hello.
20 stories
·
0 followers

Stop aggregating away the signal in your data

1 Share

stackoverflow.blog - By aggregating our data in an effort to simplify it, we lose the signal and the context we need to make sense of what we’re seeing.


Tweeted by @EdwardTufte https://twitter.com/EdwardTufte/status/1525824511630839817

Read the whole story
jthaman
3 days ago
reply
Maryland
Share this story
Delete

"causal inference is all about the aggregation of individual effects into average effects, and if you have a direct model for individual effects, then you just fit it directly." twitter.com/StatModeling/s…

1 Share

"causal inference is all about the aggregation of individual effects into average effects, and if you have a direct model for individual effects, then you just fit it directly." twitter.com/StatModeling/s…

“Causal” is like “error term”: it’s what we say when we’re not trying to model the process. statmodeling.stat.columbia.edu/2022/05/14/cau…




44 likes, 13 retweets



36 likes, 4 retweets
Read the whole story
jthaman
10 days ago
reply
Maryland
Share this story
Delete

The Heel Heuristic

1 Share

You discover that your favorite movie star is a full-blown heel. He cheated on his spouse, got caught, and ended up divorced. Now, he’s estranged from his kids.

Question: How do you respond? Do you boycott his future movies? Stop watching his earlier movies? Treat him as an unperson?

Or do you shrug in disappointment, muse “Well, it’s really none of my business” - and continue to enjoy the star’s work?

These days, anyway, almost everyone takes the latter course. So says common sense, and so says this Twitter poll.

Is this because being a full-blown heel is “No big deal”? Hardly. Almost everyone agrees that the star’s behavior is quite blameworthy. They just don’t think that it compels any action on their part. So says common sense, and so says this Twitter poll:

Now consider each the following other actions.

  1. Exposing yourself to acquaintances in your hotel room.

  2. Using the N-word during standup comedy.

  3. Affirming the biological reality of gender.

  4. Using a racial epithet on camera when you didn’t know you were being filmed.

  5. Posting a sexist tweet.

Question: Are any of these actions remotely as bad as being a full-blown heel - a man who cheats on his wife, gets divorced, and stops seeing his kids?

Honestly, I doubt almost anyone sincerely thinks so.

Which brings us to a great puzzle: Virtually no celebrity gets “cancelled” for being a full-blown heel. In stark contrast, notable celebrities have been cancelled for each of the offenses on my list. What’s the logic?

The most plausible response is roughly: “Being a bad father has roughly the same bad effect if you’re famous or obscure. But being racist or sexist has a vastly worse effect if you’re famous than if you’re obscure, because you’re a role model.”

Yet on reflection, the most plausible response remains weak. Why can’t a celebrity father be a role model for other fathers? And if you question the size of this role model effect for fatherhood, why not question the size of the role model effect for racism and sexism as well? How many people actually look to Louie C.K. to decide how to treat women, or Michael Richards to calibrate their racial sensitivity?

The real story, I warrant, is just vividness and herding - what I’ve previously called “the Unbearable Arbitrariness of Deploring”:

I can explain what I cannot accept.  When I witness the unbearable arbitrariness of deploring, two unsympathetic types of explanations come to mind. 

First, people’s negative emotions depend far more on the vividness of the evil than its badness.  A hundred stories about celebrity harassers would upset the world far more than ironclad statistical proof that 10% of celebrities harass.  Indeed, it’s likely that one detail-rich story about a celebrity harasser would upset the world more than the best statistical study ever performed.

Second, people’s negative emotions are intensely social.  People don’t want to rage alone.  They want to get mad with their friends and countrymen.  So when a new round of ugly stories pop up, almost no one asks, “Is this really the best target of our collective anger?”  Instead, they jump on the bandwagon.  Who cares where we’re going, as long as we’re united in negativity?

What is to be done? I propose that we strive to defuse the power of vividness and herding with what I call the Heel Heuristic. Here’s how it works: Whenever a scandal erupts, ask yourself: Is this really worse than cheating on your wife, getting divorced, and no longer seeing your kids? If the answer is no, tell yourself, “This is really none of my business” - and forget about it. Even though the story is gripping. Even though other people hunger to discuss it.

If others keep pushing you to care about issues that don’t deserve your attention, respond by sharing the Heel Heuristic with them. Ask others: “Is this really worse than cheating on your wife, getting divorced, and no longer seeing your kids?” and “If you can overlook being a heel, why not overlook the latest scandal as well?”

Don’t get angry. Instead, encourage others to share your equanimity.



Read the whole story
jthaman
25 days ago
reply
Maryland
Share this story
Delete

100% of excessive inflation is due to bad monetary policy

1 Comment and 2 Shares

Always.

This shouldn’t even be debatable. I don’t know of any respectable theory (monetarist, Keynesian, Austrian, etc.) where this is not true.

And yet according to this recent Vox article, economists have no idea what is causing the current bout of excessive inflation:

Among economists and experts, there’s no strict consensus about what exactly is to blame. There are certain factors widely agreed upon that we’ve been hearing about for months: supply chain woes, rising oil prices, shifting consumer demands. These concerns have hardly subsided. But there are other arenas where there’s more disagreement, such as the role government stimulus has played in increasing prices, and the possibility that corporate greed is an important factor. . . .

Whichever economist or expert or policymaker you ask to explain the current inflation story to you is going to tell you something slightly different. I asked a bunch of economists over the past couple of weeks what was causing inflation and how to fix it. Most kind of laughed for a second before launching into their cases, acknowledging the full answer is, to a certain extent, ¯\_(ツ)_/¯.

The Fed’s job is to insure an appropriate level of demand. On occasion, it may be appropriate for inflation to exceed 2% for a period of time due to supply issues. But that’s no excuse for excessive growth in nominal spending. When inflation is higher than it should be for demand side reasons, it is always 100% due to bad monetary policy.

And now Vox tells us that economists are now talking about “corporate greed”? Seriously?

And just who are these so-called “economists”? According to Vox they are “progressives”:

Many progressive economists and politicians are beginning to argue that it’s corporate consolidation that’s making inflation worse.

Wait, I thought it was right-wingers who were the creationists, global warming deniers, the anti-vaxxers? I thought the left believed we should “trust the science”? Since when did progressive economists become the left wing version of QAnon?

Ezra Klein has a must read interview of Larry Summers. One of Ezra’s questions is interesting. No, “interesting” isn’t quite right, it’s the single most revealing question I’ve ever seen from a reporter in my entire life.

EZRA KLEIN: So I know you’re a hard-nosed economist who looks at the numbers here. But I want to locate, I think, the emotional and to some degree even political frustration of this conversation, because a lot of the dynamics you’re talking about that then get framed as excess demand, there are things that feel just, that many of us have wanted for a long time. More hiring, wage increases, particularly at the bottom end, stimulus checks for people who have had a lot of bad years and didn’t have a lot of cushion behind them, child tax credit for families that could really use that.

And so there are a lot of policies that came together — I mean, there was a reason the Biden administration wanted to run the economy hot. There was a long period when it didn’t just feel, the economic data showed, that expansions were not reaching people on the margins. And it felt, finally, like we were reaching people on the margins. We were putting a lot of firepower to do that. But even in this terrible time, this horrifying pandemic, we were giving people who needed it quite a bit of help.

And then for that to then turn into this horrifying inflation problem, which is now eating back those wage increases, potentially going to require much sharper action from the Fed— I recognize the world doesn’t have to please me, but it is maddening. And I think one of the hard questions, before we even get into Ukraine and China— I think one of the hard questions is, does it have to be this way? Did it have to be this way? Is there some way for this to end without the people we were finally helping suffering?

If anyone wants to know how we got in this mess, it’s right there in Klein’s question. Kudos to Ezra Klein for being willing to reconsider his views when new information comes in. But the passionate desire to “run the economy hot” in a misguided belief it would help workers is precisely how we got into this mess. Jay Powell and all the other the run the economy hot people wanted it to be true that the 1960s never happened. (Recall how Powell cited 1965 as a successful soft landing!) But the 1960s did happen, and could happen again if the Fed doesn’t wake up.

Klein asks Summers about the nonsensical claims that inflation is caused by corporate greed, and Summers has this to say:

So I think it’s, frankly, ridiculous to take businesses saying on a earnings call that they have pricing power as some kind of evidence of perfidy. I just don’t think that is supported by any serious understanding of how the business process works.

And by the way, another way of looking at this is, again, to look at what’s happening to wages. Wage inflation is as pronounced a phenomenon as price inflation. And we don’t think workers have huge pricing power. Or another way to look at this is as relations in terms of what Amazon is paying its suppliers or what Walmart is paying its suppliers. Surely if there’s market power in the relationship between Walmart and its suppliers or between Amazon and its suppliers, the power is on the side of Amazon and Walmart. And they’re paying much higher prices to their suppliers.

So are there market power problems in the American economy? Yes. Is it a good idea to attack those problems? Yes. Is it fine if we use the motivation provided by inflation to do some of that? Yes. Does it make any sense at all to blame inflation on market power? No. That’s not serious economic reasoning and the judgments and forecasts of those who engage in that reasoning should be taken less seriously as a consequence.

The interview is quite long and worth reading in its entirety. Summers makes many of the points I’ve been making, and he recognized the inflation problem well before I did.

At one point the discussion turns to what the Biden administration could do to slow inflation. Summers points out that they could do a few things at the margin, but their actual policy has been almost the exact opposite, to reduce aggregate supply and make the problem worse:

LARRY SUMMERS: Mostly the tools are pretty limited. And the tools that there are, are tools that the Biden administration has so far been very reluctant to adopt. If we reduce tariffs, that would make more goods available at lower prices and perhaps reduce the Consumer Price Index by 1 percent or more. But their rhetoric has gone the other way on tariffs.

If we decided to do public procurement as inexpensively as possible, that would reduce prices of a whole set of things the government buys and increase competitive pressure. But we’ve instead indicated a desire to shift from buying cheap to buying America and buying in ways that protect certain key constituencies.

These policies would not reduce aggregate demand, but by boosting aggregate supply they would make the “appropriate” inflation overshoot of 2% smaller than otherwise (from a dual mandate perspective.)

If we could just stop talking about inflation and focus on NGDP growth then all of these concepts would be so much easier to explain. The language of macroeconomics is such a complete mess.

PS. Here’s what else Biden could do. He could “promote” Powell to Treasury secretary and replace him with Yellen or Summers. In 1979, G. William Miller was promoted to Treasury secretary, and replaced with Volcker.

Read the whole story
jthaman
53 days ago
reply
Maryland
Share this story
Delete
1 public comment
freeAgent
54 days ago
reply
Like Sumner, I'm curious to know who Vox's "economists" are.
Los Angeles, CA

What caused the high inflation?

1 Comment and 2 Shares

In a recent interview, SF Fed President Mary Daly listed 4 factors that caused inflation to exceed her expectations. The first three are supply issues, while the fourth relates to demand:

4. Unexpectedly high consumer demand

The final factor that Daly says she underestimated was consumer demand. “The American consumer has been incredibly resilient and incredibly interested in purchasing things when they couldn’t purchase services,” she said. At some point, she believed that Americans had “purchased as many Pelotons as we can possibly use.” And yet, the demand seems insatiable.

In February, overall retail sales increased 0.3% from January and were up 17.6% year-over-year, according to U.S. Census Bureau. And that’s set to continue. The National Retail Federation predicts that sales will grow between 6% and 8% this year.

I have several problems with this claim.  First, it’s pretty obvious that most people have effectively “insatiable” preferences for a higher living standard.  Even if at some point people have all the Pelotons they want (and I for one do not), they would simply begin to desire other goods.  I find it a bit worrisome that a top Fed official would view consumer satiation as a reason not to worry too much about inflation.

Second, it makes more sense to focus on total aggregate demand rather than just consumer demand.  In some cases, excessive aggregate demand shows up in rapid growth in investment spending, which can be just as inflationary as rapid growth in consumption.

Third, there is no mention of the role of monetary policy in creating the inflation.  Fed policy was clearly too expansionary last year, and as a result aggregate demand (M*V) rose at an excessive rate.  Fast growth in nominal spending will lead to high inflation regardless of whether consumers have enough Pelotons or not.  If the consumer saving rate rises because their garages are packed with expensive toys, then fast growth in nominal spending would lead to higher investment spending.  Or perhaps government spending increases.  One way or another, a monetary policy that leads to excessive growth in nominal spending is almost certain to lead to excessive inflation. 

When I hear Fed officials talk about inflation, it often seems as if they regard it as some sort of mysterious problem that befell our economy.  Excessive inflation is a product of excessively expansionary monetary policy.  Demand is a nominal concept; don’t talk about it like it’s a real concept.  Aggregate demand rose by more than 100 billion-fold in Germany during the early 1920s, and it wasn’t because Germans suddenly had an insatiable demand for exercise equipment.

That does not mean that all inflation above 2% is excessive.  The Fed has a flexible average inflation target, and when there are supply shocks it is appropriate to allow above 2% inflation for a brief period in order to better achieve the Fed’s dual mandate.  But when inflation is excessive even from a dual mandate perspective (as it clearly is today), that’s a failure of monetary policy.  It’s that simple.  Fed officials are perfectly justified in talking about supply problems, which do provide justification for temporarily allowing above 2% inflation.  But instead of talking about mysterious increases in “demand”, I wish they’d simply say that monetary policy in 2021 was too expansionary.   Why is that so hard to do?

Arsonists don’t need to fix the house burning problems; they need to stop burning down houses.  The Fed doesn’t need to “fix” the inflation problem; it needs to stop creating inflation.

(1 COMMENTS)
Read the whole story
jthaman
56 days ago
reply
Maryland
Share this story
Delete
1 public comment
freeAgent
56 days ago
reply
Never admit to making a mistake, Fed. You do you.
Los Angeles, CA

Inflation is a disease – Milton Friedman

1 Share

Article URL: https://www.youtube.com/watch?v=B_nGEj8wIP0

Comments URL: https://news.ycombinator.com/item?id=30743725

Points: 81

# Comments: 99

Read the whole story
jthaman
59 days ago
reply
Maryland
Share this story
Delete
Next Page of Stories