One of my guilty pleasures is watching the TV show “Alaska State Troopers” on the National Geographic Channel. The pleasure part comes from seeing spectacular scenery and wildlife like moose and bears I’ll never be able to visit in real life.
However, the guilty part comes because most episodes also pose the question as to whether some of the people the Troopers will have to meet are actually more foolish than the animals.
In one incident, a man took his 15-year-old daughter with him to a party, driving his truck up a side road onto a foothill and parking it there among other party-goers’ cars with no room to turn around. All of the adults at the party proceeded to consume far too much alcohol to legally drive, and no one paid attention when the 15-year-old decided to join the drinking, too.
So now it’s late at night. The sun is going down — this is Alaska in summer after all — and the father realizes he and his daughter have to go home. Therefore, the intoxicated father decides that this will be the perfect time to teach his almost-as-intoxicated daughter how to drive the truck. Of course, since there is no space on the road to turn around, the father instructs the daughter to back the car down the hill by looking in the rear view mirror to see where she’s going.
You can already imagine that this adventure was not going to end well. Fortunately, the truck went off the road and into a ditch so quickly that it had no time to work up any significant speed, and no one was injured. However, the attempts of the father to then convince the arriving Troopers that alcohol had absolutely nothing to do with his truck being in the ditch were comical.
Future space exploration will also contain a similar steering-behind-you problem. Star Trek imagined a technology called “inertial dampening” that allowed starships to accelerate between “rest” and warp speeds and back almost instantly without smearing everyone aboard into a biofilm along the bulkheads. Physicists have no way to imagine inertial dampening, but they actually have conceived of technologies that can get an unmanned probe to a nearby solar system (a mere six light years away) in a human lifetime. Figuring out the details is mere engineering. 😀
Project Daedalus is a design for such an interstellar probe originally created by the British Interplanetary Society in the 1970’s. A 54,000 ton, 620-foot long, two stage starship (sketch shown in figure) would be constructed in Jupiter orbit and then launched toward Barnard’s Star with a fusion reactor engine. To acquire enough helium-3 for your fuel, you’d first have to build a fleet of balloon-suspended robotic mining factories to extract the isotope from Jupiter’s upper atmosphere for twenty years, but, like I said, mere engineering details.
Operating continuously for two years, the reactor would boost Daedalus to 12% of light speed –or about 22,000 miles per second — before shutting off for a 46 year cruise. About half way through, Daedalus would deploy telescopes to start surveying the Barnard System in order to figure out where any interesting planets there might be when Daedalus arrived decades later. Still at 12% of light speed upon arrival, Daedalus would fly through the system throwing out subprobes for closer planetary flybys, radio its and their data back to earth, and continue onward into the great dark.
The mission was planned as a flyby because stopping in the system creates a whole new set of complications. To start, the ship has to be much more than twice as large, because it takes fuel to slow the ship down just as it takes fuel to accelerate it to 12% of light speed in the first place. In fact, the ship is mostly fuel, and it consumes a lot of extra fuel to first speed up the fuel you’ll need for slowing down upon arrival. Furthermore, you can’t, as in the flyby mission plan, store most of the fuel in a first stage fuel tank which you dump on route and arrive at Barnard’s Star with a mere 450 ton payload from the tens of thousands of tons you launched from earth orbit. You’ve got to carry all the braking fuel with you to the end. It takes much longer to get all of that extra mass up to cruising speed, so the whole trip takes much longer than 48 years.
So, depending on how you’ve optimized the specs and what your telescopes have told you about the star system, years before arrival, the ship rotates so it’s pointing back toward earth and refires its engines for those years to slow to orbital speeds. And that brings me back to the steering through the rear view mirror problem again. How accurately can you know when to fire up the engines? Remember, you’re limited to on-board instrumentation and computation because earth is too far away to improve the accuracy of your work, even if earth’s ability to measure interstellar distances has gotten much better in the decades since you left. And you’ve got to start slowing down while you are significant fractions of a light year away. When you’re going 22,000 miles per second, inaccuracies in distance measurements to your targets matter. Be off by 11 seconds in firing your “retros”, and you come to a stop no closer to your destination than the earth is from the moon.
_____________________________________
The point of these little parables is this: it is really hard to direct where you’re going when you have to make decisions by looking backward and can not know exactly where you are in respect to where you’re trying to go. Yet, I now have noticed that the assumption that the government can successfully manage modern economies containing huge international flows of demand, finance, labor, technology, and consumption contains the same steering-by-the-rear-view-mirror fallacy.
President Obama acknowledged the fallacy — perhaps without quite realizing it — earlier this week when he responded to an interview question about why he had failed to cut the annual Federal deficit in half as he’d promised in 2009:
“Everybody who is out there back in 2009, if you look back what their estimates were in terms of how many jobs had been lost, how bad the economy had contracted when I took office, everybody underestimated it. People thought that the economy contracted 3%. It turns it retracted close to 9%. We lost 8 million jobs just in a year’s span, about half a year before I took office and half a year after I took office,” Obama said. “So, the die had been cast, but a lot of us didn’t understand at that point how bad it was going to get.”
There’s some truth in that, but it proves the larger point I’m trying to make: it took more than a year after the fact before the government could gather the information to notice that a recession was really a Great Recession. The Obama Administration then spent that year focused on expanding health care entitlements and “investing” in next-generation green energy projects — which we now see are becoming an ever larger economic albatross — because it was falsely convinced by its own outdated data that a strong economic recovery was quickly guaranteed.
Similarly, two years earlier, before the housing bubble popped, the Bush Administration, the major banks, the Federal Reserve System, and the Democratic congress were all being fed equally out-of-date information from the bureaucracy that led them to believe there were no major risks to the system, so rear-viewing the economy is a bipartisan risk.
And we are still doing exactly that!
The Bureau of Labor Statistics publishes more than most people ever want to know about employment statistics through the menus here. I’ve been spending a lot of time wandering through the information since last fall when the reported unemployment rate began to fall, but I noticed that employment was rising only among males. The number of employed adult women was continuing to decline.
In fact, the number of seasonally adjusted employed adult women declined compared to the same month in the previous year (when “seasonal” factors should be similar) in 10 of the 12 months of 2011. By comparison, the number of seasonally adjusted employed adult men rose by over 500,000 per month compared to the previous year in every month of 2011. So, either there was a sudden gender-related new trend in labor participation over the last year, or the data samples were missing something important.
Turns out it was the latter. In the January 2012 data, when the BLS updated (as it does annually) population estimates — as is necessary to project employer sampling results over the entire economy — the number of men employed was only 90,000 more than in December. However, the number of women employed shot up by 700,000. And, due to the same correction, the number of adult women officially counted as “not in the seasonally adjusted labor force” went up by 780,000 as well (there was negligible change in the not-in-labor-force number for adult men between the two months).
Now, these 1.5 million women didn’t suddenly materialize on New Year’s Eve. They’ve clearly been here throughout 2011, but policy makers are only now finding out about them as 2010 Census data is better analyzed. The Census Bureau, according to oral reports I’ve heard, suggests these women are mostly over 55, meaning that they are approaching the age when they will be drawing on entitlements for their care. They show up most importantly now in producing a lowered labor force participation rate -– or, in other words, in the number of people we assume will be paying taxes and producing economic output to support future entitlements for all you young whipper-snappers. And a lower labor force participation rate means that the official unemployment rate (which is number unemployed divided by the sum of the number employed and the number unemployed) shrinks, even as the future debt problem gets amplified.
So just as the bureaucracy underestimated how bad job loss was getting because their data were outmoded in 2009, they are now overestimating how rapidly jobs are improving because things that actually happened a year ago are being counted as if they are happening today. This cannot simply be extrapolated to continue as if this was merely a delayed, but routine recovery from something that was not a Great Recession. You can’t just “party on, Dude!’
The BLS has to use seasonal adjustments in the first place because there are normally huge month-to-month swings in employment (for reasons, like weather and holidays) that have nothing to do with the underlying health of the economy. You add jobs for some months, and reduce them for others so that over the entire year a better picture of longer term trends emerges. Then you calibrate your formula for the next year based on later, more complete data from other benefit-related government sources on total employment.
As John Crudele noted:
“There really weren’t 243,000 new jobs created during the first month of the year. But there were 2.68 million [unadjusted] jobs lost…The positive figure was derived solely from favorable seasonal adjustments. Or, as Calculus99 wrote on CNBC.com: ‘Perhaps the reason why the consumers are feeling gloomy is because they don’t see the jobs that the statisticians tell everyone are being created.’ ”
“In a normal year, the economy picks up statistically in the spring and then swoons in the summer. This year — because the … weather this winter has been warmer than usual — there might actually be a dip in economic activity this spring.”
“…The Labor Department [also] removed a huge block of 367,000 jobs it thinks, but can’t prove, disappeared in January as companies quietly went out of business. But the assumption about the seasons outweighed the assumption about invisible birth/death jobs.”
Now, I am not suggesting in any way that these adjustments are being made to “cook the books”. They are the best our measurement systems are capable of doing with the time and budget constraints of current sampling technology, and in most stages of a business cycle, when large numbers of jobs are being gained or lost, they probably work pretty well. When you’re at the peak or trough of a cycle, when job growth is stagnating, isn’t one of those times! The “corrections” are several times as large as the jobs signal we’re trying to measure, and so errors in the corrections can be large enough to mislead policy-makers. Realize that due to sampling uncertainty alone, monthly job estimates come with an uncertainly of + 100,000 jobs before you include any uncertainties from things like seasonal adjustments and the business birth-death model. Yet the stock market and political commentary go manic-depressive based on whether or not the official jobs number is 50,000 more or less “than expected”. No one actually knows enough to “cook the books” even if they wanted to do so.
I note that we are continuing to wait (for short-term stimulative expediency) to do a “full-burn” on our retro rockets in regard to government spending. I expect we’re going to be a bit more than 11 seconds late when we finally discover that the system we’re heading toward looks entirely too much like Greece. Or, applying to the economic recovery the words of the old Meatloaf song, “Objects in the Rear View Mirror May Appear Closer Than They Are!”
Physicists have no way to imagine inertial dampening
The inertialess drive rotor cancels out inertia.
Firetag:
I think this is a great article, and point well taken. I’ve heard it said that political economics is a lot like building a house on a moving truck. It’s not like the data is entirely or satisfactorily manifest at the time of kee-jerk policy making. It is often quite later.
A term paper I did for a Macro-Economics class once compared macro-economic model variances to the old board-game “Mouse Trap”. Macro-economic models in my opinion are often guilty of a kind of slippery slope logic. If economics was truly built on certain stable constants, rather than human behavior, then this wouldn’t be much of a problem. But just like the game Mouse Trap there is way too much unaccounted for variance. If you remember watching the old commercials for Mouse Trap, you would see a happy family effortlessly setting in motion a chain of events, starting from a rather simple action, that culminated predictably into a cage falling onto the unsuspecting mouse. This is sort of how government economic policy is sold. The logic seems good, so we go for it. Nevertheless, as anybody who has actually played mouse trap knows, the game was a piece of junk. Getting the “system” to actually work as advertized was sloppy, tedious, and anything but predictable. In sum, there were all sorts of unaccounted for variances inherent in the pieces, and the board, assembly, etc, that caused the system to fail.
Sometimes, the very best thing a government can do, is the most unintuitive and emotionally unsatisfying thing possible. Nothing, just sit and wait.
Cowboy:
“Mouse Trap” is a very good analogy. Of course, the complicated and subject-to-error schemes our policy makers are building are intended to get we mice OUT of the recession trap without dumping us into an inescapable debt trap.
My kitties are already sharpening their claws in anticipation of mouse-kabobs, I’m afraid.
LDS Anarchist:
Glad you read the post, and I looked at your rotor web-site, but to head-off a complete off-topic discussion, what are your views on the MAIN point of the post: relying on rear-view data to plan for an economic recovery. I’d be particularly interested in your views, since, as I recall, you are knowledgeable of Europe, where the debt crisis is much more advanced.
Curiously, you did not mention a single event leading up to the economic situation of 2009. Two wars, huge tax cuts, unfunded prescription drug plan, the housing meltdown…you write as if 2009 was dropped on us from an airplane. This strikes me as intellectually dishonest, unless your only goal is to make a cheap political argument.
Don:
My post’s point was to discuss the limitations of trying to make policy with the inadequacies of current data. Data was not better in the past than it is today, so the policy blindness would have been even WORSE in the past. In fact, the existence of the two wars and the housing meltdown kind of proves the point that data limitations aren’t just limited to the Labor Department. We didn’t see the 911 attacks coming; we thought policies that used Fannie, Freddie, and the banks to provide high-risk mortgages could never blow up in our face. We did see the problems with entitlements coming at least as far back as the Clinton administration, but our political leaders of both parties decided they could handle things later after they solved the problems of preserving their own immediate political futures.
I am quite confident that a more extensive review of the BLS historical data as published at the time would turn up more oddities, but errors from before the last recession have probably been caught by the ongoing historical revision process by now.
It’s the ignoring of the errors and data uncertainties that are affecting policy decisions NOW that concern me.
Re LDS Anarchist
Sorry, I subscribe to the Battlestar Galactica method of FTL travel. 😉
Re Cowboy
Yeah, absolutely.
Re FireTag
I dunno. I’m a little frustrated by several things. The Obama and Bush Administrations indeed say that no one knew. But people DID know. There were many people who predicted the collapse, and many people who predicted about how bad it was going to be. The problem is we all thought they were crazy. And this is what rubs me the wrong way, sometimes Chicken Little is just being Chicken Little, and sometimes the sky really is falling. But the fact that they claim no one knew, alerts me to the fact that they don’t even listen to Chicken Little, let alone just disagree with him.
I guess I don’t need convincing, because I already believed what you wrote. It just felt natural to me. Maybe it’s the engineer in me. I’ll give a quick recapitulation of my view in engineering parlance in the hopes it will translate to something other people can understand:
The economy is a really big system. It has inputs, and outputs. The question is, how do we get the desired outputs? Well, one way (the way we’ve chosen) is to put in a controller (alongside the market, which I think is a good thing). The controller needs access to the outputs of the system, and it gets this via measurements (sensors, if you will). As in all systems, the measurements aren’t perfect and have error associated with them. So now we have two problems, designing the controller to obtain our desired output, and filtering out the error, or designing the controller while accounting for the error. In our gov’t, we have chosen to design the controller without either filtering the measurements, or accounting for them appropriately. As you say, it’s not entirely our fault, it’s very complex, but be that as it may, we haven’t accounted for those errors appropriately.
In reality, those noisy measurements are compounded by the fact that they’re delayed measurements. As you point out, we have outdated data. So guess what happens when you use delayed measurements in a control system? Unless you know the delay almost exactly, and appropriately account for it, you go unstable.
I understand that I’m not an economist, but I know a hell of a lot about systems, and I know how to identify when one looks like it might fail. You can’t have a system with bad, and delayed data, and expect to appropriately control it. Or rather, you can, but it will be marginally unstable at best, and will go unstable easily.
As Andrew has pointed out in the past, just because we don’t yet have a perfect model of the economy (BTW, if we did, we could dispense with the market altogether and just have a central planner), doesn’t mean we should quit trying to figure it out. I think this leads to the ultimate question (as I see it) of the post: is it possible to have a good enough model of the economy that would justify dispensing with the market? Personally, I think it’s problematic because of the agency of the economic actors, but it may be possible. As long as we can’t predict or appropriately measure the decisions and desires of the economic actors, I think we have to have markets, and therefore will be subject to this bad data that drives too many of our decisions. In my opinion, economics and the gov’t have neglected an important resource in that regard (probably due to privacy issues): the internet. In theory, the internet could be used to obtain more accurate information and more quickly, which would lead to better controller design.
JMB:
I appreciate the point you’re making, but I’m not sure that we understand the nodes of the internet and their feedback loops to and from the current “controller” to anticipate any new forms of instability they might have. A lot of research to do there — if we don’t go bankrupt first.
One problem I’d like to solve — and I’ve been struggling with it since I first started working at the academic-government interface almost 40 years ago — is getting policy-makers to understand the limits and strengths of their own data inputs. As much as academics may wish it otherwise, economic data does not drive policy choices as much as it acts as a support or obstacle to implementing pre-existing policy biases.
JMB:
I should also add that you’re right about the notion that people who warn about things people don’t want to hear are not believed. This time there is a consensus on BOTH sides of the political aisle about what’s coming, but a very Louis the Fourteenth-like “after me, the flood” attitude.
To quote Mark Steyn column in National Review last week:
“Have you seen the official White House version of what the New York Times headline writers call “A Responsible Budget”? My favorite bit is Chart 5-1 on page 58 of their 500-page appendix on “Analytical Perspectives.” This is entitled “Publicly Held Debt Under 2013 Budget Policy Projections.” It’s a straight line going straight up before disappearing off the top right-hand corner of the graph in the year 2084 and continuing northeast straight through your eye socket, out the back of your skull, and zooming up to rendezvous with Newt’s space colony on the moon circa 2100. Just to emphasize, this isn’t the doom-laden dystopian fancy of a right-wing apocalyptic loon like me; it’s the official Oval Office version of where America’s headed. In the New York Times–approved “responsible budget” there is no attempt even to pretend to bend the debt curve into something approaching reentry with reality.
As for us doom-mongers, at the House Budget Committee on Thursday, Chairman Paul Ryan produced another chart, this time from the Congressional Budget Office, with an even steeper straight line showing debt rising to 900 percent of GDP and rocketing off the graph circa 2075. America’s treasury secretary, Timmy Geithner the TurboTax Kid, thought the chart would have been even more hilarious if they’d run the numbers into the next millennium: “You could have taken it out to 3000 or to 4000” he chortled, to supportive titters from his aides. Has total societal collapse ever been such a non-stop laugh riot?
“Yeah, right.” replied Ryan. “We cut it off at the end of the century because the economy,according to the CBO, shuts down in 2027 on this path.”
The U.S. economy shuts down in 2027? Had you heard about that? It’s like the ultimate Presidents’ Day sale: Everything must go — literally!”
Maybe a prophet is what we call a Chicken Little only after the sky DOES fall, but in this case, people seem to not want people to know about long term economic problems for society in order to solve short term political problems for themselves.
Aren’t there smart people at MIT that have figured this out yet? No…well, I have no shot at it then. 😉
All I think is while we can agree there is a problem with relying on historical data to accurately take into account all factors to project future situations, we are far better off with slightly dated data than just not trying and doing nothing at all.
Could we be in the Great Depression #2 right now had they not learned from the past and taken some measures (even if they didn’t solve the whole problem)?
I also disagree with JMB that people really knew and predicted the collapse and just weren’t listened to, any more than half the ESPN analysts “knew” the Giants were going to win the SuperBowl. There are arguments for both sides, we don’t know until it happens and we look back to confirm it. Its all probabilities. Eventually, someone can guess right, but did they really know? (Think, May 21 2011 predictions…pushed to Oct 21, 2011 Judgement Day…and still guessing??)
That’s my attempt at giving 2 cents. 🙂
Heber13:
That’s a good two cents, if the currency hasn’t been devalued too much. 😀
But then in regard to #9, by the same logic, shouldn’t we be trying to do something NOW about controlling spending more than just calling a minor slowing of spending increases a draconian cut even if we don’t get everything exactly right?
Tax increases are already locked in by current law to have a major destimulating effect at the beginning of 2013 as current cuts expire. If we don’t also lock in real spending cuts NOW for the later part of the decade, we can’t brake the spending in time to avoid the 2027 cliff predicted by the CBO.
Re FireTag
Yes, that is critical.
Agreed 100%. Not only do they not want to know, but the system itself incentivizes short term thinking. Personally, I think this is one place where the Constitution probably needs to change. We currently rely on the moral uprightness of our politicians to think long term and “do the right thing.” But it’s a flawed system with flawed people. I think a reworking of the terms of office for the various branches might help some. Although I recognize that would introduce more problems as well.
Re Heber13
But this was Cowboy’s point above, that sometimes it IS better to do nothing at all. Let me draw an analogy for you. In a car, if you oversteer to avoid an obstacle, you require countersteering to correct your error and stay on the road. But if you oversteer your countersteer, you then need a larger countersteer to counteract the original countersteering effort. Eventually, by providing too much control input to the car, you cause the car to go unstable and you crash and die. You are better off taking your hands off the wheel, and letting the car return to its stable equilibrium of driving straight. The thing is, this happens ALL THE TIME. In airplanes, in cars, in virtually every system, if you do too much, you wreak havoc.
This is a tricky question and is encumbered with lots of speculation. Would you prefer to have your wrists slit, and repaired immediately, or would you rather have daily chronic migraines for the next 10 years? Some people believe that if they hadn’t done anything, the banks would have collapsed, we would have seen horrible economic conditions for a short time, and we’d well on our way to a healthy economy by now. In other words, to some people, doing something may prevent acute disaster, but prolongs the recovery. I don’t know enough about it to be convincing, but I think they’re worth listening to. The Great Depression isn’t as cut and dry and our political pundits would suggest.
That’s just silly Heber13. That’s like you telling your kid “well, you might not go to jail if you steal from the store, it’s just probabilities.” I mean sure, it’s all probabilities, no one can predict the future perfectly. But some schools of thought are better than others. There were people who predicted with pretty good accuracy what was going to happen. To then say “oh, well, they didn’t really know” is ridiculous. The bottom line is, everyone ignored them, called them crazy, and went on with their lives. Nevertheless, I think there is an important point here, and that’s that economic forecasting is tricky business, and even those who correctly predicted the collapse have made bad predictions other times.
The collapse of the housing bubble was broadly predicted by a number of sources, but the momentum desire to milk the bubble for all that it was worth, outweighed reason. I bought a home in 2008, and became a little obsessive in studying the housing market. I predicted the collapse, though I am not suggesting that I was a unique expert. The fact is, everybody knew that the market would collapse. When people say, “nobody could have predicted this”, what they really mean is that “we couldn’t predict the timing within less than five years”. Across the board home values, that normally appreciate at a rate of 3-6% annualy, we’re growing at rates as high as 13%. Anybody with a little experience in compound interest understands that given the rates already exceeded inflation, home prices would fall (off of a cliff) as the cost exceeded the markets willingness and ability to bear, coupled with the saturation of new construction. Also bear in mind, that a large share of the development was in speculation, from individuals who had no intention of living in the home. This fosters a situation where the supply can quickly outpace the natural demand. And, this isn’t even physics. It’s pretty plain vanilla economics that every “economic actor” understands, whether they are trained economists or just people who earn money and spend it.
As for doing nothing in an economy. That is not a permanent strategy, but the logic goes hand in hand with Firetags point. Firetag is arguing, among other things, that part of the problem with economic policy making is that the quality of the data is poor and untimely. Now, that certainly leads to a good discussion about improving our data gathering and system of metrics, but that is another discussion. With our current limitations on data, we are subject to variances that are already inherent in the system, but that we don’t understand. Furthermore, by constantly feeding the emotional desire to “do something” we exacerbate the overall variance in the system. Statistically what that means for economics, is we can increase both the probability for worsening the economy, as well intensifying the severity, depending on our variables. I’m not an engineer, but I have experience in process improvement with business, and I can say that whether we are talking economics, or firm production, Statistical Process Control is governed by the same principles.
I am attaching a good video demonstration on an old experiment developed by W. Edwards Demming, called the red bead experiment. This shows how constantly adjusting a system/process on the basis of limited data, actually produces worse results, even though intuitively we have a desire to fix things. Conversely, he shows that when you collect more information, and begin to discover the causes of variation in a process, you then become better equiped to adress and reduce the overall variation.
jmb:
Let me add a link in regard to #9 that will allow everyone to see the White House budget’s own chart for themselves:
http://blog.american.com/2012/02/this-is-the-debt-chart-obama-and-geithner-should-be-ashamed-of/
One looks at that chart and concludes that the President either intends to never do anything about the debt or that he is attempting to deceive the voters (including his base?) about what he plans to do should he actually be reelected.
And, by the way, the Washington Redskins recovery is working, too. After all, they beat the Giants twice in the last season. (Please ignore their 3-11 record against everyone else.)
Cowboy:
Part of the difference between successful and failed leadership is precisely the knowing when the need to wait for better data is outweighed by the decisiveness of the moment. Or, as the old Monty Python joke goes: “And now — the moment you’ve all been waiting for… and there it goes!”
“Part of the difference between successful and failed leadership is precisely the knowing when the need to wait for better data is outweighed by the decisiveness of the moment.”
Agreed, though I would add, that effective leaders also understand what impulses are driving their own reasoning.
When we try and use a central controller it seems the impedance is so high that the feedback loop mine as well not even exist, or, if it does, it creates such oscillations that that it becomes uncontrollable. The fallacy we are looking at is that the goals of each individual vary so widely that the central controller truly has no control. So in order to have any feedback that is worth its electricity is to have a controller for each family so the feedback is almost instantaneous, but when we introduce the central controller the feedback corrupts the true controllers and the whole system goes berserk. So, scrap the whole system and come up with a new one that actually represents the reality we are trying to model.
People were calling the housing bubble in 2004 and even before. It was obvious, the only thing that was perplexing is, how on earth did it last so long? It’s that central controller screwing everything up. It’s a natural consequence for monopolistic bureaucracies to grow ever larger until they cannibalistically consume themselves, which doesn’t need to happen overnight, on the contrary, it usually takes a while.
Firetag, great story about the Daedelus project. And not just the economic analogy, but an interesting discussion of how science fiction is (gradually) becoming science fact…though w/o Jim Kirk bedding the green-skinned alien chick.
It’s EZ to pontificate about the evils of Government over-regulation, the bloated size of Government, and the inability of the President and his economic advisers to ‘drive’ the ‘bus’. Reality is that’s what has happened, it’s gone terribly wrong, and we now have to deal with what ‘icky-tasting medicine’ can be pushed down the throats of the American public. Judging by Ron Paul’s good-but-not-even-close-to-good-enough showing, they won’t gobble up Libertarianism. So that leaves who will be the best “manager”…and the Republican faithful won’t stomach the Mormon.
This is looking like 1996 all over again, only the outcome leaves far more at stake this time. Then I could just be like Homer Simpson and vote for Kodos…
Jon:
It’s hard to scrap the system in mid-flight, but people do need to think about their responsibility to the care of their families and neighbors if we do become Greece-like in the problems of debt, because in that scenario government can’t magically cure the problem.
Douglas:
Jean Luc Picard would never have bedded the green alien chick, and Riker never made Captain until he stopped such irresponsibility. So I guess there is still hope for progress.
Firetag: Jean-Luc is the captain you want when negotiation is needed (he got it out of his system taking on three Naussicans). James Tiberius Kirk is best for butt-kicking (though he must have a large file in the Federation office of Child Support). Anakin Skywalker would kick the crap out of them both (and give Riker a wedgie just for effect) before his finished his morning ‘caf’ and go home and make whoopie with Padme.
FireTag,
Yes, it is hard to do mid flight, but unfortunately the controller is so out of whack that we don’t have much of a choice in the matter. A good engineer knows when to cut his losses and start over when the solution is found to be wrong. It may hurt and feel like a waste, but, in the end, it will do the company much better in the long run.
“So, scrap the whole system and come up with a new one that actually represents the reality we are trying to model.”
Yet another knee-jerk reaction based on the impulse to act quickly. It’s not that the model necessarilly needs to be scrapped. Particularly if we can’t say exactly what’s wrong with it. In fact, as I understand it, Firetag is debating the model at all. Rather he is arguing about the quality of the data we feed into the models to yield an output.
Cowboy:
My post IS about quality and limitations of the data, and one point I do make is that the current data’s noise level overwhelms the signal at peaks and valleys of the business cycle, even though it might be good enough to give a clear signal when growth or recession is strongly underway.
That brings me back to the link in 14. The signal of a coming debt tsunami is clear enough to show through the uncertainties. Both sides of the left-right political spectrum, with obviously different economic biases, now concede it exists. Whether it hits a few years before 2030 or a few years afterward only affects which of us see our own elderly years threatened, and which of us see the blows fall on the hopes of our children or grandchildren instead. We’ll be at the “pick our poison” stage if we continue to postpone dealing with the issue for a fraction of the time we’ve already tried to ignore it.
Jon:
This is an Apollo 13 Houston-we-have-a-problem kind of failure. It’s going to take creative rigging to find something to get people to safety. We can’t just exit the system and walk home to redesign Apollo 14.
Cowboy,
My point being, that the quality of the signal doesn’t matter if the system can never properly use the signal to begin with. Looking at the the symptoms doesn’t solve the actual problem, the root cause needs to explored. I wouldn’t call it a knee-jerk reaction, I would call it looking at reality for what it is and moving on to better pastures. Like I said, a good engineer knows when to scrap the old and start all over again, even though it is painful to see all the hard work being “wasted.” Unfortunately, monopolistic organizations typically refuse to do this, they are not the good engineers we should be looking to to solve our problems.
FireTag,
A better analogy is, we are in deep space our ship is cracking and about to melt down over a long time, the pain will be horrendous, but we have to opportunity to let our children to have to suffer like we are by building them a new ship and us sinking with the old. It seems like it would be immoral not to sink our own ship so our children can have a much better ship that would be very difficult to sink at all. The signals are clear that this needs to be done, unless we just want our children to suffer the same fate we are.
Also, continuing with the above analogy, when we do get the ship going for our kids, it won’t take too long for the kids and us to build a new one for ourselves also, are suffering need not be drawn out. Which, I guess, contradicts what I wrote above, but I don’t feel like rewriting. 🙂
Jon:
I can sympathize with your “scrap it” perspective. I think for the sake of the other “kiddos” on Wheatandtares, you and I should scrap our prevailing model of conversation. If prior experience is any kind of an indicator, I have a pretty good prediction where this conversation will go. Good luck to you.
Hindsight as they say is 20/20 is certainly true in this case.
Gee, people disagree on how it happened, when it happened and how the remedies taken did or did not work? That’s a surprise!
If we actually listened to the most impartial experts and not the partisan hacks we hear all the time, we might actually avoid some of these problems.
However, we have 1) a highly manipulated economy, which inhibits fast action. 2) we have a political system that thinks of itself first and our economy and the Americian people last.
So, to point a particular finger at a certain person, who is an easy target is just laughable. They are all to blame.
JMB #12 “There were people who predicted with pretty good accuracy what was going to happen. To then say “oh, well, they didn’t really know” is ridiculous.”
I just think you’re playing Monday Morning Quarterback here, which I guess is still Firetag’s point…we are steering by the rear view mirror validating one group’s predictions (among many predictions), and you fairly mention those groups that predicted with accuracy in the specific example we are discussing were not always right in other instances.
I understand the concept of not over-reacting, or not introducing more variance to the system by mucking with things, but I think we are more in agreement to do nothing is not a permanent solution, but more what Firetag said, knowing when to do something and when not to. Which brings us back to forecasting with probabilities, and without certainty.
“I understand the concept of not over-reacting, or not introducing more variance to the system by mucking with things, but I think we are more in agreement to do nothing is not a permanent solution, but more what Firetag said, knowing when to do something and when not to. Which brings us back to forecasting with probabilities, and without certainty.”
I won’t try and speak for JMB275, but that is exactly what I am saying. I enjoyed Firetag’s post and agree overall with his perspective. The only reason I see to interject a point is because phrasing it like this:
“knowing when to do something and when not to.”
without really explaining how a person knows when to do something, can sound a lot like trying to calculate “charismatic leadership”. It sounds a little sexy, but at the end of the day we don’t even know what that is really, so it just becomes a catch phrase and loose comment that leads us down meaningless paths. When you follow up your statement with this:
“Which brings us back to forecasting with probabilities, and without certainty.”
Then “knowing when to do something” starts to mean something.
Re Heber
I like what Cowboy said. But also, at least for me, we should default to not doing something if we don’t have an appropriate level of confidence in either the data, or our ability to control the system. So my primary complaint, overall, with our approach to the economy is that we act without high enough levels of confidence in both the data we have, or how our actions will affect the system. But our overwhelming psychological and political urge to do SOMETHING motivates us to ignore our low levels of confidence.
I agree, we’re all playing armchair quarterback here, no question. But then again, so is Chicken Little. Unfortunately, Chicken Little is right sometimes. That’s why I’m not too quick to dismiss all the armchair quarterbacks until we have better information.
jmb275,
I don’t think we are necessarily playing armchair quarterback’s since the actions of others directly affect us and we are the ones that are supposed to be the only quarterbacks of our own game but someone else is doing it for us, when most of us didn’t even ask or give permission to those playing it for us.
Jeff,
There is no such thing as “impartial experts.” That’s like saying, “If we actually listened to the leprechaun voices in our heads and not the partisan hacks we hear all the time, we might actually avoid some of these problems.” Everyone has a stake and therefore their views will be skewed to whatever benefits themselves the most. We could have the best data in world and it wouldn’t matter, people still want to serve themselves rather than others.
Jon:
“Everyone has a stake and therefore their views will be skewed to whatever benefits themselves the most. We could have the best data in world and it wouldn’t matter, people still want to serve themselves rather than others.”
That’s why we create general rules about data use outside of contexts where we have strong emotional biases so that we can hold to a “rod of iron”, so to speak, when our emotional biases ARE buffeting us.
Indeed, I think your observation about people acting in their perceived self-interest is a pretty high-confidence data point in itself. So, I think we each should individually be “realistic” about that when we decide how to get society to do something, or in devising our own plan of action if we can’t get society to do something WE perceive would be in their interest.