Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Jason Smith Profile picture
Jason Smith
@infotranecon
Physicist who messes around with economic theory. Author: A Random Physicist Takes on Economics + A Workers' History of the United States. He/Him.
Jun 28, 2023 • 32 tweets • 11 min read
Haven't done a twitter talk in awhile. Despite my commentary about economists freaking out about the term "greedflation", I don't think recent inflation was driven by greed any more than anything to do with money is driven by greed.

Gonna start with supply and demand! When supply and demand match up with no excess demand or excess supply we have a (Walrasian) equilibrium

This implies the distributions are directly related

Which implies the information content of those distributions is directly related
Jan 12, 2022 • 8 tweets • 3 min read
CPI data continues to be consistent with a 10-11 month width (FWHM) shock from Mar 2021 to Feb 2022. Monte carlo using the covariance of the estimated shock parameters (gray lines) shows that the bulk of uncertainty is towards the upside of the best fit in terms of duration.
Jan 10, 2022 • 8 tweets • 2 min read
"I mean what differences do Republicans and Democrats even have? There's no policy disagreements I'm aware of — just two groups that don't like each other."
politico.com/news/magazine/… tfw you savvy yourself into a gray void
Apr 17, 2021 • 21 tweets • 12 min read
1. I'm a week away from celebrating the 8th anniversary of the blog. When I started, it was a fading form having reached its pinnacle of achievement on May 30, 2012 with Cosma Shalizi's review of Red Plenty. 2. While substack might be leading to a resurgence of longer form (but partially paywalled) independent writing platform, the old school blog seems dead.

So during this "anniversary" week, I'm going to build a list of my favorite posts from my blog as a kind of Irish wake.
Apr 17, 2021 • 6 tweets • 3 min read
Still estimating ~ 590 thousand COVID deaths in the US. The latest surge is starting to show up in this data; hopefully it will be muted by vaccinations of vulnerable people. Image The beginnings of a statistical deviation with data biased high: Image
Apr 17, 2021 • 5 tweets • 2 min read
“Because this one price is rapidly rising all prices must be rapidly rising.”

Hard to find a hole in this logic! Lol. They repeated it on Marketplace the other day using lumber as the example. That NPR/Marketplace comment was what this was subtweeting.
Feb 12, 2021 • 18 tweets • 7 min read
I thought I'd do a short thread on why I think "Regression Discontinuity Design" (RDD) is a poor method, especially for noisy data in economics. I'm going to generate a "time series" where the only changes are pure noise. Here, the underlying variable x is constant in time (mean x = 1), but has normally distributed noise with a standard deviation of 1.
Nov 14, 2020 • 4 tweets • 2 min read
Let me start by saying I love the three pages of disclosure and research design information in Nature articles. However, the results do not seem to be qualitatively very good.

The information we'd want to be able to glean from this additional mobility data is the future path, and that is pretty hit or miss.
Nov 14, 2020 • 6 tweets • 4 min read
The latest surge in reported COVID-19 cases in the UK seem to be passing through the peak as indicated a month ago.

(The xmas holidays could start up another surge, however.) Tweets from Oct 12:
Nov 12, 2020 • 4 tweets • 2 min read
JOLTS hires rate data looking more like an ordinary recession — about 1/3 the size of the 2008 recession. Image This is consistent with the job openings rate which has looked like an ordinary recession all along (albeit one that started in 2019).

It's also showing it's about 1/3 the size of the 2008 recession. Image
Nov 12, 2020 • 8 tweets • 3 min read
The estimate of the path of cases in the UK has been fairly stable for the past month (the long dashed line is the estimate from Oct 12th, and the dotted line is the current estimate).

Using that baseline, we could see the size of the effect from the xmas holidays pretty well. Image The path of cases in the US has just started to stabilize (the rainbow lines show estimates over the past week) — but still hasn't started to turn over yet. We'll likely get another surge with the Thanksgiving and xmas holidays before the current peak shows up. Image
Oct 18, 2020 • 6 tweets • 3 min read
The seemingly early rise in deaths relative to cases is likely due to the "upper midwest super-spreader event" (late Aug/early Sep) — we haven't begun to see the rise in deaths from the case surge in the rest of the US yet. For example, Florida — cases rising, but no sign of a surge in deaths yet (it's due to come in the next couple weeks).
Oct 18, 2020 • 4 tweets • 1 min read
TFW you forget that true “Enlightenment” values would mean you disclose the aims of right wing donors to the Hoover institution so people can judge whether the source of information is biased. Love the misrepresentation in the screen capture as well — makes it look like they’re blocking a purely political image about the media but in fact it was the tweet below it that gives misinformation about masks.
Oct 16, 2020 • 5 tweets • 3 min read
I haven't looked that the performance of the 10-year interest rate model in quite awhile — it's been going for more than 5 years now. And while the recent data seems to be quite a bit outside the average path forecast (green) ... Image But the interesting part (to me at least) is that the forecast was based on a counterfactual path of the inputs — if we plug in the actual path of the inputs, it accounts for the latest drop fairly well! Image
Oct 16, 2020 • 4 tweets • 2 min read
Was busy with the real job this week, so I missed the inflation report that basically said inflation had returned to normal after the COVID-19 fluctuation. Image You can see it in the CPI level more clearly — a blip and then a return to equilibrium slope. Image
Aug 8, 2020 • 17 tweets • 6 min read
0. Inspired by several recent twitter threads and some DMs, I thought I'd lay out my personal (not terribly original) view of what is meant by "intelligence" — no one really knows! But it's why I think IQ is garbage, we need a social safety net, and 2+2=5 is important. Thread. 1. We're all born with a common, genetically initialized neural network. That initialization sets up a few (probably unrelated) "dimensions" in the abstract space of potential outputs.
May 25, 2020 • 9 tweets • 2 min read
If this was true, then Warren would have run away with the nomination — instead she won no primaries.

I think his mistake is a mash up of the *effect* of US policy over the past few decades and Murc’s law (“only Democrats have agency”). Any theory of US politics and the Democrats must start with the fact that since 2001 they only had power for 2 years (2009-2011) — during which they made the largest expansion of health care since the 60s.

Was it enough? Nope! But it’s a good pole star to align your theory.
Apr 16, 2020 • 5 tweets • 2 min read
It looks like the explicit shutdown order in WA state only accounted for about 1/3 of the decline in open table reservations in Seattle — most of the decline was "voluntary" and more likely triggered by news of deaths. Similar behavior in NYC — about 10% drop around the time of the first death in WA state, but a subsequent voluntary drop around Cuomo's state of emergency declaration with the shutdown itself only getting the last 1/3 to 1/4 or so.
Jan 15, 2020 • 21 tweets • 8 min read
1. I haven't done one of these lunchtime (PST) Twitter talks in awhile, so here's a new one on the basics of information equilibrium that I've tried to make accessible.

At the heart of it, it's a matching problem. For example, matching labor force with jobs. 2. Unlike matching theory which tends to look at search costs, information equilibrium is more about constraints in general — like distance, education, economic conditions (firms' forecasts).
Oct 12, 2019 • 57 tweets • 10 min read
Inspired by Jo Michell's review, encouraged by the availability of a free sample section, and enabled by a Saturday with nothing to do while the family is out of town, I am going to live tweet my reaction reading the first couple of chapters of Macroeconomics . Don’t think it’s their fault, but the Mercator projection should be outlawed except for navigation.