Things I Wish I Said

Making group decisions

noreply@blogger.com (Greg) — Thu, 18 Feb 2010 07:55:00 +0000

There's a huge amount of advice available for decision-making, well-being, spiritual development, and the like at the personal or individual level. It's far from a solved problem. More attention needs to be paid to the problem of figuring out what larger groups of people ought to do, though. Just a for-instance, the Bible has quite a bit of material on how to behave personally, but the best known group governance advice is treated as uniform individual admonishments ("Noone should kill other people.")

This seems to be a common feature of thinking aimed at the governance of groups -- "if everyone would just follow these uniform mores, everything would work out well!" Well perhaps so, but if you haven't noticed, that's not happening. Exhortation would appear to be fruitless. Making group decisions and deciding on group governance is not a process well modeled by gas laws (treating all the individuals as interchangeable and with simple inter-personal mechanics), even, perhaps surprisingly, for quite large groups of people.

I'm not sure what the technical terms are, but I even wonder how applicable lessons from smaller groups are to larger ones. For instance, countries are heterogeneous groups of people with largely involuntary memberships. (You can't just declare that you're not bound by a country's decision-making process, and conversely, the country can almost never just kick you out.) The sanctions available to countries, then, are radically different from those available to smaller groups which are more homogeneous (at least in some dimension) and have voluntary memberships, and from small groups like families which are genetically related. Do many lessons from families apply to corporations, clubs, or churches? Do the principles of organization of these kinds of communities apply well to countries? At first glance, it sure doesn't seem so. Even if there are principles of human psychology which apply to all these regimes, it seems more than just plausible that they manifest quite differently in different types of organizations.

Informally, quite a bit of the thinking on this topic that I'm aware of without digging in very deeply seems to be aimed at figuring out how to make country-level governance operate as an extension of smaller homogenous-and-partial group dynamics. That is, the idea is that there should be some kind of proper subset of the country, which is easier to think about in terms of the dynamics of voluntary and partial group processes, and which should be in charge of making decisions that bind the whole country. The thinking is mostly about establishing the composition and selection and size constraints on this group. That is, the theory is that if you can figure out how to specify "the perfect ruling elite" with just the right rules of behavior on it, then it will be appropriate to vest governance of the whole in that body.

Whether you are a royalist who thinks direct personal ownership of the country by a bloodline of monarchs is the optimal elite, or a republican who thinks that a set of rules specifying the operation of some sort of legislative body is correct, I wonder why we seem to be so drawn to this way of thinking. What evidence is there that this is the best (or even a good) way to do governance at the level of a country?

Having a Title and Description

noreply@blogger.com (Greg) — Thu, 18 Feb 2010 06:20:00 +0000

I was listening to the Stack Overflow podcast today. I was struck by something that seems like an obvious procedural thing that is so often missing from software projects: having a title and description. This is often called the 'elevator pitch.' Or, as the podcast calls it, a 'vision statement.'

The idea is to answer the question, "So, what are you doing?" or "What is your project about?" Here's a stab at a necessary and sufficient list of constraints on the answer to that question.

1. It has to be short. That is, think about it as the title and paragraph description on SourceForge, or the title and a snippet in a search result or in an RSS reader. The canonical 140 characters is a good upper limit.

2. It has to be written. Written like *written*. Like writers do writing. Not like "I put a bunch of ascii characters in a row."

3. It has to be concise and descriptive. More nouns and verbs than adjectives.

How come? First of all, it should be short because it should be easy to explain. It should be easy to explain because it should fit into the world in a way that is easy to explain. If it isn't, it is probably a waste of time.

It should be written in order to solve two problems. One is communicating it to people who aren't in the elevator with you. This also puts a proper scope threshold on the project. A project which isn't worth titling and describing is not a software project. It is something you should do with awk or grep.

The description should be mostly nouns and not many adjectives. This makes sure that you know what it is you are trying to do. The writing of this description commits you to some thinking. Fine. You may be wrong, but fortunately this needn't be your last project. You can change the description and try again.

These constraints imply a certain level of thinking and iterating and refining about what it is you are doing. That's fine. That's basically writing like writers do.

Being Smart

noreply@blogger.com (Greg) — Wed, 17 Feb 2010 05:32:00 +0000

Here are some things I've learned about being smart. All else being equal, it's good, but there are things that are more important. Whether or not being smart is genetic, these skills can largely be cultivated.

1. Listening. There are thousands of really smart people, and the weird thing is that a surprisingly large number of the best, in whatever area, are constantly blabbing about whatever great thing they thought up. All you have to do is figure out who they are and pay attention. That's substantially easier than thinking up good ideas yourself. It requires skill at evaluation, because such people will throw off a dozen horrible ideas for every good one. That's expected. But it is easier to cultivate a careful ear than to arrange to be born a genius.

2. Persistence. 90% (or whatever) of success is showing up, and there are lots of forces aligned to make persistence pay off. Working hard and sticking with something is key. An idea has low value compared to sustained execution. This has to be balanced against adaptability and knowing when to make course corrections, of course, but that's really just another way to look at persistence: you persist despite the need for course corrections, not fighting them.

3. Planning. You can cultivate the ability to be tremendously productive by being attentive to what the most important thing is to do. It isn't always easy to plan that ahead, but it often is. Getting things done by setting intermediate goals, doing simple workable things first. Practicing.

What's the news?

noreply@blogger.com (Greg) — Sat, 07 Nov 2009 07:18:00 +0000

I'm interested in following the rapid evolution of the news business. Something that puzzles me is the conviction on the part of many people who are experts in this (i.e. Jay Rosen), that Twitter is somehow emblematic of the new news system. That's true to an extent, but only insofar as what the news system has allowed itself to become.

As the news system has become more driven by individual personalities, and replaced the apex of the profession with the blow-dried TV anchor, coverage has gotten more and more shallow, driven by the logic of the news organization itself instead of by facts and reality. In short, news is now optimized for sharing unattributed gossip about celebrities, and basically treats every story in that way.

In that sense, Twitter is great for news -- the perfect online match for the attribution- and research-free milieu of today's journalism. But that match is dancing with the devil. Any modern news system that is going to be successful is going to look very little like Twitter, and rely on such knee-jerk low-information tools in only the most rudimentary way. Recent examples as I write this: Balloon Boy and the Fort Hood shootings. In these and similar stories, the news might as well be reading twitter streams -- they put on ignorant "experts" and relay breathless and almost invariably wrong reports from on-the-scene reporters.

But what is going to be successful in a world where passing on vapid chatter is zero cost is adding the values of reputation and authority. That is only going to be earned the hard way -- by taking the time to not get it wrong. To get it right. To explain. To find the facts behind the facade. To investigate. To ask the follow-ups.

What does Twitter have to do with that? If the existing news business abdicates its role of spending the time to get the story to doing low-value "curation" (that is, largely link-blogging), then someone else (probably what we now call "bloggers") will step into the role. The real lesson of Twitter is that curation is easy, and can be done largely through algorithmic aggregation. Unless you are occupied with doing the thinking, investigating, and explaining, you're simply not doing news. That today's news infrequently thinks, rarely investigates and seldom explains, is not evidence that Twitter will save it, but that Twitter will destroy it.

Why people can't think morally

noreply@blogger.com (Greg) — Tue, 15 Sep 2009 04:12:00 +0000

There's an old saw about a fox and a hare sitting down to decide what's for dinner. This mental image, though, is very similar to arguments humans have about ethics. That is, when deciding what claims people have on each other, we traditionally start with actual people. This premise is absolutely emphatic in western law, where the issue of "standing" (that is, actually having been wronged) is critical in any legal action.

This starting place leaves a lot to be desired, however, when engaged in moral philosophy. The most glaring flaw is that individuals manifestly don't start life on equal footing. Their circumstances of their birth vary, as do their environments and genetic histories.

Which seems more morally strange, that two babies, through no action of their own, might be born one to wealth and power, and the other to poverty and struggle, or that at birth, one baby has any sort of ethical claim to participate in the benefits that may be enjoyed by another? Perhaps these two intuitions are the basis for left wing and right wing sympathies, but perhaps the difficulty in both of them is that they lead us astray.

That is, the situatedness of these claims is why we can't get traction in resolving the intuition gap. Imagine, for instance, that these future babies were sitting in Babyland considering how they might organize the world. Would they both agree to structure the world such that one of them (chosen at random) would be well off and the other suffer? Or, if they made the rules, would they change this circumstance? If so, how much? If we take this frame of reference as superior, it gives us a way to reason about how we want to organize the world which would be most satisfactory to those currently in Babyland.

See Peter Singer and Aaron Swartz for more on this.

Legitimacy

noreply@blogger.com (Greg) — Thu, 20 Aug 2009 06:16:00 +0000

There's an interesting new poll breaking down what people who suspect Obama isn't a natural-born citizen believe.

It is really pretty crazy that 24% of Americans believe this, but that's about the number who still thought George W. Bush was doing a good job in 2008. Perhaps these two beliefs are correlated.

But two other results struck me as even more interesting than that. First of all, the basic theory by people who talk about such things is that Obama was born in Kenya. That's what the faked documents are made to show, and where the real action is in the "discussion" (such as it is). Yet of respondents, 10% think Obama was born in Indonesia, and only 7% in Kenya. (More about the remaining 7% presently.) What does this mean?

What it must mean is that most people formed some kind of early belief about this, based on email forwards or something, and have interpreted all subsequent talk as confirmatory of their belief, when in fact they aren't even hardly talking about the same theory any more. These countries are basically on opposite sides of the world, the proposed mechanisms for Obama having been born in them are different (as they'd have to be).

And the other 7%? Well, 6% of respondents say that they don't think Hawaii is part of the United States. Others probably know that it is, but believe that since his parents weren't born in the United States, that makes him not a "natural born citizen." These geographical and legal misunderstandings mark the bearers as not just not paying attention to an arcane conspiracy theory, but in fundamental default on basic facts of the political reality around them.

So what's going on?

One possibility is that people who tell pollsters they don't think Obama is a natural born citizen don't actually believe it. That is, this belief is more of a statement that Obama oughtn't to be president rather than indicative of what they think. But why? Why express this opposition in such a silly and baseless way? It is manifestly clear that this theory has no chance of ever bearing any political fruit, and in fact may end up being harmful to the cause. But as is clear by the fact that virtually all of the people who believe in this are unmoored, not only from basic geographical facts, but from the conspiracy discussion itself, they are probably not well equipped or care particularly much what the political realities are. They may be in the mindset that opposition to Obama implies that they should believe he isn't even qualified to be President. That is, it is just a facet of the unacceptability of his presidency that infuriates them as a whole. Evidence suggests this may make up quite a substantial fraction of the people holding this view.

A fruitful exercise is to look at the circumstances surrounding George W. Bush's election in 2000. Most people opposed to his presidency believed that the Florida election was not fair. The fact that it hinged on such technicalities, and the realities of the error-prone ballots, made it appear that the outcome of the election was within the error bounds of the voting process itself, and that consequently, control over the process of vote-counting was determinative. The fact that Bush's brother and his campaign officials were in a position to exert such control made that belief irresistible.

I'm not saying these beliefs are based on the same kinds of footing; what I'm saying is that we don't have to look back very far before finding an instance where opposition to a president was tightly coupled with various theories of his illegitimacy. Polls in 2001 might have shown that a lot of people believed various versions of how Bush won Florida, all of which would be expected to occur in people opposed to him, and which formed part of the overall complaint about him.

What's up with agency?

noreply@blogger.com (Greg) — Sun, 16 Aug 2009 06:40:00 +0000

The agency problem seems to be getting worse. Or perhaps it's as bad as it's ever been, but we're in a payback period now where it's more noticeable how bad it actually is. Whatever the case, we should take advantage of the situation to try to think of ways to improve it.

In business, it seems like stock ownership isn't proceeding under the agency theory. That is, managers of firms don't appear to act as agents of investors, at least, not to the extent that investors wish or believe they ought to be. Instead, shareholders are more accurate characterized as investing in the talents of the managers. Or perhaps better yet, as making side bets on the bets the managers are making. Is this because of incestuous board relationships and other conflicts of interest? Or will the downturn cause investors to stop wanting to make these kinds of side bets?

The prevailing wisdom on the problem is that you attack it by means of incentives. That is, you assume that you are hiring a crook, and that your responsibility is to make sure that the crook is cheating for you, instead of on you. I am suspicious that whole approach may be wrong.

In a similar fashion to the microeconomics result that charging parents for picking up their kids late at daycare leads to even more tardiness [Freakonomics], I suspect that this implicit posture that managers are crooks and need to have their interests mechanically aligned with owners ends up producing a system in which managers feel entitled to exploit their information asymmetries to get around these rules. (After all, if "the incentives aren't properly aligned," they can't really be blamed for exploiting them, right?)

Perhaps the answer is the same: a return to moral argument. But a diffuse set of shareholders of a public company has a lot of obstacles in making moral demands on a company's management structure. Presumably the board is the vehicle by which this happens, but there's a lot of risk there of over-coziness, an agency problem on stilts (the board is itself the shareholders' agent), and classical information asymmetry.

So how would shareowners exert moral pressure in a more direct, socially powerful way? Putting it another way, how should they make better decisions about hiring morally competent managers, rather than ignoring that and hiring ones they think are managerially competent (but perhaps morally questionable)?

In government contexts, how do voters exert direct moral pressure on lawmakers and executives to provide the leadership they are elected to provide, rather than using the information asymmetry and systemic lags to engage in featherbedding and other kinds of corruption?

Both areas suffer from common problems: historical methods of direct social pressure available to the majority of the stakeholders simply aren't effective. Voters by and large don't go to the same parties as politicians, so if they stop inviting the politician to parties because he or she is corrupt, the politician can just go to the parties of the people whose influence they are transmitting. Same goes for CEOs and shareholders.

Url shortener FAIL

noreply@blogger.com (Greg) — Mon, 10 Aug 2009 05:46:00 +0000

I think I've gotten more shrill about shortened urls. Especially after reading that tr.im is closing.

These things are a toxic blight, nasty to use, worse to see, and are a nasty (un)planned obsolescence introduced ill-advisedly into the core of the internet.

Furthermore, can we all agree that bit.ly has powned twitter? How much should bit.ly charge twitter not to nuke all their links at this point? One billion dollars? Sounds cheap, especially since pretty much the only value in the place is those links.

140 chars seemed hip when sms seemed like a valid practical hurdle. It no longer is, and it is easy to imagine a better protocol. Shortened urls ought to be unlamented collateral damage.

The monopoly on financial force

noreply@blogger.com (Greg) — Tue, 28 Jul 2009 04:54:00 +0000

What's the balance of economic power between a country's government and its banks? Take the Unites States, for instance. The Federal Reserve, in response to the 2008 financial crisis, has increased its balance sheet (the quantity of assets it holds) from a bit under $1 trillion[good graph] to over $2 trillion. That's a lot of money. What's the balance sheet of Bank of America, the country's largest bank? $1.8 trillion. How about Goldman Sachs? $884 billion.

This kind of scale severely hampers the ability of the Federal Reserve to make moves it thinks are important. Or, to use a more common phrase, these institutions are "too big to fail." It is commonly argued (not least by President Obama) that Sweden's approach to its banking crisis was much easier because the banks were much smaller. The current (crisis-inflated) balance sheet at Sweden's central bank (Riksbank) is some 200 billion kroner. The Nordbanken (Sweden's BofA), was taken over in the Swedish banking crisis at $8.67 billion dollars, or about 70 billion kroner.

The balance sheet isn't the only piece of the equation -- presumably the Fed can throw its weight around significantly more effectively than can a bank, so it has more leverage in terms of what it can accomplish than does a bank. Still, though, it is clear that unless this leverage is much higher than seems likely (like 100:1 or something), the scale of the financial industry is much larger than that of the central bank, with large individual financial instutitions being as big or bigger. Still, it is worth noting that in Sweden, having the banks being a bit smaller than the central bank seems to have made the degrees of freedom quite a bit more accomodating. The ability of the financial sector to staff the offices of its regulators and effectively lobby Congress is another manifestation of its large scale relative to the government institutions meant to regulate it.

Is this really a good thing?

The premise of modern governments (the 'Westphalian state') has been the monopoly of force. That is, the state is responsible for (and is in control of) all use of force within its borders. Indeed, the definition of a 'failed state' is the inability of the central government to control the existence of armed militia groups operating within its borders.

This isn't to say that large banks are just like armed revolutionaries or something like that, but a pretty good rule of thumb that a country shouldn't have financial operators at or above the scale of the central bank. This keeps the play at a level where quick, prudent action is possible when necessary, and enables the government to move all-at-once in ways that these gigantic balance sheets of financial sector actors makes impossible.

Free markets and health care

noreply@blogger.com (Greg) — Tue, 30 Jun 2009 05:46:00 +0000

What is the role of free markets in figuring out the level of resources we should dedicate to healthcare? There are several problems with how the market works in our current system, including:

The healthcare marketplace is full of distortions: monopolies granted to drug and equipment companies in the form of patent protections,
Indirect consumption of virtually all healthcare products, leading to an enormous decoupling of the cost assigned and the cost actually paid,
Rent-seeking by the medical profession in artificial barriers to entry,
An enormous lock-in factor that occurs at the point of consumption in many cases, when the patient is not in a position to make critical decisions,
Distortions caused by information presented to non-experts (by which I mean drug company advertisements on televisions)

There's so much distortion on both the supply and demand side that what we have can hardly be considered to utilize market forms of resource allocation at all. Instead, what we have is many of the worst kinds of information asymmetry and artificial barriers to entry which commonly plague markets, without many of the competitive pressures that usually make them still function effectively.

One thing that is just about guaranteed to help (a lot) is to make sure people are consuming their own health insurance. That is, they should not be in company-based risk pools, but should have the freedom to join risk pools on their own. That is, the same regulations and tax structures governing employer-based healthcare should apply to other forms of risk pooling. That is, any organization should be able to enjoy the same collective bargaining and regulatory advantages as companies do when negotiating rates for members. This makes consumption of health insurance a lot more direct for basically everyone, and removes some of the distortions caused by the current indirect consumption system.

Another change that would be advantageous is to remove some of the artificial barriers with respect to things like prescribing drugs. Anyone should be able to write prescriptions for drugs for their own use after having gone through sufficient training. For most regularly-prescribed drugs, a day would suffice. This would free the healthcare system of a tremendous amount of waste and friction, as people with chronic conditions needing various common drugs wouldn't have to wrestle with doctor's offices and pharmacies to get them lined up. There are probably some medications for which "sufficient" probably means "medical school." Fine. An approval process for the training program would be needed (and it probably wouldn't hurt to put physicians and pharmacists through it, too, before they were allowed to prescribe or prepare the medication), to be administered by the FDA.

In addition to greatly relaxing prescriptions requirements for self-use, an accompanying change would expand prescription authorization to other healthcare professionals, allowing commonplace treatments to proceed without high-cost, low-value physician supervision.

We also need to examine the way we do quality assessment of treatment and caregivers. The current system of lawsuits and punitive damages and malpractice insurance and hidden outcome data is pretty untenable. An alternative that would release a lot of tension in the medical community is raising the bar quite a bit for non-negligence torts. In exchange, outcome statistics ought to be available. Negligence probably ought to be treated much the way it is now, but the criteria for determining negligence might need to be changed to make it more restrictive. The current system is very easily abused, which leads to a lot of defensive medical practice that is hugely costly. When caregivers see clients as potential litigants, they feel obligated to use a lot of unnecessary and expensive treatment options, for fear of how the failure to have used them might play out in front of a jury. So while the litigation expenses in the current system are themselves only a few percent, those get magnified substantially by the change in behavior among healthcare providers in response to that threat.

URL shorteners

noreply@blogger.com (Greg) — Sat, 09 May 2009 05:08:00 +0000

URL shorteners are a toxic blight on the web.

But there are probably times when you need them -- like when you need to SMS someone a link or something. Some urls are > 200 characters long, and won't fit.

So OK, there will sometimes be shortened urls. Fine. But if we're going to have them, can we please, pretty please, do something about their information content? Something like

http://sho.rt/65tyhK

says nothing. But how about this:

http://sho.rt/nytimes/65tyhK

That's a lot better! Now we have a convention where a shortened url for a page under www.xyz.com gets transposed to http://sho.rt/xyz/[noise]. That is a vast, vast improvement.

For virtually every domain, that's probably enough: just take the domain organization name and leave it at that. It might be worth making a few sub-domain exceptions for sites with a lot of different and popular sub-domains, like http://sho.rt/maps.google/88gB6 or http://sho.rt/video.yahoo/29T4, but ordinarily, things like http://sho.rt/flickr/e5Kr8 are going to be great, providing an extra few characters of information does volumes to self-document what's going to happen when you click the url, which would be good for users.

Publishing and awareness

noreply@blogger.com (Greg) — Fri, 08 May 2009 00:18:00 +0000

Clay Shirky:

"It makes increasingly less sense even to talk about a publishing industry, because the core problem publishing solves — the incredible difficulty, complexity, and expense of making something available to the public — has stopped being a problem."

Well, many of the physical obstacles, sure. But being heard remains very difficult. After all, making something "available" to the public has been possible as long as there's been markers and cardboard.

It is indeed hard to overestimate the impact of the web. The web has made it far easier for far more people to create a scalable vehicle to provide what they know to anyone who wants to hear it. That's a huge change. But that change has done nothing to the laws of human attention. It could be that your blog is potentially more influential than your cardboard sign, but in most cases, no-one is reading either. The problem of getting heard is still there.

I think Dan Carlin had a lot of valuable things to say in this podcast. The gist is that the news business began repudiating ownership of stories long before the web. They started reporting on each others' news stories in order to be fast. Rewriting stories from other sources was a cheap way for the news business to eat its own seed corn (its credibility) in exchange for pumping some viewership numbers. The whole vapidity and blow-dried nature of modern American news is simply an extension of that practice.

All this makes me think several things.

It will be no sorrow to see modern stenographic news die. They earned it and richly deserve it.
The death of the late-twentieth-century news system will leave a lot of fertilizer in which to grow new news businesses.
There will never again be 1000 news outlets all purporting to cover national news (while in reality just retyping each others stories). There will probably be just a handful.
News will fragment, with the unbundling proceeding at a breakneck pace for things like sports and business news first (where there is a possibility people will pay for content), and political and local last.
Reporters and advertisers will be more likely to speak directly to the public rather than bundling their work, but there will always be synergy in that bundle, especially if there are a few dominant venues in which news is consumed.

There has been speculation over which few English-language global news organizations will survive. A few years ago, a reasonable prediction would have beeen that CNN, the New York Times, Fox, the BBC, and probably a few other big newspapers were sure bets. Now I'm not so sure. It would seem that there are no existing news organizations willing to see into the future and make the changes necessary to survive the transition. That's really discouraging, because the continuity of established voices (like the New York Times) is important. You can go out and start a news company, but you can't go out and start the New York Times. It takes decades.

Shorter Marx

noreply@blogger.com (Greg) — Wed, 22 Apr 2009 05:38:00 +0000

Brad DeLong has a nice article about Marx as political and economic theorist.

My takeaway is that Marx made a couple critical errors which are not just defects in his work, but which extrapolate to some of the excesses to which his ideas contributed.

First, the idea that there's a historical imperative, that there's an arrow of inevitable progress, is the kind of shocking mistake that leads to fanatical extremism. It isn't just wrong, it is the sort of Messiah-complex wrong that can get you Stalin or Pol Pot. It doesn't have to, but a big helping of humility and feeling like those who disagree maybe don't belong in re-education camps would be awesome.

Second, the idea that value is related to labor investment, is not just wrong-headed and delusional, it is the kind of error that gets you aggrieved victim kinds of reprisals. The absurd thought that participating in trade is exploitation rather than opportunity is pretty hard to overcome no matter how clear-eyed your economic history is. This mistake doesn't just implicitly set up class warfare and retaliation (with the accompanying messianic belief that opposition is evil), it pretty much explicitly endorses it.

So while it might be said that Marxism, like capitalism, is one of the two biggest social theories that have never been tried, I'm not sure that's really true. I think an honest assessment has to conclude that while these defects in Marx may not inevitably lead to gulags, they are the seeds out of which this kind of extremism grows.

Feeling unproductive

noreply@blogger.com (Greg) — Fri, 10 Apr 2009 05:08:00 +0000

It is hard to get things done. There's a whole industry around solving the problem. There's no question that a lot of people feel less productive than they wish they could be. Why is that? I think there's several reasons.

One of the main ones is expectations. I think that's often because thinking about something that you want to do is a lot easier than doing it. Even knowing that, it is hard to predict how much effort something will take. Let's say you read the humans can run a marathon in two hours. So out you go to the street, and two hours later, you've only gone nine miles! You feel very unproductive, but the reality is that while it may be true that someone has run a marathon in two hours, virtually no-one can do this. Or at the least not without years of training. So despite moving forwards at a steady clip, you may feel unproductive because you're comparing yourself to an unrealistic goal.

This happens quite a bit to all kinds of projects for several reasons. First, an estimate is made with imperfect knowledge, and it is often the unknowns that end up requiring a lot of effort. It's as if instead of estimating it was two hours to run a marathon, you estimated two hours to get to the top of that hill there, except for it turns out the hill is further away than you thought, and there's a cliff and a river in the way that you couldn't see.

Second, and often more importantly, estimates are made at emotionally optimistic times. There's a bug in human wetware which basically manifests itself as euphoria over a new project. That euphoria leads to very optimistic estimates of how easy it'll be, how little time it'll take, and so when reality doesn't match up, we feel unproductive.

Third, another characteristic of human psychology is that almost everyone believes themselves to be a couple levels of competence up from where they actually are. So when you think about a marathon, you think of yourself as a faster runner than you are, so you estimate the time shorter than you can actually do it. Same goes for other kinds of projects. This is probably why feelings of unproductivity are often demoralizing: they make us feel like we aren't as good at the things we thought we were good at as we thought. That's hard.

Another set of reasons productivity often feels low to us is about the mechanics of why estimates are often low: that is, what happens during the execution of the task itself that makes it take longer and require more effort than it seems like it ought to. I think this often boils down to three things: friction, distraction, and misdirection.

Friction is basically secondary work that has to get done to accomplish the main goal. If you're running a marathon, you have to drink water. The effort it takes to get the water, drink it, and deal with the wet pavement aren't directly involved in getting to the finish line, but you have to do it. There's always friction associated with every task. Sometimes parts of the task feel like friction, but aren't (they're just parts of the task you weren't accounting for), but often the friction is annoying and dispiriting because it is friction and nothing else. If you go grocery shopping, it can take a lot longer than you thought because you have to walk around the store a couple of times looking for that one thing, and then stand in line to pay. That's just friction: ideally those parts of the task would occupy zero time, but they don't. We're good at ignoring even predictable friction in estimates.

Distraction is time spent not working on the task, but instead working on other things, whether other (possibly important!) tasks or things that ultimately aren't that important. Stopping to read a magazine while shopping could be a distraction. Or you may realize while in the store that you need to do some important bank business and they have an in-store branch. Either one makes "shopping" potentially take a lot longer than it "ought," and so can feel demoralizing and unproductive.

Misdirection is effort you think is accomplishing the goal, but it turns out it isn't. It's as if while running the marathon, you think you are on the right road, but it turns out you overshot the course and ran a mile out of the way. Now you have to run back. You were running well the whole time, just not where you needed to be going. Unfortunately wasted effort like this is a fractal problem, and is probably a key source of misestimates. In our estimating minds, we are champs who would never make silly mistakes or get sidetracked. In reality, this can happen, and often does, and when it does, it can take a lot of time and feel very unproductive. It can happen on sub-sub-tasks which turn out not to contribute to the goal, or the whole goal itself may turn out to not be what you thought it would be.

What to do? I think some of the modern task management thinking is wise, but I suspect it often won't help. It is wise to be organized and on top of things. If your main problem is distraction, this is probably a big help. But I suspect that for most people, friction and misdirection are more of a problem than distraction. For those people, time management systems may be a source of friction and their granularity a risk of misdirection, and actually hurt more than they help.

An idea for the 'toxic waste' problem

noreply@blogger.com (Greg) — Wed, 25 Mar 2009 06:11:00 +0000

So there's a problem with our banks. They have assets that are troubling them. They're troubling because they are illiquid, resulting in a lock-up of credit and a general mistrust in each other's ability to pay. If you think the other bank has a 1% chance of failing tomorrow, letting them hold your money is crazy.

But these assets have a kind of information asymmetry problem. The bank says, "How about $1 million for this asset." The buyer says, "I'll give you $50,000." The bank says, "Why so little?!" The buyer says, "I don't believe you that it's worth $1 million!" The bank says, "Well, we paid this gentleman in a suit over here to put an AAA price tag on it! Of course it's worth $1 million!" And the buyer says, "Yeah right" and walks away. The idea is that the bank knows more than the buyer about these assets, so the mere fact that it is trying to sell them is a signal that you shouldn't buy them at all. This is kind of like the lemon car problem: if you don't trust the used car salesman, the fact that he is selling the car at all is a red flag.

But what if the banks had to do the following thing. They parcel out all their mortgage-related securities into bundles. At that point, some public and some private parties (or perhaps partnerships somewhat similar to those envisioned by the Geithner plan) say how much they'll pay per bundle. The trick is that only some bundles can be sold, but they all have to be sold for the same price, and furthermore, the order of choosing is the following: first the government picks its bundles, then the private investors pick theirs, and then the bank keeps the rest.

This would create an immensely powerful incentive for the banks to create bundles of precisely equal value: since they go last in picking the bundles, they had better make them all identically valued. This is like splitting cake between two kids: you have the first kid cut the cake, then the other kid choose the first piece. This makes the first kid very, very careful to cut equal pieces.

So if the bank made unequal bundles, then the private parties and public parties would underbid, take the most valuable ones, and leave the trash for the bank. The government's leverage works like this: it relies on the private participants to set price, but it goes first in picking. Perhaps it simply picks at random. Then the private party is incentivized to bid an accurate price -- if it bids too much (on the theory it'll just grab the sweet bundle), the sweet bundle may simply disappear. (Competition is intended to make the bids not be too small.) The leverage is that the government buys 10 bundles for every 1 bundle private parties buy.

In terms of cake, this works like this: the bank has the cake, we want to give them a fair price for the cake to reduce the uncertainty in their balance sheet. The problem is, no-one will buy a piece of cake because they're worried that some pieces have no filling, and the bank knows how to cut slices with no filling, and will just sell you those. So the market for cake freezes. Furthermore, the cake is simply too big for anyone to buy up the whole thing, which would ensure that they don't get cheated. The government could buy the whole cake, but has no idea what it should pay. So what we do is make the bank cut 16 slices of cake, then ask private parties to bid for the right to get two or three slices. The government then buys 10 slices at the market price. The government picks 10 slices at random. The private parties get their pick of another two or three, and the bank selling cake gets the remaining couple.

As a side benefit, now that there's an agreed-upon price for these slices, a) the bank can write down whatever losses it took on the cake, b) it just got a fat roll of cash, c) it can now sell it's own slices on the market as bundles, or re-bundle the remaining assets and repeat the trick.

This would undoubtedly leave some banks bankrupt. (It is analogous to the 'stress test' that's been proposed by the Treasury.) Those can be taken over and basically nationalized and dealt with. Those who survived to tell the tale would have accurate balance sheets that counterparties could trust, and credit in theory would be available.

The bottom line? Taxpayers are exposed to the same risks and getting the same performance from their portfolio as the hedge funds and private equity that bought the bundles from the banks. Even if bank employees who have inside information quit and join the hedge funds to cheat, the fact that the government picks first and at random, and that the banks go last, leaves very powerful incentives for the banks to make the bundles as even as they know how.

Who makes the laws around here?!

noreply@blogger.com (Greg) — Sat, 07 Mar 2009 05:57:00 +0000

I liked this post about the high-finance shenanigans and what we should do about them.

I think we should go further, though. Congress should pass a law saying that any CDS contracts held by or owed to American companies are non-binding. If the parties want to continue to pay them, fine. If they don't want to, fine. It is as if they simply never happened.

This would have much the same effect as the bad bank strategy that's been floated, but even the threat of that legislation should be enough to force the value of these instruments to plummet to zero, where they belong, wiping out the miscreants who got us into this mess in the first place. Good riddance.

It'd probably also be more efficient, as the banks would then, on their own time, trade these worthless assets away in some kind of gigantic orgy of write-offs. When the dust settles, the people who were defrauded by people selling them junk bonds hidden inside AAA rated firms, like AIG or whoever, should be allowed to file personal civil liability lawsuits. Then they can all exhaust themselves suing each other, keeping vast armies of paralegals fully employed until the sun goes nova.

Social capital and the trust deficit

noreply@blogger.com (Greg) — Wed, 04 Mar 2009 05:07:00 +0000

I think part of the story of the 2008-2009 financial problems are the ills of the rising importance of the social aspects of the organization of the economy. There have obviously long been important social aspects to the economy and how business is organized, but it seems to me that our economy in the past few decades has come to rely more and more on the social or psychological dimensions of value than it has in the past.

Some examples. The main obstacle to creating a new manufacturing company in the 1900s was capital. You needed a lot of money to buy the machines. The main obstacle to creating a new information company in the 2000s is people and social connections and finding a place in the psychological environment. Think about a classic social enterprise like news (or, ta da!, banking). The value of a newspaper has always relied on its reputation and its occupying a particular niche in the psychology of a region. But whereas it used to be that sufficient cash could buy the presses and distribution systems, and those were the main obstacles, today those obstacles to creating a news organization are much smaller. There are still the barriers of building a good reputation, which are high ones, but the point is, the decrease in capital barriers for information economy businesses means that the social aspects are relatively much more important.

Even a classic manufacturing thing like making shoes looks far different. Nike is mainly a branding/marketing company, which outsources the apparent content of its entire business. The value Nike is adding is around patents, licenses, knowledge, connections, marketing and branding. In similar ways, the outsourcing of manufacturing has impacted other industries as well, rendering their social dimension more important than their capital-dependent dimension.

A standard way of talking about this is the "U of value" -- the observation that in modern American product companies, most of the value (in terms of where the profits go) come at the beginning of the development cycle -- in product design and building the connections for manufacturing -- and at the end, where the marketing and branding are. That is, the economy in the 2000s is much more dependent on the ability to coalesce brilliant teams of design experts, build networks of supply and distribution, and field marketing campaigns than it is about TQM on the assembly line.

Fair enough. But what this means all around is that the economy is much more dependent on something which policy-makers have a harder time getting a handle on -- social capital. When capital is the main bottleneck to economic activity, then financial operations with the money supply are a great lever. When the main economic activity bottlenecks are people talking at parties, building trust, making connections, traveling to visit their counterparts and bonding with them, gelling with teammates, and such, it is much harder to get a handle on that economic engine through monetary means.

The banking system is a good example. Banks have always relied on their reputations for a big part of the value they provide. That's not news. But as with anything else, falling capital barriers means that actors which might not be big, but with a lot of social capital, can have tons of influence. The bubbles we've seen could be described as ills of socially-organized economies. People follow each other. They rationalize actions relative to what they see others doing. They choose who to trust. (Or, perhaps more accurately, they shop around for someone to tell them what they want to hear and then trust that source.)

The point isn't to say that people weren't important to the economy in the 1900s, but to point to the banking industry as a key initiator of the current problems as evidence that these kinds of more social and psychological issues of trust, risk tolerance, and personal connections are now not just peripheral features of the most esoteric parts of the economy, but are a lot more deeply embedded in the economy than they used to be.

The primary mode of transmission of this crisis -- credit -- is an indication that psychological factors are key. Credit is basically the financial representation of trust. If there's a trust shortage, there'll be a credit crunch.

Which means that when this economy is sick, the ways we learned to cope with a sick economy whose principal engine was capital may not work as well. We need new schemes. More psychological schemes. Schemes tailored more towards trying to make sure that trust can be safely bestowed.

New Deal vs. WWII

noreply@blogger.com (Greg) — Mon, 09 Feb 2009 08:36:00 +0000

One of the main criticisms of the stimulus package being considered right now is that Roosevelt's New Deal project didn't work, only WWII did. (via Yglesias)

But here's the thing. Think about what the New Deal built, and think about what WWII spending built. The New Deal built power lines, dams, bridges, etc. All kinds of stuff we depend on to this day. What did the colossal wartime spending build? Basically things that were all blown up.

It is true that New Deal spending was definitely smaller than war spending for World War II, but that ought to teach us something! And that is this: it apparently doesn't matter if you hire people to build things that you will then just tow out to sea and sink. What matters is the process of hiring them.

This time around, we have the luxury of not having to hire people to build things that blow up, and we can be grateful we don't have to hire them to risk death fighting some evil force bent on world domination. Instead, we can hire them to build another round of modern infrastructure which can provide enormous value to our society for another half-century.

Unemployment benefits and the deflationary trap

noreply@blogger.com (Greg) — Sat, 31 Jan 2009 06:36:00 +0000

As we wrestle with what to do about the financial crisis, I find myself convinced that better/smarter unemployment benefits are a key part of it. There is plenty of talk about how people who are unemployed are good targets for stimulus because they'll spend, and about how they'll benefit from the jobs created by infrastructure development projects. Here's another angle.

Krugman &co keep warning about the deflationary trap. Basically, the problem of having falling real wages and prices, which then tend to get baked into business and personal planning, creating a very hard circumstance to avoid. How can good unemployment help this? By making workers not mind as much going without jobs rather than staying, and making businesses less reluctant to do layoffs. Paradoxically, if businesses ask workers to stay on and accept pay cuts, this can feed into a deflationary spiral. The model is that since the cost of labor goes down, firms can cut the cost of their goods to capture the reduced consumption. Consumption is reduced because workers expect their wages to go down. Thus the deflationary trap.

The news business

noreply@blogger.com (Greg) — Tue, 13 Jan 2009 05:25:00 +0000

The news business is in trouble. There are various theories about what's going on, but basically, it is much harder to have an ad-supported newspaper than it used to. Why? Aaron Swartz points out that it used to be hard to deliver ads to people, and now it isn't. Is this really the issue?

The basic premise is that ad money follows the audience. The audience is now more fragmented, it is true, but audience laws still follow a power law and there will still be larger and smaller vehicles for ads. The ease of advertising online I don't think can be blamed. While it may be easier, that should allow more people to do it. It isn't clear why this makes the total market shrink.

I suspect that the ugly truth is that newspaper advertising was never as valuable as it seemed. It was just that there was no alternative. Another ugly truth is that newspapers aren't as valuable as they used to be. This is for two reasons. First of all, there are way too many. There is no reason why there are 100 microphones in front of someone at a press conference. A couple will do. The market is being asked to pay dozens of distributors to do the job of just a few. There's no dramatic regional angle why we need 20,000 similar stories on the strife in Gaza. Wouldn't 2000 do the trick? Or 200? Or perhaps 20? There is a gigantic oversupply of news. And at least the Gaza story is substantive. We need barely any coverage of the various celebrity-scandal-o-the-day type that many newspapers waste effort on.

In other words, if the market were consolidated into say a dozen or so large national newspapers, the quality of news would probably actually go up, not down, and those organizations would be able to draw the kinds of audiences to make a go of it. Local papers would be very heavily locally focused, perhaps syndicating articles, perhaps wrapping their sections with the national bits. Perhaps going free.

Oddly, I wonder whether it is in fact online entities which will end up becoming these organizations. I wonder if the newspaper business is too wedded to the past to be able to adapt, and that with perhaps a couple of exceptions, will just die out, and be replaced by expanded online news organizations with growing budgets and an appetite to become news hubs with print components to add to their electronic presences.

Torture policies

noreply@blogger.com (Greg) — Tue, 16 Dec 2008 07:56:00 +0000

Wow. I just read the Executive Summary and Conclusions of Report on Treatment of Detainees in U.S. Custody, released unanimously by the Senate Armed Services Committee. It isn't really news, at this point. Most of the facts have been discussed in previous books. But the report puts together a timeline and the names attached in an official way that will hopefully be fruitful for further investigations.

What's been clear for a while is that the plan on torturing prisoners in US custody was conceived and directed from the center of the Bush Administration. The report spells out the details of the progression. Weeks after the terrorist attacks of September 11, President Bush issued a memo that said that detainees in this conflict could be treated without regard to the restrictions of the Geneva Conventions. This was interpreted, by people with intimate access to the President (and who may have been responsible for helping him come to that view in the first place), as authorizing them to construct legal fictions authorizing the adoption of torture as a standard operating procedure to be used on US detainees captured in the conflict.

Which they promptly did. Jim Haynes, the DoD legal counsel (and a Bush appointee), railroaded some legal cover for Rumsfeld to rely upon to authorize the use of torture for detainees.

The source for these torture techniques was the SERE program, which is a program run by the US armed forces to teach US military personnel how to resist illegal torture if they are caught by an enemy willing to break the Geneva Conventions. The techniques themselves are the ones used on US military by their enemies in Vietnam and Nazi Germany, and have been known for a long time to elicit false confessions. In other words, the SERE program is a perfect place to learn how to illegally torture detainees, and that's where Bush administration officials looked. It is clear to any onlooker that any legal fiction invented to justify the torture of enemy detainees using these methods is illegal and probably counterproductive.

There were hiccups as military lawyers raised fierce objections to torture being introduced as a standard technique for the US military. Rumsfeld authorized torture at Guantanamo in Dec 2002, then rescinded the order in Jan 2003. Haynes convened a Working Group, which relied on made-to-order legal opinions from Yoo and Addington, thus providing a legal fiction on which Rumsfeld relied to re-authorize torture in mid 2003.

By that fall, SERE trainers (who had instructed US military personnel in torture techniques at Guantanamo) had traveled to Iraq and Afghanistan to teach military personnel how to torture prisoners there as well. As the techniques were adopted (piecemeal) in various locations, they became "confused" (according to the post-Abu Ghraib investigations), and contradictory information was in circulation. Some SERE instructors thought the techniques were being used in dangerous ways. Subsets were authorized, then de-authorized. Meetings in which some generals objected weren't communicated to everyone.

In other words, there were two messages: one which said "it is OK to torture detainees in these several ways." This message came directly from President Bush and the civilian leadership of the Department of Defense, and no doubt found many US military personnel in support. The other came from many military lawyers, many generals, and no doubt many other US military personnel. That message was "do not torture detainees." From the beginning, many military lawyers and generals had protested vigorously to torture, and it is no surprise that in this climate of controversy and the attempt by the civilian leadership to authorize illegal torture practices in standard operating procedures under cover of railroaded legal fictions, that the exact details became confused.

It was in this climate that the abuses at Abu Ghraib took place.

The SERE torture techniques that the program trains US military to resist include, specifically: nakedness, stress positions, hooding, being treated like an animal (that is, leashed, forced to perform tricks), cold, poor food, slaps to the face and stomach, waterboarding, humiliation, and plays to specific phobias. Furthermore, as the Levin-McCain report makes clear, it is exactly these torture techniques which the Bush Administration was interested. A good example is the idea that Arabs have a fear of dogs. The SERE techniques include exploiting these fears. Several books and testimony shows that this was discussed explicitly in White House meetings, written into policies for Guantanamo and in Afghanistan and Iraq, and used on prisoners. This specific idea is a good example of how this list of torture techniques, adopted from a program known transparently to those who drew from it to be training US military to resist illegal torture, spread through US military operating procedure.

These techniques are manifest in the photographs of torture at Abu Ghraib. Forced nakedness. Stress positions. Intimidation by dogs. Being forced to behave as an animal. Cold. As such, the conclusion of the Levin-McCain report: that the actions taken by the US military personnel shown in the photos were, by and large, understood by them to be authorized as part of the torture regime which had been approved at the highest levels of the Bush Administration, the Pentagon, and in their own facility.

There are still some blank spots in the report. Why did it take the SERE program and its lawyers so long to realize that their specialties were being illegally exploited offensively by the military? (It was not until late 2003 that they started objecting to this use, and not until late 2004 that their lawyers and generals put a stop to it.) Did SERE trainers who traveled to these locations understand that they were training US military personnel to illegally torture detainees? It seems hard to believe they couldn't have. Did their superiors, who must have known of their trips to these locations, understand the purpose? It seems almost certain they did.

There are details to fill in. The details of how the practices spread from location to location are unclear. For example, the general in charge of Guantanamo visited Abu Ghraib as the torture techniques were coming to be practiced there. How important was that event in their adoption? It seems clear that many generals accepted these illegal practices, and many others did not. What did the objectors do to try to stop these illegal practices, and why did it take so long for them to be successful? Are there records of their objections with high level Department of Defense leadership? What was the disposition of any such objections. For example, by Dec of 2003, the military lawyers had informed Rumsfeld that he could no longer operate under color of the pro-torture legal fictions invented on his behalf. Yet it took at least another year for that to actually become felt as a matter of policy. This seems almost certainly due to active unwillingness to disseminate this policy, but the details of that aren't spelled out. It seems that reviews like this may have been as a result of high-level objections to torture policy, but who said what when isn't detailed in the report (at least, those portions released at present).

Perhaps the declassified version will fill in some of these details. It is important to figure out where the threads of responsibility lie. The reason is that many of these acts are war crimes of the sort for which Nazis and others were executed for by the United States.

Licensing Natural Images

noreply@blogger.com (Greg) — Sat, 06 Dec 2008 17:08:00 +0000

Sometimes it is hard to get a good picture of how much value humans get from having animals and their habitats exist. One idea that might make this more clear would be to charge for the use of their images. That is, in the same way that it costs a company or organization money to use the likeness of a celebrity on their products, it would cost them money to use the likeness of a fox, or eagle, or valley, or mountain.

Of course the hard part is figuring out who would administer these licenses, but a revenue-maximizing monopoly would be a decent way to start (same as with a celebrity). Such a venture would then be obligated to use some large fraction of the revenues on conservation efforts. Or it could be structured as an agent contract. That would make it easy to figure out if the agency price was fair: compare to what human celebrities can negotiate with their agents.

Complicated Assets

noreply@blogger.com (Greg) — Wed, 03 Dec 2008 03:47:00 +0000

I was reading Matthew Yglesias and Ezra Klein complain about the complexity of understanding the assets involved with the banking meltdown. These are smart people. And while it may be true that the underlying bets are (in retrospect) easily identified as bad, the reality is that it is hard to figure out the securities involved.

I'm also reminded of a successful businessman whose advice is to "buy what you know." Often, I think people don't heed this advice, but when there's a flight to liquidity, they sure do. "You want me to buy a what?!" one imagines traders without the expertise in CDOs, CDSs, super-senior tranches, and the like asking their counterparties.

Thus, precisely when you wish there was a liquid market for these complex securities is precisely when a) people don't want to buy things they can't comprehend, and b) no-one will trust you an inch to explain it to them! (After all, you moron, you thought you understood them and you just got destroyed the economy and your institution.)

Bailout Fail

noreply@blogger.com (Greg) — Tue, 02 Dec 2008 04:51:00 +0000

What's up with the bailout? Tons (literally) of money spent, and still the stock market is in turmoil, economic indicators are bad, and the credit crisis is still severe.

So why isn't the bailout working? Well, it sort of is. The TED spread (the gap between treasury note yields and LIBOR interbank rates) is down from an insanely high peak to more like "intense crisis" levels. But this would have happened with a sensible intervention policy as well. What's making TED drop, in my opinion, is the belief that government will step in to backstop major banks. I'll bet that in fact, a more sensible plan might have actually eased the credit crisis more. The theory would be that with crazy giveaways like the Citi bailout, the public will start clamoring for heads to roll and the backstop guarantee will evaporate as politically impossible (and damn the consequences), whereas if it was conducted in a more sensible way, such a clamor might be further away.

By "sensible" I mean that the government could be demanding far, far more concessions than it is getting. The banks are using bailout money for shareholder dividends, for instance. The government isn't even getting the shares in the banks it deserves for the value of the bailouts it is handing out. The Citi and AIG bailouts are the prime examples here. There's little doubt that Citi would crash and burn without government assistance. The government should demand a much, much higher stake than it did. Given that the alternative is liquidation, is 99% the right number? (That is, a nationalization.) Given Swedish history, it might be. One could argue that the right number is the one that pulls the most non-government capital back into the game. Too little runs the political risk of intervention fatigue. Too much may run the risk that investors, if they pick the wrong bank to invest in, will get their shares confiscated if it goes under. Caveat emptor and all that, but there's definitely that argument that you can go too high. (Unless you're willing to nationalize all the banks in one go.)

It would seem like even Treasury Secretary Paulson's original second-rate plan of having a government-lubricated market for trash mortgage securities would be better than the straight-up giveaway he's ended up doing. (Insert lubrication joke here.) Even better would be direct government investment in banks in return for radical concessions, up to and including nationalization for the worst offenders.

More from Robert Reich and Mark Thoma.

One Reason Why States Lose Power

noreply@blogger.com (Greg) — Tue, 25 Nov 2008 05:36:00 +0000

There are a few reasons why the trend over the years has been for more centralization of power with the Federal government and a diminishing of the role of the states. One main reason is that voters hold the Federal government responsible in hard times. Why? The current economic crisis provides an opportunity to see why.

The lessons of the Great Depression (and other smaller recessions) is that government spending is the flywheel that keeps the economy moving when downturns threaten to turn ugly. There's a catch, though: structural deficit spending is constitutionally prohibited to most states. In that context, states find it all but impossible to make the kinds of policies required to shepherd their economies through hard times. What does the average person take away from this? A perfectly understandable view that when the chips are down, the state government cuts back, but the Feds pour in resources. The state government cuts back because it is unlawful for it to do otherwise---shrinking revenues during a downturn compel it to spend less. But that's precisely the time when the Federal government is able to borrow and spend. So for many people, when they need it most, it is the Federal government that can step in.