I sincerely hope Tufte can successfully counteract the Washington tradition of applying the methods from How to Lie with Statistics. A good visualization can just as well expose the fraud from a contrived one.
> From what I've read of his work, Tufte has almost a compulsion to force data to tell the truth
I think it's naive to think that there's just "one truth".
What's the truth about NASA spending during the Apollo program?
It was so-many billion that added up to a glorious step into the cosmos?
It was so-many billion that added up to a race with the Soviets to prove that freedom is better than communism?
It was so-many billion that was a big whose-dick-is-bigger game?
It was so-many billion spent to bid up the wages of well payed first world engineers while third world peasants continued to live in poverty?
These are all true, and there's no way that graphic design can pick the one true truth out of the other truths.
Tufte can do some great work on displaying the multi-trillion dollar buget, but it's not going to get at "the" truth.
Here's the visualization I'd like to see:
FDR's Depression budgets weren't concentrated in areas with the highest poverty, but in areas where his political support was wavering. Obama's "stimulus" is similarly spent in areas of the country, and areas of the economy where buying political support is needed. I'd like to see Tufte do THAT visualization.
I'm sorry to hear that he's going to be on the government payroll - I'd love to have his skills deployed as a non-partisan, non-aligned outsider, as opposed to someone inside the tent looking out.
By a similar token, it can also be naive to expect the truth to be plain at the outset, or even for a sizable period of time afterwards. This is doubly (trebly) so for political matters (c.f. Gulf of Tonkin). What's sought is honesty.
To quote an old Murakami novel: "Speaking frankly and speaking the truth are two different things entirely. Honesty is to truth as prow is to stern. Honesty appears first and truth appears last. The interval between varies in direct proportion to the size of the ship. With anything of size, truth takes a long time in coming. Sometimes it only manifests itself posthumously."
Obama's "stimulus" is similarly spent in areas of the country, and areas of the economy where buying political support is needed. I'd like to see Tufte do THAT visualization.
Interesting. I think this is exactly the kind of claim that when it is posted on any forum on the internet should always be backed up by a reference to clear evidence.
I'm not saying I disbelieve you, but simply that if it's as clear cut as you make out it should be straighforward to provide the evidence.
Yes. What if the data plainly presented show that the stimulus plan was plainly a waste of money?
P.S. How many of you, when you hear announcements of how many jobs have been saved by the stimulus plan, immediately do the mental arithmetic to figure out how many dollars of taxpayer funds were spent to save each job? I think I could do better in my family in keeping us all actively employed in work helpful to society with the same number of dollars per employable person.
Before you do that math, make sure you're not counting raises, theoretical jobs from as-yet-un-funded projects, etc. Here's a nice map of where all the fake saved-jobs are:
Although you have to adjust upward: faking the existence of tens of thousands of jobs is obviously difficult work, and I bet more than one person is working full-time on the faking.
Tax credits, Pell grants, unemployment support: it's not all jobs.
But you have to blame the administration for this [perception], at least in part: They sold the stimulus in terms of job creation, rather than simple economic growth.
How many of you, when you hear announcements of how many jobs have been saved by the stimulus plan, immediately do the mental arithmetic to figure out how many dollars of taxpayer funds were spent to save each job?
I do not, because that is a fool's calculation.
The value created is not one year's salary, it's the difference in the present value of an individual without a job and the present value of an individual with a job, including all estimable secondary effects (both positive and negative.) If you simply divide it and compare it to a year's salary, you'll sit there and go 'HURRR THAT IS DUMB', but in reality, you're proving only that you don't really understand policy analysis.
----
I'm reminded of a situation that occurred many years ago in my career, when I was still working in a corporate setting. My team hit a seemingly intractable combinatorial explosion problem while trying to create a tool for the sales force. It was incredibly difficult, and quickly blew through the allocated budget (which we'd told them right from the gate, was inadequate).
The management just assumed we were lying... that we weren't up to this task, or we were just lobbying for more money and man-power for our group (I was trying to make a very specific high-end hire to help, and I had already purchased some serious iron). In fact, they were so convinced that this couldn't be this difficult to accurately analyze that they started shopping consultancies, and most agreed with us, that this would be difficult and expensive... but one did not. One said "yes, we can build it, no problem."
A year later, the consultancy had burned through a pile of money, and had nothing to show for it.
Only then did management recognize that maybe this seemingly simple configuration problem was not, in fact, as simple as they imagined. And perhaps we weren't just lying or waving our hands when we said "this shit is not as simple as you think it is." (and even then, they didn't admit it outright. I had an all-day meeting which consisted solely of them asking me stupid question after stupid question, as to why various simple approaches wouldn't work... and me having to draw out specific examples on a whiteboard, indicating the business and technical problems with these simple approaches.)
It's sad, but common, for people to assume that other people's jobs are easier than they really are. Even very smart people fall into that trap... but generally speaking if intelligent, hard-working people are advocating something that seems stupid to you, it's probably because you don't understand the problem or the solution.
Thank you for the detailed reply, not all of which was responsive to my reply. My claim is still that if comparable amounts of money were put in my own pocket, I could make sure that my family members who are of employment age would be going about producing more externalities beneficial to all society than many of the workers whose jobs have supposedly been saved by the current stimulus package. Reasonable minds can differ on how to count the present value of this worker being unemployed today and for some reasonably expected time thereafter, versus that worker being employed today, but that inquiry, to be worthy of relying on for endorsing a public policy, has to be done carefully and with great attention to detail. One first detail to look at is the actual dollar amounts involved.
Is it really foolish to ask if money is being well spent?
> My claim is still that if comparable amounts of money were put in my own pocket, I could make sure that my family members who are of employment age would be going about producing more externalities beneficial to all society than many of the workers whose jobs have supposedly been saved by the current stimulus package.
You might be able to. But I can probably find 4 families on my own block that would just spend the money on tiger print shag carpeting for their bedrooms.
Is it really foolish to ask if money is being well spent?
Only if you ask the question as poorly as you did. Taking two arbitrary values out of a complex multi-variate equation, and dividing them generates nothing but noise.
If you're legitimately interested in quantitative public policy analysis, there are excellent graduate and doctoral programs that deal both in general policy analysis, as well as more specialized studies dealing specifically with labor and unemployment.
If that sounds ridiculous to you, perhaps it'd be wise to recognize that the analysis is highly skilled technical work, and if you haven't devoted considerable energy towards gaining those abilities, you're just a noisemaker. You might as well be a donut-baker who holds strong opinions about graph theory despite having never programmed nor studied CS.
* My claim is still that if comparable amounts of money were put in my own pocket,*
Even if your claim is true, governmental solutions need not only to be good, they need to be replicable and scalable.
If step one is "identify a million individual families who are able to use money more efficiently than established market-proven businesses", then it's going to take a very, very long time to reach step two.
> The value created is not one year's salary, it's the difference in the present value of an individual without a job and the present value of an individual with a job, including all estimable secondary effects (both positive and negative.) If you simply divide it and compare it to a year's salary, you'll sit there and go 'HURRR THAT IS DUMB', but in reality, you're proving only that you don't really understand policy analysis.
You're assuming that the job will last beyond the year and that we couldn't simply send that person $X dollars and produce most of the same benefits.
If you think it's controversial among fiscal conservatives to do job promoting stimulus programs, just wait until you suggest just sending poor people full-time salaries worth of extra cash for nothing. Good luck with that.
Your comment, particularly the part where you talk about "simply send that person $X and produce most of the same benefits" underscores your complete and utter lack of comprehension of his post.
Ok, although IANASS (not a social scientist) (edited to add: and sorry anamax for the uncivil tone, long day):
Economic activity has benefits for society at large -- we get stuff out of it, and there's a feedback loop where more people with money in their pocket means more demand, which means more people employed, which means money in their pocket, etc.
Recessions are when the economy gets out of balance (historically inventory oversupply, more recently it tends to be financial meltdowns) and that feedback loop goes negative -- less money in pocket, so less spending, so less production.
Counter-cyclical spending is, according to economic theory, the one point in the cycle where the government should be encouraged to spend money like they're in vegas, up 50k and hanging at the strip club. Breaking this negative feedback loop is priority #1 -- debt can be paid back later, and if you've been breaking out your counter-cyclical toolkit, you've already lowered interest rates to near-zero so this is really cheap debt, 1% interest or lower -- if you produce any return at all, you've got a win in addition to the counter-cyclical effects.
Once you've decided to spend, you've got to decide how. Giving money to people in lump sums (which was in fact about 40% of the bill, the bill was 40% tax cuts) is fast and simple, so it certainly has a place, but keeping people in a job is far more effective because you create externalities. That job is producing something of value to the economy, that person is gaining skills/experience leading to future employability, and assuming that business is ok, their labor might lead to additional job openings at the same company later. These are all sustainable improvements that build on themselves and contribute back towards that positive feedback loop. Just spewing money out to people without any targeting whatsoever can only be a brake on recession -- it doesn't really help lay the foundations for the bounceback.
1. The government can increase spending, but they can't create money. Either they get it from taxes, or they get it from devaluing the currency (i.e. taxing holders of currency, or currency-denominated assets). How would it help to take $1 away from an employer building widgets, and use it to hire a Diversity Coordinator at the DMV?
2. What's the long-term effect on business decisions if the only two modes the economy is in are a) growth, or b) near-zero interest rates? It looks like this would lead to lots of leverage.
3. You're assuming that it's always better for someone to have a job than to lose it. So you'd end up subsidizing candlestick makers (probably by taxing light bulb factory owners).
At best, this philosophy merely gives everyone a reason to misbehave. At worst, it fines the productive, growing parts of the economy, in order to reward the unproductive, shrinking parts of the economy. America in the 1930's. Japan in the 1990's. How many lost decades will it take to discredit this stuff?
WRT to #1, before you get to the "or" items, government borrowing crowds out private, since a national government (or at least ours) is viewed as a much safer creditor in the earlier periods of economic trouble.
In the longer term, sovereign debt defaults, explicit or implicit (your "or" items, including the inflation tax), inevitably follow this type of counter-cyclical spending (statistically, since this tends to be a regional or world-wide thing).
Government debt is an entirely different product than private debt. Different interest rate and all of that.
Regarding the sovereign debt defaults, did we have one of those after the great depression? Sovereign debt defaults are associated with financial panics. Don't bind up intermediate (correct or incorrect) steps with the root cause.
It may be a different product but it's still in the same market for people who want to lend out their money. The different interest rates are because of different perceived levels of risk; normally a government can offer less because they're viewed as more likely to return the money.
If they aren't in the market or to the degree they are less so, investors have no choice but to put their money elsewhere.
As for your question, I don't think we can easily say. E.g. the BLS says the dollar lost 28% of its (CPI) value from 1940 to 1946 ... but I'm not willing to assign blame due to the discontinuity of WWII.
If Japan defaults in its 3rd Lost Decade, will that be the cause of the initial '80s panic or decades of the wrong medicine to deal with it?
IMO decades of the wrong medicine in Japan's case. In the case of an Argentina where a ton of crap hits the fan and they default a year later, it's a lot harder to blame the medicine.
(regarding Japan -- the consensus is that they didn't do enough stimulus, and that in addition to their high savings rate is why they fell into a deflationary spiral. the government printing money and handing it out with coupons for blow and hookers would have been a better strategy than what they did)
1) They can indeed create money : wiki "Quantitative Easing" (edit: and reduce it, too! that's the federal reserve's fundamental job)
2) Stagflation is much worse than recessions and bounceback with low interest rates. See: the 1970s. There have been a lot of recessions and a lot of livelihoods ruined over the last 100 years or so to reach our current understanding -- it's likely not perfect, but it is based on experience.
3) That's not what I'm advocating in the slightest, I advocated for targetted government spending to help create/retain jobs, specifically on things that have high externalities for the country. We don't need any more candlesticks or buggy whips, but every state is laying off teachers like crazy and we could certainly use some new bridges and high speed rail.
Regarding your last statement -- a lack of sufficient stimulus up front is the most universally agreed upon reason for the protracted nature of BOTH of those downturns. Like, all of the literature says the exact opposite of what you're saying. Maybe they're all wrong, I dunno. I do know that in both cases, decision makers were motivated by feelings of avoiding borrowing as a fundamental value, which is the right approach all of the time except for specifically during a panic or the beginning of a recession.
1) They can print dollars, reducing the value of existing dollars.
2) The government was growing massively in the 50's, 60's, and 70's. If government growth was the solution, why did we get so much of it before the problem?
3) Is the government more likely to subsidize new, growing industries, or old, stagnant ones? Government subsidies will "save" the most jobs when they target industries with low capital requirements (like the ones that rely on cheap labor) rather than ones with high capital requirements (like high-end manufacturing). Venture capital is a tough job; deciding that some guy in DC is going to do an even adequate job at it is a stretch.
It is not "universally agreed." I think we can all agree that both of the cases I cited included an unprecedented level of government spending, coupled with an unprecedented recession.
"I was healthy, then I sneezed and you demanded that I start using leeches. I've never been sicker."
"It is universally agreed that the problem is insufficient leeches."
You need to quit trying to boil things down to conservative ideology talking points -- the size of government is largely irrelevant to the choices faced in an economic downturn. The only people who care are right-wingers who try to drag it into everything.
1) Oversimplification, read an intro to economics book about how the chain of lending actually creates more dollars and wealth without just printing them
2) Not addressing what I was saying. The size of government is irrelevant here, we're talking about stimulus spending -- totally orthogonal.
3) Not addressing what I was saying in the slightest.
It is in fact almost universally agreed that in both of those cases, the government spending was too little too late. By the vast majority of economist who've spent their lives studying this.
Maybe your "gut feeling" is smarter than their decades of research. But I'd suggest you read this: http://www.xkcd.com/675/
EDIT: I'll add a history lesson so we can all learn something today.
1929 - stock market crashes
29-32 - hoover does nothing, economy spirals downwards
32 - FDR elected, massive counter-cyclical public works projects, economy recovers
37 - FDR raises taxes and cuts back on spending as an austerity measure, too early, double-dip
In the case of Japan, similar story except the high savings rate exacerbated it, leading to deflation -- the government printing money and handing it out on a street corner would have been more effective than what they did
I presume that views are either "ideology talking points" or "universally agreed-upon facts," right? Case closed!
I have read intro econ textbooks. The notion that taking a dollar from one party and giving it to another party creates wealth is absurd. At best, you can argue that it will debase the currency (if not, why doesn't it work with any commodity? You can say that if we confiscate X amount of wealth, and give it away, it turns into 2X amount of wealth. But if the wealth is denominated in, e.g., gold, or houses, or publishing rights, it doesn't have similar fecundity.)
Thank you for the comic. It ably makes your point. However, we're talking about a social science, which is at least partially influenced by the need for grants, public approval, speaking engagements, etc. You can't just put the Depression in a supercolider and run tests on it. Because of this, it's more analogous to other ideological popularity contests, like English Lit and Scientology. People have spent decades studying both of those, but it hasn't made them great writers or superhumans. Similarly, economists who study lots of Keynesian economics don't end up making awesomely profitable global macro trades--even Keynes was smart enough to invest based on something other than his own philosophy.
Hoover's spending was unprecedented. It was hardly "nothing". Another timeline to consider:
Dawn of time-Present: Governments attempt to control currency, typically lead to inflationary bubbles, followed by busts.
> 2) Not addressing what I was saying. The size of government is irrelevant here, we're talking about stimulus spending -- totally orthogonal.
Actually, it's not orthogonal. In fact, it's quite on point, as stimulus spending tends to become permanent spending, that is, part of govt.
We don't have counter-cyclical spending in the US. We have huge increases during down times and smaller increases during up times. Occasionally the private growth during up times exceeds govt growth. However, it's basically a one-way ratchet.
Sigh, well if you're determined to bring silly ideology into everything, of course you can find a way to relate it. But it's entirely beside the point, unless we take your assumptions as gospel (they're not). Adding "but I think this will happen" when it is in no way mandated or apparent doesn't change the facts.
We do in fact have counter-cyclical spending. You're seeing it now. Obama already announced a spending freeze on domestic programs, so we'll be seeing federal domestic spending decrease in real terms over the next few years.
If you're wondering the last time that federal spending decreased in real terms? It's not "never". It's "the 1990s". (protip: inflation is relevant).
> But it's entirely beside the point, unless we take your assumptions as gospel (they're not).
I didn't "assume" that US govt spending grows monotonically in the past. I observed it. If you think that my observation is incorrect, let's see supporting data.
> Adding "but I think this will happen" when it is in no way mandated or apparent doesn't change the facts.
I didn't say that monotonic spending increases are mandated - I said that they happened. If you're going to claim that this time will be different, you get to explain why.
> We do in fact have counter-cyclical spending. You're seeing it now. Obama already announced a spending freeze on domestic programs
Actually, he didn't. He announced a freeze on some types of discretionary spending. That "freeze" excludes the vast majority of the budget and the places where the increases have been occurring.
Similar "freezes" have been announced in the past and didn't happen. Since this one isn't actually reducing spending (contrary to the "now" claim), it's somewhat silly to assume that this time will be any different until it actually is.
> If you're wondering the last time that federal spending decreased in real terms? It's not "never". It's "the 1990s". (protip: inflation is relevant).
I didn't say "never", I said almost never.
The 90s drop came from decreased military spending due to the end of the cold war. The military budget is now small enough that plausible future cuts are less than planned increases in other spending. In other words, that's a one-time event.
> Economic activity has benefits for society at large -- we get stuff out of it, and there's a feedback loop where more people with money in their pocket means more demand, which means more people employed, which means money in their pocket, etc.
Yes, there is. However, the amount of feedback, which economists call the "multiplier" is a matter of considerable dispute. Obama's current point-woman on this has published research showing that the multiplier for much govt activity is typically less than 1. (She's saying something different now.)
> Giving money to people in lump sums (which was in fact about 40% of the bill, the bill was 40% tax cuts) is fast and simple, so it certainly has a place, but keeping people in a job is far more effective because you create externalities.
You're assuming that the person is producing something of value and at a reasonable cost..
More to the point, the original poster asserted that a job which cost $200k was worth more than just giving the person $70k/year even if we ignore such factors. (Numbers made up.)
Note that Krugman recently wrote that just giving money to unemployed people doesn't cause problems. This is in contrast to his statement in "Macroeconomics" that unemployment payments encourage people to stay unemployed longer than they would otherwise.
> That job is producing something of value to the economy, that person is gaining skills/experience leading to future employability
I'd agree, but this flies in the face of the arguments used to justify higher minimum wages.
> Counter-cyclical spending is, according to economic theory, the one point in the cycle where the government should be encouraged to spend money like they're in vegas, up 50k and hanging at the strip club. Breaking this negative feedback loop is priority #1 -- debt can be paid back later,
I know the theory. However, it assumes that govt does something that the US govt doesn't do, namely save during the other part of the cycle. Given that, it's unclear that the spending during the down part makes sense.
To put things in terms of another HN argument, it's unhealthy to eat the calories that you'd need to train for the Ironman if you're not actually training for the Ironman.
Frankly, I don't think there's a need to complicate this discussion with mentions of Keynes or interest rates. Their error starts far earlier, at the basic measurement of value.
Joe works at WidgetCo and makes $60,000/yr. His loaded costs are a little over $100,00/yr.
WidgetCo pays this because his skills are worth about $125,000/yr to them, so it's a great deal.
At this point, if you are a well-meaning, but rational cherub, what is the highest rational price to pay to get Joe working again? I'm guessing anything up to $125,000.
----
But that's not the end of our grossly simplified model. We also have the fact that Joe spends his paycheck on niggles. He fucking loves the shit out of niggles. As he consumes $60,000/yr of niggles, he personally pays for 1/2 of an employee at the niggle factory (and that guy spends his money on something too...)
So now we have to recognize that if Joe loses his job, there's going to be a negative ripple effect, and if Joe gets a job, there's a positive ripple effect.
So now, as a well-meaning, but rational cherub, what's my highest rational price? Well.. at this point I'd probably call in an economist to help me figure it out, but it's clearly more than $125,000/yr... it's likely close to $200,000/yr.
----
And now somebody hands me a "stimulus" proposal, and it turns out that I could create a job for Joe for about $150,000/yr. So let's see... $200,000/yr of value for $150,000/yr... sounds like a great investment to me! Very good value for money, even if it does lead to an annoying headline indicating "$60,000/yr jobs cost silly cherub $150,000!"
----
Obviously this is grossly glossed over (I didn't even mention the obvious fact that if people are buying widgets, they must provide value), and actual models of costs and benefits of various scenarios are quite complex, but I simply meant to demonstrate that simply dividing the total cost of a package by the number of jobs is silly, and misses the point widely.
Isn't that an example of the "multiplier effect" ... which has never really been seen in the wild (or so I understand)?
E.g. because on the other side of the ledger the government crowds out private borrowing (see my comment in this thread to byrneseyeview) and/or taxes others to get this money, plus this almost always is accompanied by attacks on businesses of all sorts, on Main Street as well as Wall Street. Extreme examples include 1937, when capital in the US went on strike and we had the Recession of 1937–1938 and didn't really recover until WWII changed the game.
In serious times of trouble, can you point out examples where this has worked in more than the short term? It didn't for the US in the Great Depression, and Japan is now nearing the end of it's second lost decade (we through they were pulling out of it, but that turned out to be a side effect of the world wide bubble that popped a little later this decade).
The meat of my post was simply an attempt to show that the economic value of an individual is far greater than their salary, thus demonstrating that if one looks only at specific jobs and salaries, the analysis is not complete.
As for more specific discussion, I'm not motivated to spend weeks putting together semi-legitimate models of various stimulus efforts, and doing less than that would be pointless. That said, when you talking about crowding out effects and such, you should be aware that economists are aware of these effects, and they are part of the math that sits behind a stimulus model.
The only thing I'll address in passing about the roosevelt recession is that it emphatically did not occur during a period of government stimulus, rather it came at a time when the government was adopting austerity measures and increasing taxes, which is pretty much the opposite of the situation today.
That said, I stand by my original point that proper analysis is complex, and that there really isn't any substitute for it. Anything less is just noise... so let's leave it to the technocrats.
Analysis of this sort is indeed very complex and I did increase the scope of the thread beyond what you were addressing.
As for the "Roosevelt Recession", it's worth pointing out that sooner or later you have to tighten the screws (later might be when external to your country effects come into play, e.g. right now I think the biggest question is when will the Federal government run out of people willing to finance its 1.5 trillion dollar a year deficit (yeah, its supposed to go down to 1 trillion soon, but who believes that absent some major political shifts?)).
That late '30s experience suggests that counter-cyclical spending might only delay the day of reckoning, with perhaps the Japanese showing what happens when you mostly try to indefinitely delay it ... and they can finance their deficit spending for at least a little while longer (well, until too many retirees try to draw down their assets).
The japanese are generally regarded as not having done nearly enough in counter-cyclical spending over the last 2 decades, specifically at the beginning.
The problem in 37 is generally understood to have been "too much, too soon" in terms of austerity measures. Penny-wise and pound-foolish if your austerity measures cause the economy to collapse back into recession.
Hmmm, I should also point out that a lot of this has to do with whether you have a liquidity ("It's a Wonderful Life") or solvency problem. There are a lot of short term fixes you can justify for the former that for the latter work only if the surviving institutions can earn their way back to solvency ... until which they're zombies.
The Austrian School is dedicated to zombie killing ^_^.
This is unresponsive to the idea that private actors individually acting in the aggregate may make many more better externalities than a few "targeted" government programs funded by deficit spending.
I knew he was off the rails as soon as he indicated that my claims were based on an assumption of duration of employment. That statement only makes sense if you use the same model that the parent used, in which the only value captured is the worker's salary.
On the plus side, the Obama administration certainly knows this too. This suggests that they do actually think it wasn't a waste of money (otherwise, they'd get someone they could bribe to say it wasn't). Now, they could easily be wrong, but we might manage to get some idea.
Elizabeth Warren has been fairly vocal and, from what I can tell, honest in what has been going on with stimulus money. She also seems to have been fairly successful (or was last time I checked). I don't know anything about this guy but if he is like her I feel like we are in good hands.
