The only way to make a market in air is to address specialized cases like scuba diving. Likewise, we used to devote 90+% of the workforce to food production; now it's under 10%. That's automation. Is there a malnutrition problem? Does food use up a greater or lesser fraction of our GDP? Is our high unemployment explained by the lack of good farm jobs? Would we be better off going back to the times when 90% of labor had to be done on farms?
I meant to say "scarcity is irrelevant to their problems if they CAN get their hands on it."
The escape by 'reversion to the old economy' doesn't work for the reason Cassandra keeps raising: the economy is transformed by the availability of super-cheap products. If you can't afford the stuff produced by lower-labor-cost automation, you certainly can't afford to pay a blacksmith to forge the thing by hand.
In the old days, there was enough small-farm agricultural backstop to all this that at worst you could revert to bartering. Now, however, we've got a society of people who wouldn't have the first idea how to raise their own crops, and have no access to land (and certainly can't buy any). So what do they have to barter? Nothing that robots can't make more cheaply, and their client base is poor enough that they'll need to keep chasing the lower prices too.
We're already seeing a 'jobless recovery' with cycles of contraction driven by low consumer demand. What I'm suggesting is the reasons behind that are simply going to continue to increase as trends.
I think the root of your model, Grim, is right here:
"The robots aren't demanding anything,..."
That simply cannot be true. They require raw materials to make in the first place. They require maintenance, lubrication supplies, new parts to replace worn ones, etc. Sure, maybe you make robots for doing those things too, but they also have demands. There has to be a cost to running robots. You assume it will someday be cheaper than human labor for all things, I think that unlikely. Every tool has it's biases. Robots will have theirs as well. I used to talk to my students about why we taught them to hand draw first, and what the biases were of even different CAD/3D programs much less vs. hand drawing. While most production of drawings in architecture is computerized now, there is still a place for hand drawings- mostly as tools for thinking in ways that working through the tool of a computer and software might be less favorable to. I'm sure there are plenty of things like that in the world.
There's also the point at which even if machines can do it all, human labor would get so cheap that it would be cheaper to use them- defects and all. The machines can't always be cheaper and better. It's just not possible. You could argue that it's true but by then humans would be so cheap as to be essentially slave or indentured labor- perhaps, but I think unlikely, and still that cancels out your absolutist end picture.
I think also, consider the experience of dining out and being waited upon. Now, I am much like you and often prefer to use the automated kiosk or checkout, and would probably like a restaurant that used ipads to order, but I'm not sure if I'd like it all the time. A good server brings something to the experience, and for some people the experience of being waited upon by another is worth quite a bit. Spas basically exist under this premise almost completely. People themselves have marketable value. To say that anything can be done better by a robot is not only to devalue other humans, but to devalue humanity- including yourself. I think we're pretty much hardwired to resist that idea at it's core- no matter the objective reality, there is still the subjective.
I think your analogy of the food web is an interesting one as well, and also points to the problem of your argument- in nature, when one thing dies, it does impact other things, but while many may be negatively affected, inevitably, others are positively affected. The elimination of the dominant dinosaurs made way for the dominance of the mammals. Nature abhors a vacuum, and we are part of nature- I think even in some ways our semi-abstract constructs like economies are framed in the natural realities we are within.
I wasn't proposing that we go back to pre-industrial agriculture; I was asking whether the change has been one for the better, since it's an example of automation making something very important much cheaper and easier to produce, while eliminating jobs that used to be necessary when it was more expensive and difficult.
Any time we specialize, we do run the risk that it's hard to de-specialize rapidly or without serious dislocation. We have to weigh that problem against the significant wealth we get from specializing. Automating something that 90+% of us used to have to work at is an example of specializing; now under 10% do it while the rest of us specialize in something else. But of course automation is not the only thing that enables us to specialize; it's just a way of letting us make the change really fast.
It seems to me that the biggest problem with your Tomorrowland filled with robot manna is what happens if something goes wrong with the robots? We'll all have forgotten how to reap and spin our own clothing, and if we don't learn quick, we'll get cold. But that's a continuation of a problem that's been building since we quit making our living by picking fruit and grubs out of the jungle--and very different from the unemployment problem.
"The robots aren't demanding anything,..." That simply cannot be true.
Of course it's true. Robots aren't the kind of entities that demand (at least not yet). Their owners are making any purchasing choices regarding the demand for parts and service. That's not lost from the model, it's just properly assigned to the human owner.
You assume it will someday be cheaper than human labor for all things, I think that unlikely.
A lot hinges on whose intuition is right. But it doesn't have to be everything: you get the food web collapse if you lose even a significant percentage of the species, because they all interact and affect one another.
There's also the point at which even if machines can do it all, human labor would get so cheap that it would be cheaper to use them- defects and all. The machines can't always be cheaper and better. It's just not possible. You could argue that it's true but by then humans would be so cheap as to be essentially slave or indentured labor-
That's wrong, Douglas. Humans have some fixed costs, even as slaves. In fact, slaves were pretty expensive as a form of labor -- Irish free labor was used for the dangerous jobs, because it was disposable as you had nothing but the day's wages invested in them.
But to employ a human for anything, even as a slave or an Irish day laborer, that human has to be fed and clothed. Even allowing for the principles of substitution we talk about in economics, you can't reduce the costs of employing human beings below a certain point. There's no reason machines can't undercut that point in a permanent and pervasive way.
I hire two women to come in and clean every so often. I don't make it my concern how they feed and clothe themselves. That's part of their calculation regarding how much work to take on. If they make it cheap enough, I'll hire them. If they don't, I'll do the work myself instead. Their fixed costs are their own issue, as are all their decisions about how big a house to live in, how expensive a car to drive, how many kids they can afford to support, etc. If working for me doesn't do it for them, they can work for someone else. If no one will hire them to clean houses at a price that works for them, they'll do something else. Labor costs whatever people can bargain for.
Only to a point. If the cost of labor drops below where a full day's work will enable them to get food, shelter, clothing, and other basic needs, they'll die. There's thus a hard limit to how low a human wage can go. The dropping price of goods associated with automation and declining labor costs can push it further, but there's still a bottom to it.
There's no such bottom to automation or computing power, as we observe from Moore's Law. Even assuming substitution, at some point human labor can't cost less. The cost of machines and computing power have no obvious floor.
They don't have to get all of their needs from me in order to avert death. If they need more from me than whatever labor they can bargain wages for, it's not wages, it's charity. Trying to recast it as wages just confuses your economics.
What you're describing is a position where a would-be laborer might be tempted to make a take-it-or-leave-it offer: "If I can't get this much, well, I'll starve, so if you won't pay it I'll just sit here and await my fate instead of working for you." Most laborers probably would decide instead to accept the lower wage and hope for best when it comes to finding some other way to make up the difference--assuming it was literally true that it was impossible to lower costs in order to adjust to the lower wage, which in fact almost never turns out to be the case, which explains why you practically never find emaciated dead bodies in the streets of America. What's more, there are lots of ways the laborer might make up the difference, including another job, help from family, or public charity.
In the case of someone trying to bargain with me to provide a service, if the demanded price is too high, I just won't buy the service, it's pretty much that simple. I'll find a way to do without, or I'll do the work myself, or whatever. After all, if I pay more than I can afford, I'll just starve too, right? It's not only the laborer who has to take reality into account, it's the proposed employer as well. If we can't cut a deal that works for both of us, there is no deal. We both go our separate ways. But that doesn't somehow convert the laborer's wished-for higher wage into a real wage. It's no wage at all unless someone hires him at that price. No power on earth can make an employer materialize who can afford that wage if it's more than the labor is worth to the employer. All you can do by interfering is prevent the employment from happening in the first place.
They don't have to get all of their needs from me in order to avert death. If they need more from me than whatever labor they can bargain wages for, it's not wages, it's charity. Trying to recast it as wages just confuses your economics.
With all due respect, you've been saying this to me for years. I think you're cleanly missing my point. It's not about whether they can get their wages from you. Maybe they have two jobs. Maybe they have six.
Nevertheless, there are 24 hours in a day, and they have to sleep at least some of them. Assume jobs are abundant and easy to get, and that automation has made the prices of consumer goods lower than they are today. Even so, there's going to be some minimum figure that they have to get for their labor if they aren't to starve. It's not a set figure, but there's a floor that they have to get if they aren't to die.
In principle, there is no reason why automation can't compete below that floor. The ordinary laws of economics will require employers to go with the automation rather than be driven to destruction by paying labor costs above the floor against competitors who are getting automated 'labor' below it.
Thus, it is not only possible but likely that increasing technology will put most of humanity out of work. Enough, I suspect, to trigger the systemic collapse I've been talking about. If that happens, there won't be anyone to provide "charity," because their former customers won't be able to afford their products. It's a disaster for everyone, not only the poor or the worker.
Yes, and you've been trying to convince me for years that 2 + 2 may equal 5 if the need is great enough, or if the alternative would be a catastrophe.
If automation makes things unaffordable, why has everything gotten steadily cheaper and more available to workers since automation began? All those jobs lost when we invented the power loom: how is it that the former spinners and weavers don't go naked now? Why can people more easily afford food now that we have gasoline-powered farm equipment operated by 1/10 the workers? The buggy-whip maker doesn't starve, he gets a job in the next town, which he can now easily get to using the new cars that have come onto the market cheap. The abacus whiz doesn't starve; he takes a new job made possible by the cheap computers that are now pouring into the market. If that experience doesn't square with your theory, which of the two is likely to be wrong? Where do you get this rock-solid conviction that wages will inevitably fall faster than the cost of goods, when it's never happened so far, and there's no logical necessity for it to start?
Does having volunteers do essential work for free impoverish the workers who otherwise might have been paid to do it? How is that different from getting robots to do it?
"Only to a point. If the cost of labor drops below where a full day's work will enable them to get food, shelter, clothing, and other basic needs, they'll die"
There are places in the world today where many people " live on less than a dollar a day". That's a pretty low floor to undercut. Few machines can cost less than that factoring in capital cost, maintenance, fuel, etc. Those costs are real and you can choose not to pay those costs, but you won't have the benefits of the robot work production either. That's why when you say "Of course it's true. Robots aren't the kind of entities that demand (at least not yet). Their owners are making any purchasing choices regarding the demand for parts and service. That's not lost from the model, it's just properly assigned to the human owner." you're wrong- your model depends on the function of the robots- and you can assign the 'demand' to the owners all you want, but the machines must have what they need or they do not produce. Period. Thus I'm not convinced by "There's no reason machines can't undercut that point in a permanent and pervasive way."
As for "you get the food web collapse if you lose even a significant percentage of the species, because they all interact and affect one another." I'll repeat that nature abhors a vacuum, and there's always someone who prospers at the demise of others. It's true in nature and true in economies- but because of growth, doesn't mean those who see demise (such as buggy whip makers) will remain in that state, unlike extinct species- humanities versatility is what makes it all work.
Tex,
Charming lawyerly rhetoric aside, I am not arguing that 2 plus 2 equals five, any more than I'm fielding an argument about 'robot manna.' Since you apparently don't understand why I think those questions have already been answered, and why this is not like earlier cases of technology changes, I'm not sure what else there is to say. Clearly you take this to be another iteration of the same problem we've seen in the past, which will turn out the same way as all the others. If you're right, we'll just wait for all these new jobs doing 'something else' that won't or can't be automated to appear.
Douglas,
I don't see any point in chasing the question of whether robots 'demand' the things they need to operate or their owners are the ones who 'demand' those things. In economic terms, it seems more sensible to me to say that the owner does, since the owner is the one with the capital who invests it in maintaining the robots. I don't think that loses the point that robots need, inter alia, lubrication and occasional replacement parts.
It's true that a dollar a day is a low floor, but look at how expensive computers were sixty years ago vice today. The movement has been exponential. The fact that it's a low floor should rather be alarming than comforting: once we get there, we're talking about extreme poverty for the people competing for the (increasingly many) jobs.
Now you can say, and I gather you want to say, that we'll just never get there. Or you can say, as you do say, that something else will come up that we don't know how to automate yet. So what is the counterargument? I've heard the buggy whip arguments since I was 18 years old, and they have held in the past. What's different this time?
Uh oh! If you're bringing up the "lawyerly" thing again, I must have driven you almost to the point of losing your temper; it's not like you to argue by insult. I apologize if "robot manna" was offensive in the same way. I thought it was efficient (if impertinent) shorthand for "almost infinitely cheap goods flooding the market without noticeable effort by human beings via automation."
Thinking about this more last night as I was composing myself for slumber, I wondered if we were in part arguing about labels. We have the value of some work to an employer, and we have the worker's need for income. If there's a gap between the two, some of us call it charity and some of us call it a fair wage. The question that bothers me is: if we conclude that the economy contains insufficient resources to ensure that charity is paid, how will we ensure that it contains enough resources to ensure that the fair wage is paid? Isn't it the same dollars?
And if we lump the excess in with the base wage, aren't we just robbing ourselves of information about whether a particular job is sustainable long-term, or whether we need instead to refocus our effort in another direction? That's what prices are supposed to be for. They're not a ceiling on generosity, they're information about where a certain kind of balance is reached: a point at which the work itself will continue to fund the wage longterm, as opposed to a point where excess resources will have to pour in from outside in order to preserve the job from elimination.
Thinking of it as a fair wage also leaps to the conclusion that the natural source of the charity is the employer. Granted that some of us have a strong belief that that is so, I think it's better to confront the choice openly than to obscure it by choice of terminology. For one thing, the employer's resources may be more limited than our other choices for how to solve the problem.
I think in part the issue is that I think of this as a distinct conversation from the conversation about minimum wages, etc., and I gather you're running that discussion in with this one. What I'm talking about here has nothing to do with fairness, or even with moral philosophy. I'm talking about a practical economic problem.
So abstract away the concern for the worker that usually bothers me -- today I'm not interested in whether workers live or die. I'm merely pointing out that they will die if a certain level of needs are not met. Now, "needs" don't do anything to drive economies: it's what you can "demand" that has economic import. The difference is that demand means you can pay for it. If you need something you can't pay for, that's a problem, but for today it's your personal problem. We don't care. You can die if you can't pay for food.
The practical problem is that poorer workers can demand less; if they end up dying, they demand nothing. Economic growth requires sales, which means that suppliers have to meet demanders in the market. Why is the economy so stagnant? Well, in part it's because we have a very high percentage of Americans who have become much less able to be demanders -- if the 10.8% U-6 number is a good estimate, tens of millions of them.
Now the way this ties in to the automation issue is this: workers' wages can't go below a given level without them dying off. If they get poorer, it hurts the economy because they can't demand products, which means sellers find a restricted market for their goods. If they die, it's the same problem but worse because now they aren't buying anything at all. You're going to get a cycle of slow growth and contraction in the market for as long as this downward trend in wages continues. It seems like it might continue nearly indefinitely, because technology has that exponential growth in efficiency. We won't see buggy-whip jobs replaced with auto-manufacturing jobs, or rather, we won't see auto-manufacturing jobs replaced with robot-manufacturing jobs. Manufacturing is easy to automate. So we shift to service jobs; but service jobs are easy to automate. So we shift to technology and design jobs, but even these employ fewer people as computers improve: how many people with slide rules were working to land Apollo 11? We can do those calculations on every desktop now.
It seems like this should be a nice set of problems to have, because we're gaining the ability to make more goods more cheaply. We should be looking at a rising tide of prosperity. The problem seems to be demand: if we can lick that problem, all these forces suddenly are very productive forces that can be extremely profitable while improving human life. That's what we want to happen.
Charity could be one way of doing that, but it seems to be morally damaging to live on private or public welfare. We need to find a way for people to earn a living that is extra-economic, as it were: because we need them to be demanders, they need a living, but because we need them to be moral we can't just give it to them. And that's true even if we don't care about them at all, but are just thinking dispassionately: an economy requires demand, and it optimally requires a more or less virtuous population that won't steal and cheat.
I've not yet been able to grasp how it is that all those workers starve and die amid all the plenty. They won't even need charity if the stuff is so cheap anyone can grab it: food, housing, and everything else worth working for will be as free and endlessly available as air. In the interim, while we'll getting to this utopia, In order for there to be a problem, wouldn't their wages have to fall faster than prices? As I mentioned, I don't see why that would start happening now when it didn't happen as a result of any of the other automation we've engaged in over the last 10,000 years. You seemed to feel that you'd answered that question repeatedly, but I honestly have no idea what your thinking is on the subject. I'm not trying to be dense or require you to repeat yourself.
There may be a moral problem with a lotus-eating society (no jobs = perfect leisure) if no one has to work in order to supply the necessities of life, but it's neither a charity problem nor a starvation problem as far as I can see. I'm not sure why, but we don't have any common ground on what a job is for, or what supply and demand are for. What's all this stuff about a "demand" crisis? You only need demand if you're worried about keeping suppliers in profitable business, but in the perfect-automation society, we don't need suppliers to profit any more. All the goods apparently are being supplied without any effort from people, so no one needs to be motivated and no suppliers have to take in enough profit to stay in business.
If I'm misunderstanding your hypothetical future world and there still is a need for people to be involved and putting out effort, then that's what the jobs will be: people doing whatever is necessary to make the almost-hands-free automation continue properly. Just like it's always been, since we invented the wheel.
Bear in mind I'm not saying people will need to supply each other with services in order to avoid starvation; I'm just saying that's what people do, and they'll probably keep doing it. But if they all decide that's a waste of time, don't they still have food, housing, medical care, and everything else they need from the robots? What are they missing?
I'm not trying to talk about minimum wage specifically, I'm trying to address your worry that people will need more stuff than they'll be able to generate the price for by working. It's true that it's the same worry that tends to lead people to support a minimum wage, but in the world you're supposing, we seem to have gone all the way past wages. We don't use the concept of wages to describe how people have air to breathe; everyone just has all he needs without having to invest any special resources to make it available.
If the automated goods aren't freely available to everyone, then I still need to understand what's causing them to be withheld and why everyone who is denied access won't simply go back to work providing themselves with goods in a non-automated way. If the answer is that won't work because the goods are available so cheap, then I'm back to wondering why there's a problem with starvation. I really can't figure out the scenario where one or the other process isn't in play. Sure, there could be a problem on the way when food is automatic but medical care is not, so no one can make a living selling food in order to generate funds for his heart transplant. But that just means that jobs will move out of food production and into whatever is still scarce. When the last necessity of life stops being scarce, the need for jobs disappears along with it.
It's difficult, I suppose: the standard examples are so well known, and have worked so well in the past.
Maybe we should return to a concrete example. Why don't we buy American-made pants anymore? They're still available, but they now cost $40-50 a pair. They used to cost around $12-15 a pair. Mexican made jeans cost around $8-10 a pair. There's a delta there, but it was manageable at first. But as the process led to more factories closing in the USA, to pursue cheaper labor costs, the price of the traditional goods rose due to a collapse of supply and the higher labor costs. Now, the goods are out of range for a couple working on jobs of the new kind: half the wages of jobs a generation ago, according to the figures quoted in the newer economics post from today.
Why don't you go back to buying the $15 jeans? You can't make them for $15 anymore. Why don't you buy the $40 jeans? You can't afford them. In fact, you may not be able to afford the $8 jeans, which after inflation for 20 years now are $15. The jeans are cheaper, in a way, than the $15 jeans would be after 20 years of inflation. But the economy has geared downward, so that now you can't really even afford the cheaper products that were a bargain a generation ago.
That's what I think is going on. 'The Great Reset,' economists are calling it, but I think it goes further. I think the problem is that we've severed the link that allowed ordinary people to be demanders in the market, and that's why we're watching an atrophy over time even as products become cheaper and more available.
Have American-made jeans become more expensive because of inflation, or as some kind of result of automation? Unless it's automation, I'm not sure how they relate to this point. Or if the jeans went up in price because American labor got more expensive, I really am not seeing how this explains the automation problem for workers. And anyway, why would an American be forced to buy the more expensive American product if equally suitable clothes made in Mexico were available? That's the whole point of the advantage of trade with countries who can make things cheaper, even including the cost of shipping them a long way to us. If we get to the point where Americans can't make jeans at a competitive price with Mexicans, then people who used to make American jeans will gradually move out of that business and into some other business where Americans have a competitive advantage. Workers in all countries do that all the time. Countries that rest on their laurels slide into trouble.
So, automation is just one form of depressing labor prices. Since we're on the front-edge of that wave, we have to look back to the last wave to see how things play out. What happened is that, at the time the depressing mechanism became available, people had been paying $15 for jeans. They could now pay $12. This minor price point difference meant that the whole industry shifted to pursuing that $12 pair of jeans -- maybe getting it lower, even to $8. Major savings, at first! Americans were out of work, but they just have to adjust (as you say).
But now the $8-12 jeans cost $15, but American workers are being paid half what their predecessors were being paid. Even $15 jeans -- the ordinary price a generation before -- is out of range. So they don't buy new jeans, they shop at thrift stores. Or they chase the next wave: Malaysian jeans instead of Mexican ones.
Is that good for jean producers, our analog for the robot-owners of the next wave? No: their customers are too poor to buy even the cheaper product. Is it good for the American worker? No: jobs vanished, and the new generation of jobs pay half-wages from the last generation. Is it good for the Mexican worker? No: the jobs vanish to Malaysia. Is it good for the Malaysian worker? No: robots are the next step.
It's a system in which we are chasing efficiency in a way that benefits no one, ultimately. But if we could fix the demand problem, then the efficiency gains would increase owner profits while also allowing all these workers to 'demand' (rather than merely 'need' or 'want') the products we can more efficiently make.
There's an is/are problem there: I suppose a pair of jeans "is", but jeans "are." Bother.
Automating jeans manufacture is not supposed to be good for jeans labor, but that doesn't mean it works out badly for labor. Labor isn't trapped in the jeans business. When jeans get cheaper, labor has access to cheaper jeans, and it changes it job from producing jeans to producing something in which it has a competitive advantage. That's the point of specialization: you do the thing you're best at and let the other guy do the thing he's best at, then you trade. Now you both have more than you would have had if you'd both continued to make both things in a second-best fashion.
Our story is getting a little mixed up, because we're now including the effects of both U.S.-based automation and outsourcing manufacture overseas. They're not the same thing. But they're both examples of how goods can get a lot cheaper--via automation or some other kind of specialization or both--which means more good stuff is available to workers cheap.
I keep coming back to the indisputable fact that food and clothing consume a smaller portion of the American families budget than they did 100 years ago. How do you think that happened?
Labor isn't trapped in the jeans business.
Right, so this is the buggy whip argument. It's a totally standard, well understood argument from classical economics about how technology is good in the long run. This is where we are hanging up. So what's different with contemporary automation and computerization? I mean, I'm not asking you to agree. I just want to know if you can tell me what I think is different? Have I been able to convey that much, or not? I think the buggy whip analogy is broken, even though it worked in the past. There's something different now. Can you say back to me what I think is different, with the understanding that I'm not asking you to endorse my thinking thereby?
This is a funny structure of conversation. I ask, "What's different about automation so that it used to work out great but in the future it's not going to?" And you say, "Asked and answered." I say, "No, I really don't know what you think about this." You say, "Can you tell what I think about this?" Well, no. The answer is, I can't tell you what you think about this. I'm asking what you think because I really don't know what you think. I haven't been able to put it together from what you've said. The picture of disaster is very vivid, but how we get from here to there is still completely unclear to me. I'm truly sorry if you think you've explained it and I'm not paying attention.
I've had many similar conversations with a certain professor of philosophy I know back when he was trying to teach me Aristotle. "So what is nature?" "So, just what is motion?" "It's right there on the page, can you just tell me what the words mean?"
So I'd tell him what I thought I'd read, and he'd say, "So, you just said back to me what's on the page. But what does it mean?" It was very frustrating, but it was an alien way of thinking until I mapped it down.
At the moment, though, I think it might be best to take a rest from the conversation. I'm sure it will come up again, and perhaps a fresh start with a new set of examples will at least help me express it in a way that makes sense to you.
How very odd. I've asked a series of questions, and you feel you've answered them in some way? If that's the case then, yes, it's best we give up. My impression was that you studiously ignored my questions, perhaps because they seemed irrelevant to how you look at the issue--though they're clearly relevant to what's puzzling me about your approach.
Sometimes when someone predicts disaster and you can't understand how he gets there--AGW comes to mind--all you can do is wait and see what happens, and then figure out whose prediction was closer to reality. Cause and effect and notoriously difficult to unravel in complex systems.
Let me start with I may be as unclear as Tex on where you are going with this, but I find it interesting and am willing to pursue it further, if you are also interested in doing so.
”I don't see any point in chasing the question of whether robots 'demand' the things they need to operate or their owners are the ones who 'demand' those things. In economic terms, it seems more sensible to me to say that the owner does, since the owner is the one with the capital who invests it in maintaining the robots. I don't think that loses the point that robots need, inter alia, lubrication and occasional replacement parts. ”
You injected that bit, not I. My point was that machines have demands also (call it what you like- needs, requirements, ) and that creates a place for people in the system again. Problem solved by noticing that there is no problem to solve.
I’m reminded of when computers came out and they were all talking about how someday soon, we’d all be living in a paperless world. Today we use more paper than ever, because printing has become so easy. Often, the problems that come around aren’t the ones we thought we saw coming.
”Now you can say, and I gather you want to say, that we'll just never get there. Or you can say, as you do say, that something else will come up that we don't know how to automate yet. So what is the counterargument? I've heard the buggy whip arguments since I was 18 years old, and they have held in the past. What's different this time?”
I’ll argue that in the essential ways, nothing is different. Clearly you believe that it is different. I’m not sure how exactly. In almost every example you’ve used to buttress your sense of this, I see a lack of convincing evidence, which reinforces my current understandings. The only thing I can see is that the buggy whip maker is out of a job, and the instrument that put him out of work- the automobile- creates manufacturing jobs for others. You posit, I think, that automation eliminates people from the supply side of the equation, but I’ve argued that’s not so (and frankly, you’ve dodged the counter-argument), and so we’re still where we’ve started.
(cont'd)
”So abstract away the concern for the worker that usually bothers me -- today I'm not interested in whether workers live or die. I'm merely pointing out that they will die if a certain level of needs are not met.
That was why I brought up the ‘dollar a day’ issue. It would seem to be an impediment to your argument. You’re arguing an extreme position, and that’s pretty difficult to do. I'm pointing out that even if you're partially right, I don't think we're pre-saging the self-elimination of humanity.
”Why is the economy so stagnant? Well, in part it's because we have a very high percentage of Americans who have become much less able to be demanders -- if the 10.8% U-6 number is a good estimate, tens of millions of them.”
This has far more to do with human interference with the market (a la increased regulation and Obama’s command and control policies). I think discussing the current downtrend as though it supports your argument when there are a good many other reasons we can quite reasonably attribute to causing or exacerbating those problems, is a mistake. You’re describing a cyclic problem, and using an anecdote that has specific, reversible causes to be an indicator of it’s existence? Color me unpersuaded.
”So we shift to technology and design jobs, but even these employ fewer people as computers improve: how many people with slide rules were working to land Apollo 11? We can do those calculations on every desktop now.
So why no significant diminshment in the number of controllers in Mission Control between Apollo and Space Shuttle missions?
As for design- how do you train a computer to come up with what people want next- that they don’t even know they want yet? We ask people to do this all the time.
Also, the demand side of the market equation is like the question of ‘how much space do you devote to closets in a house design?’ the answer- everything that isn’t needed for something else, because no one ever had too much closet space in their house. Whatever we’re allowed, we fill and seek more- it’s our nature.
Douglas,
I am quite tired of the conversation, but I will attempt one more time since it would please you. The signal difference between the buggy whip example and the current example is just that the buggy whip example assumes -- a valid assumption at the time -- that changing technology will create new jobs for human beings. That assumption is increasingly invalid.
You say that a dollar a day is an extreme position. But as technology improves, I don't know that it is. Here's how things stand today:
“You’re getting a skilled laborer in a robot once it’s programmed and set in a job. So let’s take that average-sized robot operating at 75 cents an hour. If you project that over the life of the project, which might be 8, 10 or 15 years, much of the cost savings results from not needing a higher-paid, manual laborer. And many times that’s $15 to $20 an hour versus 75 cents an hour.”
So the amount it costs to run the machine now is $0.75 * 8 hours in a man-day = $6 per man day of labor. But of course the machine can work 24 hours a day, excepting refit and repair.
We don't, then, have to see robots get all that much cheaper to compete with the dollar-a-day guy. Plus, of course, to get to a dollar a day you have to set up a factory in a terribly poor and distant part of the world, and ship the stuff: that's not necessary here. What is necessary is an initial investment -- money up front that you don't need to hire a worker -- but the investment is repaid many times over.
There are some jobs that are harder to automate than others, but the window on that is closing. Just today, it is reported that an AI built by Tufts scientists solved a tremendously difficult problem in regeneration that has stumped the entire field for a hundred years. The AI worked out how it could be done: in other words, it designed a process that would achieve the desired effect, and far better than any human being, at the very highest levels of education and training. We simply couldn't do it.
It did it in 42 hours.
You want to tell me that these things can't design a better cabinet? It's true that human beings are going to have to decide what we want the machines to do, but it isn't true that they won't be able to do the things far better than a human designer can.
So I think the buggy whip argument simply fails. It is not true, anymore, that changing technology implies new industries with jobs for human beings. Manufacturing changes brought about by technology will increasingly be automated. Mental work is increasingly automated. Even design work, past the point of humans deciding what they might like, is increasingly going to be automated.
That won't put every human being out of work any time soon. It will, I think, put a huge percentage out of work. Just look at the expected impact of driverless trucks.
That's what is different, and it's why I think the old answer about the buggy whip just doesn't apply to the world we face anymore. And as I've also said several times, this should be a solvable problem. It could be a very good world to come out of this automation and technological improvement. In fact, I really think it's going to be, if we can solve this problem of demand. We need all of these people who are going to be out of work to be able to come to the market and demand the things they need (and, ideally, at least some of the things they want). If we can lick that one problem, ideally without the government just giving everyone money (see the parallel discussion on the moral hazards of charity in the Catholic Business Ethics post), it will be fine.
If we can't lick it without the government just giving everyone money, it will be less fine: we will end up with fellow citizens who are dependent on government in a way that is morally hazardous and dangerous in a Republic. If we can't lick it at all, even if only because it proves to be politically impossible to lick, we're looking at a loss of a large enough part of the 'food web' that I think there's a danger of systemic collapse.
Thanks for humoring me but really, we can let this go- time is precious and I don't think either of us will persuade the other anyway.
I will only point out that the supposed AI puzzle solver is really doing advanced, high speed analysus, not- I repeat not design. It ran a bunch of operations with a high number of variables until it found one that worked for a specific pre-assigned problem- computation, not design.
I also think that some things are too complex to figure out in a way that is of use to us- especially as history has proven time and again that we aren't very good at looking into the future. Probably because no one person, not even one group of people, get to make the decision as to where we end up going.
Post a Comment