The Recession the Unemployment Rate Can't See | Eric Pachman on the Data Beneath the Jobs Report

Summary

---

Excess Returns Podcast Summary

Episode: The Recession the Unemployment Rate Can't See | Eric Pachman on the Data Beneath the Jobs Report

Date: July 14, 2026
Host: Matt Zeigler
Guest: Eric Pachman, Data for the People

---

Episode Overview

This episode dives deep into the real story behind U.S. employment, labor market data, and inflation—exploring why headline numbers like the unemployment rate may be masking a labor crisis already underway. Host Matt Zeigler is joined by data journalist Eric Pachman, founder of Data for the People, who uses innovative public datasets and visualizations to challenge conventional readings of the jobs report, highlight the fragility in wage and job quality, and connect these trends to inflation and social stress.

---

Key Discussion Points & Insights

1. The Flaws in Official Labor Market Narratives

• Separation of Surveys:

  • Employment data comes from two BLS surveys:
    • CES (Current Employment Statistics): Surveys businesses; forms the basis of the non-farm payrolls number.
    • CPS (Current Population Survey): Surveys households; informs the unemployment rate.
  • Both have plummeting response rates and can tell conflicting stories.
  • Eric: “You will not see the recession coming with the unemployment rate. It’s already in the data… The crisis is in labor force participation.” [07:21]

• Labor Force Participation is the Real Red Flag:

  • BLS press releases downplay stories of declining labor force participation.
  • The main issue for the economy is the shrinking pool of working-age people, not just those actively job-seeking.

• Data Visualizations Overlooked by Policymakers:

  • Pachman’s dashboards allow users to drill into employment by race, age, sex, and veteran status, revealing grim declines in male participation and a surge driven by women in specific sectors.
  • Eric: “Go out and thank the next woman you see for supporting this labor force and helping it to grow.” [10:55]

2. Where the Job Growth Really Is: The Medicaid Care Economy

• "Medicaid Care Economy" Drives Job Growth:

  • Since 2024, 60–400% of state-level job growth has come from healthcare, particularly social assistance: home health aides, elderly and disabled care.
  • These jobs are underpaid (median ~$35,800), largely filled by women (80%) and immigrants (30–40%), and dependent on Medicaid funding.

• Risk of Medicaid Cuts:

  • Upcoming federal budget cuts threaten this entire job growth engine.
  • Eric: “Through the end of 2025… like 32–33 states would have had negative job growth if it wasn’t for Medicaid care economy.” [17:20]
  • Threat to caregivers and families is significant—investors and citizens should be alert to state-level funding risks.

3. The Wage Ledger & America’s Single-Income Stress Test

• Wage Ledger Visualization:

  • Interactive tool for exploring wages across 830 jobs.
  • Top job by size: Home health and personal care aides—lowest quartile pay, yet driving headcount growth.
  • Eric: “We are building basically all of our job growth and betting it all on people that are largely women, largely immigrants, largely living at or below the poverty line.” [21:09]

• Single-Income Precarity:

  • Simulations show a single median wage often fails to meet the poverty line for a family of four—any shock ($15,000 expense) pushes most geographies into red.
  • “If you switch to the 10th percentile income in any city, everything is blood red.” [51:56]

4. Inflation: The Data Underneath the “Noisy” Headline

• CPI vs. PCE:

  • CPI (consumer focus) and PCE (broader, more weighted to healthcare) are both constructed with arbitrary exclusions (e.g., stripping out food and gas for "core" inflation).
  • Eric: “Core is… bullshit. The strips are arbitrary; if you wanted volatility, you’d drop all of healthcare and keep all of groceries.” [26:07]

• Hidden Inflation Dynamics:

  • 42–45% of CPI items have breached one standard deviation over their means ("anomalous").
  • Volatility and upward pressure often hidden by small category weights but collectively significant—e.g., “gardening, lawn care, hair care, coffee, beef… all spiking.” [29:18]

5. Oil, Energy, and Tail Risks to Inflation

• Petroleum Inventory Crisis:

  • U.S. is exporting record oil while drawing down strategic reserves to historic lows.
  • Misalignment of oil type/refinery leads to sky-high crack spreads and inefficiency.
  • Eric (Analogy): “Think of oil as fruit salad… every type of crude oil is different, and our refineries aren’t always designed for what’s coming out of the ground.” [33:38]
  • Closure of shipping chokepoints (Strait of Hormuz) is likened to “creating Covid for oil markets.”

• Diesel Is the True Linchpin:

  • Diesel shortages ripple through groceries, logistics, and other inflation components—all while retailers like Walmart weigh eating costs to hold prices and defend market share.
  • “Gasoline is the largest mover, but diesel is more important because that feeds into almost all the other ones.” [44:31]

---

Notable Quotes & Memorable Moments

• On BLS Press Releases:
  “Your entire story is that there’s no story. The crisis in the labor force participation rate comes first.” – Eric [05:07]

• On Modeling & Data Interpretation:
  “All models are wrong. Some are useful. They just made their model a little more useful. Still wrong.” – Eric [08:00]

• On Policymaker Accountability:
  “Shame on the people...interviewing the chair of the Fed… for not asking these questions. I just had a high schooler do this!” – Eric [22:48]

• On the Medicaid Care Economy Risk:
  “If you have a parent or grandparent who relies on home health care aid, you better figure this out right now.” – Eric [18:40]

• On Inflation Construction:
  “If you really wanted to measure pain… you would measure gasoline and groceries first and foremost. But that’s not what we do.” – Eric [27:59]

• On Diesel Shock:
  “Diesel prices were up 13% that day… people are most worried about diesel right now… That’s very unsettling for CPI.” – Eric [44:16]

---

Timestamps for Key Segments

• High-Level Labor Market Overview / Two Surveys: [02:59]–[06:55]
• Labor Force Participation and its Decline: [07:00]–[11:30]
• Medicaid Care Economy Explained: [12:37]–[18:55]
• Wage Ledger & Job Growth Quality: [19:12]–[24:10]
• Inflation—CPI, PCE, Underlying Data: [24:44]–[30:29]
• Oil Inventories, Energy Inflation & Crack Spreads: [31:01]–[41:52]
• Diesel, Supply Chains, Retail Response: [44:16]–[47:49]
• Single-Income Precarity Stress Test: [49:00]–[54:24]
• Mission and Technology at Data for the People: [55:08]–[61:45]

---

Data for the People – Mission & Technology

• Transitioning to 501(c)(3) non-profit for public benefit, open access.
• No paywalls—funded by donations from those who can afford it.
• Using AI to dramatically accelerate data journalism and create open-source tools for public and community groups.
• Eric: “Our vision ultimately is to create a shared basis of reality for America and for the world. Data can do that if we choose to use it for good.” [64:40]

---

How to Follow & Support

• Website: data4thepeople.com (number 4, not “for”)
• Email: [e-mail]
• All tools, methodology, and raw data are free and encourage critical engagement.

---

Final Thoughts

Eric Pachman and Data for the People offer a radically transparent, data-driven lens on the U.S. job market, challenging surface-level interpretations and sounding the alarm on unseen economic risks. The conversation is a sharp reminder: the headline numbers rarely tell the whole story, and thoughtful investors—and citizens—must dig deeper to understand the real state of the American economy.

---

Transcript

[A]

This episode is brought to you by Google Chrome. You think you know a browser, but Gemini and Chrome, that's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50 page restoration block. Or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it, ready to make anything online make sense. There's no place like Chrome. Check responses set up, required compatibility and availability. Various 18 plus foreign.

[B]

You're watching excess Returns, the channel that makes complex investing ideas simple enough to actually use, or better questions lead to better decisions. I'm Matt Zigler and today's guest is the author and data nerd at Data for the People, like it says on his shirt, Eric Pakman. Welcome back to Access Returns.

[A]

Thanks for having me. Great to be back.

[B]

You came on Last Call this past month. You did a segment on what the Fed is looking at when they're talking about employment and inflation. I promised we would do this episode to go deeper into each because the stuff you've been writing at Data for the People has been just on fire lately. So we're going to do the deep dive today. You said though, right before we recorded, you wanted to say thank you. You got a bunch of traffic, a bunch of people checking out the site.

[A]

Yeah, it was almost kind of shocking. I was getting these like anomaly reports from Google and everything. The search traffic. We're a sleepy little site building all organically, all word of mouth, bare bones to zero funding at this point in time. Just doing the work because we love it. And yeah, I mean, I was just on the tail end of your last podcast, like your weekly recap type thing, and all of a sudden it's like 60 people signed up for my distribution list. And yeah, distribution list, that's a heavy lift, right? Like you have to actually go to the site, read the slate, click a button, put your email address in. I get it. Like I. That that takes some, some commitment and,

[B]

and you already have a listener base with gumption. We've got some gumption out there. And I'm telling you, very grateful. I'm telling you, dear listener at home, make the man grateful. And you want this data because this, this is where I'm starting us. We talked about the Fed. The way the Fed's looking at the numbers. Eric thinks there are flaws and problems not just with the numbers they're looking at at the top level, but then as we drill down into their component parts. So drawing on public data sets like BLS data and other places, he'll give us sources as we go through these. This is not the arrangement that they're looking at this data lens through. So if you're running money, if you're handling stuff with clients, you want to see data for the people and what he's doing. So I'm going to ask you very, very high level right now I want to lead with employment, I want to lead with labor.

[A]

Yeah.

[B]

The employment trend in the US right now is employment running up, down, sideways. What direction does it look like we're heading?

[A]

Yeah, so I think that it's actually you phrased that question in exactly the right way and we have a dashboard for that. So if you go to data for the people, we just released a brand new dashboard where we take the current population statistics data. So this is the household survey. So just a little bit like little aside here we get the employment report every month, moves the market one way or another. There are two surveys that are published here. Hopefully most people know this. The ces, that is the institutional survey, that is where the non farm payroll comes from. And the cps, that is the household survey, that is where the unemployment rate comes from. So those are separate. They, they sometimes will tell completely different stories because they're totally different surveys. One is serving people, the other one is surveying businesses. Right. Both the response rates are falling off a cliff right now. So there's that, but that's a whole lot other discussion. So how much we trust them, I don't know. But they're the best data that we have. So when I was with Ban Creek in asset manager before I started Data for the People, we created a really good, I mean it was a, it's called the tree map. And go to bancreek.com, you can look at it and you can drill down into the non farm payroll data. So if you see like I don't know what last month was, it was like 50,000 change or whatever that's what is called at the BLS hierarchy level. Zero, highest level. And then when the number gets higher you go down to more granular levels until you get to like sewing machine fabricators or whatever down at level seven. So you can see a lot of detail on all the industries that's already there at Ban Creek. That's the ces, that's the survey of businesses.

[B]

Right.

[A]

Non farm payroll data. So the problem with the CPS data and again sorry for all detail but I think this is really important.

[B]

This is really important because these surveys, how the surveys are being responded to, how the information is being captured is not the headline number. And that's why I wanted to unpack this.

[A]

No, no. And it will also, like, if you can see this data, which, this is the challenge. Like the bls. Look, I'm not trying to hate on them, but their job is not to make you see the data. If you look at the release from the BLS from last month, the press release, it's like, unemployment, unchanged. Unemployment for black people, unemployment for this, unchanged. It's like, great. Your entire story is that there's no story. And then it's like, oh, and by the way, the labor force participation rate fell to 61.5%. That was down.04 points. And I'm like, what the. And then you're just like, moving on. And you're just like, whoa, whoa, whoa, time out. Like, I'm a storyteller. You lead with the crisis in the labor force participation rate first rather than like, there's no change in the unemployment rate. But this is the issue with the BLS right now and this is the issue with the Fed. It's like you have to go back and like, look, ask Claude, go back and study the history of how these different benchmarks came into place. And there was a comment on that last video. I want to address this comment. It was like, you're not, you're telling me that the Fed doesn't have access to all these tools? Of course they have access to all these tools. They know all the minute detail. But they can't all of a sudden come out and say, like, for the last 30 years we've been focusing on the unemployment rate and guess what? That no longer works. The world has changed. Like, they've just kind of painted themselves into a corner to say the unemployment rate is the one ring that rules them all. And right now I can tell you, investor person that cares about America, you will not see the recession coming with the unemployment rate. It's already in the data. The depression, the recession, the crisis, let's call it the labor crisis. It could be like it's an inflationary surge because we just don't have enough people to work. But if you go to data for the people and you look at our latest data visualization, which takes almost all the data in the CPS survey and creates a drill down tool to where you can look at every dimension. Forget about the unemployment rate, that's in there too. For you to look at it and see that nothing's happening. Go look at the people that are not in the labor force right now. You can actually see, and we just wrote about this, the revisions that are happening as the BLS improves their model and the census improves their model. And you can see the number of people that are working age people just disappear. And so you're just like, well, wait a second, where are they going? Well, they're 75 plus now. And it's like, well, the actual age of people didn't change, but the whole BLS and the whole census runs on models. And we just wrote this. There's this great quote. All models are wrong, some are useful. Well, they just made their model a little more useful. Still wrong. But now all of our projections going forward have less working people to pay, taxes to pay into Social Security to pay into Medicare. This goes way beyond investing. And I'm not here to tell you like what to believe in it. I'm here to tell you the tool is there. And so you want to answer your question on the number employed people, just go select the series that says employed and you'll see it. And then select it by race, select it by ethnicity, Hispano or Hispanic, Latino. Select it by sex, select it by veteran status. Do it all. I mean, it's all there. By the way, the veterans one is crazy. Like the veterans are falling off a cliff. Like, we're just, we're not. I mean, this is, this is well known, but I just learned it. It's like, oh, wow. You know, clearly a lot of the Vietnam era, Gulf War era, they're all dying off in, retiring out, but they're not being replaced. So it's like, where are all the people that are fighting for our country? They're not there anymore. Anyway, that's an aside. But like all the data is there.

[B]

Can we get a slide up for. Let's look at beyond the unemployment rate because this is part of what we're dancing around now. So let's talk about this. What's going on here?

[A]

Yeah. So, okay, when you pull it up, it will come up to age 16 years and, and over civilian labor force. So I think it's very, very important, if you understand how cities work, to focus on the civilian labor force. Because that is the number of bodies that create tax revenue. That is the number of bodies that can actually go out and shop for things, support businesses, until we come up with a mechanism to tax bots, AI and massively tax big tech, which, like, I'm not holding my breath. This is the business model of the US we need people to work. And so that's why you have to look at the labor force. So right now you see this long term trend going all the way back to the 1950s for 16 years and over, but now just go through and There are literally 30 different dimensions and cuts that you can look at in here. You got to spend a little bit of time with this and really have the desire to be educated here. So just go in and look at. I'm going to look at the number of 45 years in over people. Okay. Then that pulls up what's called a non seasonally adjusted chart with the 12 month average on that in the civilian labor force. Well now what if I'm concerned about or if I'm interested in the 45 years and over, but just the labor force participation rate, then you're like, oh wait, what happened here? So in 2009 that was at 55.9% and now the 12 month average is down to 50.7 and it is falling off a cliff. And by the way, I haven't even looked at this chart before. There are too many dimensions in here to look at. But now all of a sudden, since what happened in January was this whole like reshuffling of the models, the labor force participation rate is really, really dropping fast for the 45 years and over, it's down to 49.6%. So then you could say, but what about men? Oh well, so men. Oh boy, that chart is death. So men is down to 54.7%, the lowest level by far on record. The prior minimum was back in the 1990s where it was 57.5. It went all the way up to 62.9. Women, well, they're much better. Which this actually jives with the piece that we just worked about women are the driver of our labor force. We have to thank them. If you're a man, go out and thank the next woman that you see for supporting this labor force and helping it to grow. And if you don't believe me, spend even five minutes working through this database to see that men are not participating at an increasing rate and women are participating. Increasing rate women, which we can go into that has a lot to do with which industries are the driver of our labor force expansion and employment, just in general.

[B]

So let's take it there. Next. Another part that's been fascinating is as we look at those jobs that are coming out, many of which in the 45 plus crowd. So these are the people, you know, these are my brethren, these are my peers, these are the ones who are making the six figure salary, doing some job. Those are the ones that are going away right now. The weird part that's stabilizing the number in employment right now is the other jobs that are being added and they're not that six figure white collar. Go out and do something because you're 45 with a college degree job. They're coming in what you're calling the Medicaid care economy. And you've got some insane charts explaining why. Because you've been talking about this for two years. Healthcare is the driving place and this Medicaid part of it is huge. Let's, let's get some more charts up.

[A]

Yeah, so I wish I remembered all these stats right off the top of my head. But you can, you can go to. Well, let me, let me step back. So this has been about a year and a half in the making. So there is a visualization on our website called Medicaid care economy, concentration of state growth, state job growth. So back when I was at Ban Creek, you know, my background's in healthcare and so I'm looking at these healthcare job growth numbers and it's like 60% of total payroll. 70%, 80%. By the time we got to the end of 2025, it was like, I don't know, the numbers are in my reports, but it was something like 2, 3, 400% of all job growth was coming in healthcare. Everything else was contracting. Wall Street Journal covered this. Everybody knew it by that point in time. But what they didn't know because people don't tend to drill down like we do, is they didn't know where those jobs were coming from. And so people like are very confused about, well, is this all administrative oversight because of how inefficient our care economy is? Is it actual doctors, by the way? No, it's not Dr. Growth. Is it nurses? Yes, it's nurses growth. We can show you that in a visualization we have. But overwhelmingly this growth is coming from what's called social assistance. And within that, individual and family services. And within that, services for the elderly and disabled and home health care, aids. And so what is that? That is, you know, I have an elderly parent or grandparent, they don't need to be institutionalized yet, which would be very costly for me and the state and, or the state. And so you can drop them off at an elderly daycare facility. They can, you know, keep track of their meds, they'll have socialization, you can work your job and a win win. Right. Or what we more commonly know is you can hire someone to come into your home to take care of your parent well or your grandparent. Those people. By the way, we have another data visualization which hopefully you'll put up which shows you what everybody makes, 830 professions across this country by city. If you go and explore that, you'll see that these are some of the most underpaid people in really in the entire economy and is the engine of job growth. We wrote another piece on that which I'll give to you to put in the show notes. But it basically is like we've created a couple million jobs in this over the past decade or so. And so our aging of the economy, which I think most people know, the boomers sort of rolled over into the 65 plus like in the middle of the 2010s that has created huge demand for people to take care of them. By the way, most of them are women. I think it's like 80% are women, something like that. And about 30 to 40% depending on the study are immigrants. So there's another issue there with maybe a potential supply crunch with the people to take care of them. Anyway, all that said, I'm staring at this for like a year and I was like, then the one big beautiful bill hit, right? So what's that going to do? Well, it's cutting Medicaid by a trillion dollars over 10 years. And as putting my investor hat on, I'm like, okay, we need jobs to keep growing. We need people to take care of old people to keep growing those jobs. So the Fed says, oh, the job growth is healthy. And by the way, what I happen to know is that most of the home healthcare aids, the primary funder for them is Medicaid. And then Trump in his infinite wisdom, and I'm not being political here, I'm just saying that he probably doesn't understand the way this works. He decided to cut this or the administration decided to cut funding for this, which is basically the major growth driver of our jobs. Look, probably not advised very well, but Medicaid is a state based program and so you can't really understand what's going to happen unless you drill down into the state. And so that's what I did. I went down into what's called the quarterly census employment wages data. That is by the way, one of the most valuable databases if you want to understand anything about job growth. It is based on unemployment insurance claims or unemployment insurance reports from states. It covers 98% of all W2 employees. It is not a model. It is the only real source of data that we really have. At the bls, it comes out like six to nine months in arrears. So a lot of investors don't use it. But my gosh, you can actually get the individual industry down to the individual county. And so I've used this database for a lot of things, for a lot of the stories of America that I tell. But as an investor, especially with AI, there's no excuse you should be using this database as well, or at least reading my stuff because I use it a lot. So that's where this data visualization came from. You can basically look over time and you can actually press this little play button in the window n quarter and then just see how much job growth came from what's called the Medicaid care economy. So this includes assisted living facilities, it includes nursing homes, that kind of stuff. How much job growth as a percentage of all the job growth in that state over a one year period, a two year period and a three year period. And what you'll find is that that map gets increasingly more red. And now we're at a point where through the end of 2025, which is all the data we have in that one year period, like I think it's like 32, 33 states would have had negative job growth if it wasn't for Medicaid care economy. And then think of that setting up into the beginning of 2027 is when most of those cuts start in Medicaid. And so this is something that every investor needs to at least have on the back of their mind as like this is a major tail risk. Now maybe it's a tail risk to the upside because if all of a sudden we have no jobs, then people just care about rates getting cut and then everyone gets really excited about more speculation. I can't tell you which way the market's going to go. All this helps is saying that there's a major tail risk that's out there. And I helped you quantify it with this, that is going to impact the real job market. And if you have a parent or a grandparent that relies in any way on home health care aid, you better figure this out right now and figure out where all the money is coming, what staffing agency it's coming through. Because that could all of a sudden just disappear depending on what state you live in live in.

[B]

It all ties back to consumption, it all ties back to real growth in the economy. I want to hit one more part on employment before we start getting into the cost side of this and talking about inflation. I want to talk about the wage ledger because I think this is another fantastic data series. Yeah.

[A]

So this one, if you actually go to the wage ledger, I'd love if you do that because I am the most proud of this one. Not because it is the most involved, the most complicated. In fact there is a spreadsheet that lives out on the BLS website. It's called oews Occupation Employment Wage Statistics. Everybody should look at it if you care about, you know, how much people make, you know, I'm sure that there's some investing thesis for knowing that kind of stuff and how that's changed over time. The problem is is the spreadsheet is like largely unusable in Excel. It crashes, it's hard to filter. And so all we did here was take that and put it into an interactive data visualization so you can sort through that and sift through it and learn from it. And not only that, but the main reason I'm proud about it is that you'll see the author of this is a gentleman by the name of Jonathan Pickens. He is a rising senior in high school and so I am working with him. He's one of my three data storytelling fellows that we have at Data for the People. And this is kind of a work in progress project. He's doing a really big project which I'm sure after we publish this you're going to want to talk about it again because it's evaluating the changing cost benefit of a four year college degree over time versus a trade career which is very complex. But as he's building the databases and getting to learn all these databases, he's coming up with these interactive tools that on their own are very, very useful. And so this wage ledger is I thought was fantastic that he came up with. You can basically go through 830 different jobs that are shown, sort them from high to low on the annual median. What will come up is the number one paying in the country. By the way, this is only W2 employees so you're not going to find Elon Musk on here or any people that own businesses. Pediatric surgeons make a median salary of $559,000. The there are 1,190 of them. But then what you can actually do is that sort that by employment boom, what comes to the top, Home health and personal care aids 4,305,810 making an hourly median of $17.21, $35,800 a year. The paid range, you can even hover over that pay range. You can See that it is very tight. The 90th percentile makes $45,000 a year, which by the way, if you go to another one of our data visualizations, you will find in some of the higher cost areas that still doesn't cover the cost of poverty for it to raise a family of four. So we are building basically all of our job growth and betting it all on people that are largely women, largely immigrants, largely living at or below the poverty line to support a family of four. And this is where we come back to the Fed to make it full circle. The Fed is report, they're not telling us about any of this. So do they know it? Probably. But this is a political game, right? Why are they not telling us that the quality of the job growth is coming all from home health care and personal care aides that make $35,000 a year? Like these aren't jobs that are creating tremendous disposable income for average Americans. That is an important data point I would think we would want to have if we are claiming it is a strong or robust job market. But we're not getting any of that information. And shame on, sorry, I'm going to say it. Shame on the people that are interviewing them and interviewing the, whatever the guy's name is, the chair of the Fed, this new dude. Shame on them for not asking these questions. I mean, I really can't believe that the people like that are invited to the room don't have access to any of this data. The people at Bloomberg, the people at Barron's, wherever they are, whoever's coming there, they don't have data analysts that can do this. I just had a high schooler do this, so don't tell me you don't have the chops to do this. All right, I would love to be in that room and ask some questions, by the way. I would love to see what the answers are. Because then we actually get to figure out, does he really know this or does he not even know what we're seeing here? Because this is the level of detail that I think investors really need to know. But right now it's just a game of like your investors that are listening to this are going to say, oh, wow, now I know this. But like Ben Hunt says, if this is knowledge that only investors that on this podcast have, but they know that not everyone else knows this, then the game of investing is to really figure out, well, what does the consensus know? And if the consensus doesn't know about this, does it even make sense for me to act on it because the real economy is going to suffer. We know this from the data. All the data says that you can triangulate any which way. But if no one ever figures it out or it takes two years, what are your clients going to say? Like you're going to sit on the sidelines waiting for that. It's a very hard job, this job that you guys have. Thankfully, I don't have to do it anymore. I just have to diagnose all this and you figure out what to do with it.

[B]

You found a good, you found a good place. And if I get Warshawn speed dial, you're going to be one of the first ones I get to know. I want to take this to inflation next. I want to go the cost side. And the reason I want to go there is because this is another data series that's enormously clunky, enormously funny in the way that we even talk about it or measure it. So I'll ask the same question that I did about labor to lead off with inflation as we understand it at the top, top level, up, down, sideways. Where is inflation moving right now?

[A]

Up, up for sure. That one's an easy one to tell you. Like, you don't even have to do all the work that you're doing on the job side to be like, well, wait a second, it's this, the not in labor force, the participation rate, like all these things that people don't talk about. You got to drill a little bit for that. Inflation is just like, just look at the data. I mean, like, we know that it's up, but we, we published about two years ago a data visualization on Ban Creek bancreek.com it's free. The first one we ever published on CPI, we followed up with a PCE one. And all it does is it takes the underlying kind of items, about 180 mutually exclusive items, and it will size them and show you like this is the, the impact that this is having on inflation.

[B]

Just like you did with CES and cps. Just take a second and talk about cpi, pce, how you guys broke that down to show it. Because this is, People need this reminder.

[A]

Yeah. Okay. So CPI is surveyed, measured by the bls. It is supposed to be the consumer's experience of inflation. And so the weights are different from PCE because that is like the broader economy's experience of inflation. So, you know, each of them have different weights, they have different ways that they measure things. So when you drill into pce, several of the items will actually be taken from cpi and they'll be the same and then others will not, they'll be taken from ppi, you know, and so it's this mishmash of them. But the best example I can give you with pce, and given that it is like the holistic view of the economy's experience of inflation, is it has like a much higher weight for healthcare because you have employers, you have the federal government, Medicaid, like you have a lot of non consumer payers, mostly non consumer payers within healthcare. And so you'll see a much larger weight there and then you see a much smaller weight on shelter. So, you know, PCE is. The Fed says it's what they look at over cpi. The Fed also says they look at core. Right. I have been on record and written multiple times about how core is part of my French bullshit, because CORE was defined and you can do the research on your own by coming out of the 1973 oil shock or whatever. And then it was having impacts on food prices. And the current Fed chair at that time was like, well, this is noisy and so let's strip it out. And so they stripped it out and we haven't revisited that ever since. And I wrote a piece of, I can't remember exactly where this was, but trust me, I wrote a piece that actually looked at the volatility of these categories and it's like, okay, gasoline is really volatile. Makes sense to strip that out if you are really looking to smooth this out. Still a really important variable for the consumer. So we probably do care about that a lot. But food is not variable. It's not nearly as volatile as healthcare. And so it's pretty arbitrary the way that core is measured right now. If you were just doing it based on volatility, you would strip out all of healthcare and you would include all the groceries. And then if you were really building a benchmark that you wanted to measure the pain and the suffering of the average American, you would measure gasoline and groceries first and foremost, because that's what we feel and that's what we experience. The surveys themselves have all sorts of assumptions baked in. CPI is heavily influenced by shelter. And the number one item by weight is something called oer, owner's equivalent rent of residences. It itself is a proxy. It's a fabricated surveyed number that is trying to come up with kind of an estimate of what rent would be and the changes in rental from month to month for people that own their homes. So they literally just place a call to you. I don't know if anyone who's listening has ever got this call? I haven't, but they place a call and they're like, hey, what would you rent your house for right now? If you could, theoretically, hypothetically, what would you do? I just go to Zillow, be like, I don't know, I mean, like, what does Zillow say? And so like then that's what they. I'm assuming that's what people do. And then that is 30%, that's about 26 to 27% of all of CPI is just on that. And so why did CPI go up so much in 2022? It's just because shelter was going up so much. This one item was. And why is it like staying muted now? Because shelter's staying muted now. But when you look underneath that and we published a report on Ban Creek, I think something like 42 to 45% of all items out of the 180 have breached one standard deviation over their historical mean. So they've moved into, ooh, this is anomalous, you know, and you also have like a huge amount that are above two standard deviations. And so you can see this real easily on that tool. You just hover over them and you can just all of a sudden see a bunch of stuff like spiking. Gardening, lawn care services, beef, coffee, like random stuff like hair care, I don't know, like it's weird stuff that that is just kind of all going kind of parabolic. And you're. We did that report on Ban Creek and I thought that was a really good one. Just because it's like no one's talking about any of this stuff. The weights are so small that you really don't care about your laundry care services that are spiking. But it's like all of a sudden you put all of them together and you're like, none of these things actually are related and yet they're all spiking. That doesn't feel so good. And that's free. You can see that very easily in our tool.

[B]

So one of the places that you've been drilling in that I think is the most valuable. And certainly this came up when we were recording that last call a week or so ago was looking at oil and you've got some charts on the petroleum inventory seasonality that I think are fascinating, especially where we are with the straight of Hormuz, our on again, off again war, and the three two, one crack spreads. I think these are really, really useful to understand right now, even for the people who aren't oil. Investors or aren't tied to that market. Everyone, it flows through to everything.

[A]

Everyone needs to become an oil investor to be an equity investor. That's what I think. So let's start now.

[B]

All right, let's start now. Petroleum inventory, seasonality. What's going on there? Well, this comes straight out of your background. Should call this out too.

[A]

This is a new thing. And let me give you like a bit of context and background. So I have for years. I'm. I'm a chemical engineer. I worked at ExxonMobil. I modeled refineries. I know this shit, right? I tried to forget it. I can't. It's been tattooed in my brain how crude oil gets processed through a refinery. I have been trying to come up with a way, I like speaking in terms of analogies, metaphors, whatever, to explain oil to people that they'll understand. Like, I don't know, like at the level of my 11 year old, where he gets it right, Because I think people think that oil is one thing. Crude oil. Oh, we have lots of crude oil in this country. We're producing record crude oil. We'd have no problem. People who will go unnamed are saying that, and people probably are just believing that. Yet let's look at the evidence. We're exporting records amounts of oil, still importing a lot of oil. And you're like, and then we're drawing down our strategic petroleum reserves to the lowest level in 45 years, almost on record. We're getting close to that. The oil inventories, when we look at those charts, which we'll talk about, are just all crashing. And you're like, why would you export

[B]

your oil

[A]

if we're out of oil? One would think that if we had any strategy whatsoever and rather than the market or basically just selling us the highest bidder, which that could be, it is like nobody's at the wheel right now. And so of course, if I have oil and I'm a company and I own it and I can get a higher price by selling it on the market, and I'm not forced to send it to a refinery in Pennsylvania to basically say, well, we want to bring gas prices down in the US then that's what I'm going to do if no one told me to do it. And so that's one part of it, but the other part of it is that oil is not oil oil. And I actually have, by the time this come out, hopefully it's out, but I should have an op ed coming out in a major national publication. So that will be posted on Data for the people.

[B]

And as soon as it's out, we'll get a link to it. Look in the description. If it's out when this is out, it's in the description. So go ahead, keep going.

[A]

So. So I have, I have an analogy in there that I finally think can work. And so we need to tell everybody about this analogy. Oil. Think of oil as fruit, or even better, a fruit salad. So everybody knows that if you order a fruit salad, it's going to have cantaloupe and it's going to have honeydew and it's going to have blueberries, it's going to have strawberries, it's going to have apples. It's kind of a mix of all the different types of fruit. Oil is not blueberries. Oil is not one type of fruit. It is a blend of all the different types of fruits, which are really hydrocarbons. So there's some fruit which creates propane, and there's some fruit which creates gasoline. The ones with six hydrocarbons or the, with six carbons or eight carbons or something like that. There's some fruit that creates diesel, some fruit that creates jet fuel, asphalt, petrochemicals, all that kind of stuff. And so every type of crude oil is different. The type that you drill in the Middle east, that's where most of the oil has come from over the history of our world. And so most of the refineries which really haven't built a new refinery here in probably my lifetime, most refineries were designed to produce that kind of stuff, which is heavier. It has more of the diesel in it, it has more of the jet fuel in it. It is more of like the, the, the heating oil, more of the heavy stuff. So refineries develop technology called fluid catalytic crackers and cokers to take the heavy crap in the oil, billions of dollars of technology, and crack that and morph it chemically and physically into gasoline. Now, the stuff that we're drilling here, which is very recent, right? This happened in like the 20, the 2005, 2006, the fracking revolution, all that kind of stuff, the shale oil, all that, this is very light oil, lots of the gasoline molecules in it, not a lot of the other stuff in it, right? Well, now if you take that and you put it into a refinery, there are actual physical limitations of what that refinery can do. In the entire refinery, the economics are managed to maximize the economics. So you want to maximize the yields. There are specs on every product. I mean, it's very, very Complicated. And I built refinery, I've refinery models to optimize all of this based on what type of crude. And I can tell you that a refinery on the Gulf coast, you would never send, you'd never be like, oh, I'm going to take all the oil from the Permian Basin, which is literally right there, and send it all here. Because all of a sudden the yields of crap that comes out the other end, it's all broken. It's not, it's not what the refiner wants to maximize their profitability. And so what happens when you have the wrong type of oil matched to the wrong refinery? The crack spreads rise. And so the crack spreads is the market telling you this is three barrels of, it's a proxy for refining margins. So it's three barrels of oil equals two barrels of gasoline and one barrel of diesel. All three of them are separately traded products on the market. They all are just like stocks, right? You know, nobody is fixing these prices. Nobody's controlling this. The refineries get, get what they get from these prices. And what's happening is when oil gets mismatched to the refineries, the economics of the refineries are bad. Now if you have what's called a hydro skimming refinery, which is like actually one of the cheapest type of refineries that wants to process like light sweet crude that we produce, man, you're, you're a cash machine right now. You could just run that thing flat out, make tons of money, $60 a barrel in, in crack spread, which is four to five times your historical profit margin. You're making 4 to 500% profit. That's why refining stocks are probably up. I don't know, I don't check them, but my guess is they are. But if you're like that Gulf coast refinery, the big, the bad refineries, the Exxon refineries, the bp, the shell, that's not the kind of oil you want. And so what I wrote in this op ed is we close the street. We open the street. When you close the Strait, all the flows across the country, the world get disrupted. The wrong oil goes to the wrong place and all the economics are screwed up. This is Covid. When we closed the Strait, we created Covid for oil markets, right when we had Covid. Remember what it did to all the supply chains for like two, three, four years. But we were back like wearing masks in public. I mean, there wasn't any immediate near term threat after we realized we needed to mask and we had vaccines. But it still Took years for supply chains to recover. You don't even need to know about refineries. You just need to know that we just shut down the bottleneck of the key choke point. We shut down Atlanta, Hartsfield. If you think about this, from the air traffic network for two months. I mean, don't tell me that not every single flight around this country will be affected eventually. Maybe not on day one, but it takes a long time and the air traffic network is underselling it. Like that's actually a lot easier to reset because you just don't fly overnight. Whereas, I mean, think about all the ships that are sailing around the Cape, sailing all these like extended voyages to not pass through the strait and everything. It'll take months for those to actually get back to where they need to be. And then we just keep shutting it down and then oping and shutting it down. And the insurance premiums are high. Like arguably this may never go back to where it is. Especially if Iran is able to formalize the pgsc. They start charging fees on this, the capacity is permanently lower because we clearly don't want that to happen. All bets are off. We just created Covid for the oil markets from a supply chain standpoint. We're seeing it in the inventory visualization that I put together, which by the way, this is yet another example of the government not giving you the data that you really should have. They tell you what was the percent change versus the last five years. And then you're like, oh, diesel inventories are down 10% versus the last five years. You're like, 10% isn't the scariest number ever. And then you're like, well, wait a second. In 2022 we had the Russia, Ukraine thing and then everything was going crazy back then. Oil was like at 100 plus. You know, gasoline prices were at $6 average. That's one of the five years. I'm comparing it against the worst year on record. Why wouldn't we compare this against all 40 plus years of history? I don't know. It doesn't make any sense to me. And that's why I did it. So, so you can actually go and look at the diesel inventory chart, the gasoline, the Strategic Petroleum Reserve crude oil, and you can look at it against all history and you can click on and off of different decades. So what if you don't want to compare it to the 90s or the 2000s? Just click those off and then it will take those off the charts and then you can actually see how bad things really are. And it's scary. It's scary bad. Like if inventories matter at all, this is Covid happening in real time and the investing public couldn't care less. Like it's, it's like Covid without the transmissibility until it hits. And then like my worst case, which we're heading towards and I think a lot of oil analysts are talking about this. It's like one day you'll show up at the gas station, you just won't be there. Or what happens when El Nino hits? What happens when hurricane season comes through and one of them happens to take out a refinery in Louisiana? I mean, we're basically just sacrificing all of our risk mitigation for the sake of whatever. I mean short term profiteering from the administration, from the government, from investors. Like we're pulling forward all of the gains. So go party while you can. But like, I mean like, I look at this as, as a knowledgeable, not investor, but like concerned citizen that happens to know how this all works. And I'm like, I, I better prepare, you know, I better prepare my family for this. I, I better just get prepared for times to get very difficult. Maybe we end up threading that needle somehow. I don't know. I can't tell you what's going to happen. I can just give you a probabilistic assessment of it and it is not pretty right now. Don't take it, look at the charts.

[B]

Tie that back to cpi, PC. The way that most people are looking at this, because I think what you're highlighting is that's especially important is the tail risk. This represents if there's overlapping events. So this helps put applied pressure that's not going away onto the inflation numbers across the board. On top of that though, how much is this showing up in CPI and PCE now versus in that tail scenario? How much more does it move Those numbers where the Fed actually looks at this and says we have an inflation problem. Yeah.

[A]

So, oh man, how do I say this kind of like somewhat politely. So let me be honest with you, like, I haven't really looked at PCE in a while, but PCE is the same as cpi, just with different weights. You know, go look at it yourself, ask Claude, they'll give you all the weights. You know, so like one weights will be higher, one weight will be lower, it doesn't matter. But cpi, I can tell you, I did a study with this at Ban Creek. I can predict almost exactly to with like 99% certainty when the AAA retail gasoline price gets finalized at the end of every month, which if you have a BL Bloomberg license, you have access to AAA data. If you don't, it's very hard to actually find that series. It should be in the public domain, but it's tough to get. So if you have that average the entire month, compare it versus the past month or compare it those retail gasoline prices versus the last year and then the gasoline component of CPI will come in almost spot on on that every single month. That's all they use. And so I've done the correlation. It's like a 0.99% or 0.99 r squared it that that's just what it is. So I can tell you the gasoline component, we know that. So I mean Honestly it's July 10th right now, so we know exactly what the gasoline component is going to be. I haven't looked at it yet. I should have gone going into this podcast. But maybe I'll send it to you and you can kind of post in the show notes. We know exactly what that's going to be going into next week's CPI release. Everything else is just based on the methodology and the surveys and the error and the models and everything in the bls. And so I couldn't tell you that like laundry services or gardening services were going to be spiking, but yet they are like, I mean, but that means

[B]

that this is the moving data point that basically drags it, doesn't drag the others with it. But if something is the most volatile component and it's this, then this is where you want to understand where that surprise risk comes from in the whole data series.

[A]

Well, and I can tell you that that's not the part. So gasoline is the, you know, that individually is the large, the single largest mover because it's so volatile and it's a large weight. But diesel is more important because that feeds into almost all the other ones. So think about gardening and lawn care services, diesel, think about, you know, all of your grocery store food diesel. Why is that the case? I mean I also happen to work at a railroad and so fuel surcharge is a really big thing. For the trucking industry, all runs on diesel. For the railroad industry, all runs on diesel. The fuel surcharge gets passed through to the consumers. Now I'm not an expert on this, but I'm sure people are on this call. I find it very interesting that you're hearing the targets in the Walmarts of the world talking about reducing their prices in the face of clearly what could be dramatically Rising costs for them because of they're having to eat fuel surcharge as it migrates through. And this only gets worse and worse and worse. If we have this on again, off again thing and if the diesel inventories keep dropping, I mean you saw like on the day where this is telling, on the day where we went back to war last week and everyone started bombing again. That same day the EIA inventories were released and crude oil, commercial crude oil inventories actually went up because the refinery utilization dropped down. But gasoline dropped by like 2 million barrels and diesel dropped by like 3 million barrels. Diesel prices were up 13% that day. Like this is the main choke point. People are most worried about diesel right now. And that's, that's very, very unsettling for CPI because that could be like permanent long term, not permanent, but it could be long term inflation that works itself into cpi because all that kind of stuff has to get kind of passed through over time. Unless Walmart and Target and Costco and all of them are willing to basically step in front of this bus or this train, pun intended, and say buses and trains. Yeah, we're willing to kill our margins to basically protect the consumer. Which to me if they're willing to do that, that means not like the top 1%, they're immune largely until the shit really hits the fan. But that means the rest of us are really, really suffering because I mean, why would you, as you know, someone that cared about quarterly reports as one of these large retailers, you're looking at rising costs pretty obviously as fuel surcharge passes through. Why would you ever cut costs right now unless you were really, really, really concerned about volume at the low cost retailers? So that's a concerning data point. Don't take my word for it because I don't read these, these transcripts. I just saw, you know, news in passing. So go out, validate that for yourself. If they really are cutting that, that is something that, that I would really want to understand. But my own, like I shop at Walmart every week. Again, we're a nonprofit or so I have to shop at Walmart and I can tell like they, they are cutting prices like at least in my, like anecdotally I'm, I'm shocked, I'm really shocked that they, given the inflationary pressures that they're probably just subsidizing at this point in time. It's interesting, really, really interesting what's going on.

[B]

It's fascinating especially when we get into corporate earnings and where we get into this and this is A broader question that keeps coming up on excess returns is this idea of how corporations are thinking of it. If you're spending on capex in AI in other places you're cutting somewhere else. Different companies are reallocating right now. I want to land us here because this is now the combination of all this data, at least in my mind and this is with financial planning hat on. This is when I go through yet another review of taxes and spending and budgeting in an exercise and we go, here's another person who's grocery and basic income expenses, whatever. It didn't go up 3% or 5% this year. It went up 50 or 60%. Yeah. Seeing it across the board, luckily a lot of the people we care about are at the upper arm of the K, not the lower leg of the K. Part of why bridging this makes so much sense. I want to talk about the single income stress test precarity line, shout out Adam Butler, Mike Green, the people have done work on this. You as well. Let's talk about the single income stress test because I think if this economy is running on consumption, this is why this important, this data is so important in understanding what the knock on effects to everything else are.

[A]

Yeah, and this also goes back to the home health care aid thing. You know, it's what I was really interested in. I took that same OEWS database and this is one of the first visualizations I put together and published on data for the people is I was really interested in connecting the wage that the median wage for one worker. So we're not talking like now there's other data in the visualization we first started talking about where you can look at people that have two jobs and that's at an all time high. Which this visualization you just asked me about will explain why. So if you want to just work one job, this to me is where this whole America's greatness kind of went sideways and how it was kind of weaponized I think back to the unions and I wasn't alive during this period of time but my understanding of this period of time was like you could actually have a middle class to upper middle class lifestyle, buy your own home, put your kids through college on one income. And it didn't have to be like I'm the president of an investment bank income. It could be like, you know, I'm like a senior person working at a factory and I can do that. You know that is a great thing. If you can do that, it's great. I mean look, maybe your Wife, or maybe it's the wife that works. But I mean, I can tell you I've been in that situation where like my wife only worked when I was not bringing income when I'm starting, Dave, for the people. And I, my kids benefit a lot from that. Like having a parent home, having a parent present for them to take care of them, rather than just putting them in daycare, like, you know, putting them in front of a tablet. Like there are real, call it, you know, qualitative benefits that, that you get from that. So I look back and I'm like, all right, well, what if you wanted to do that now? What if you wanted to earn a median income in the city in which you live? Dayton, Ohio, let's say. Or I don't know, say Philadelphia is close to you. I don't know. So you want to.

[B]

It's an upgrade.

[A]

You want to. Yeah. You want to own or you want to earn a single income and how far above the poverty line to raise a family of four are you? It's a model, right? Again, all models are wrong. Some are useful. I find this one really useful because it's simulation. It's not saying this is the number of people because most families have two earners now, right? But what if you didn't want that? Like you wanted those old days of where you could do. So the answer, if you actually look through all of them, one, it shows you the gradient. So the more green you are, the higher above the poverty line you are. And then I also give you a little kind of. It's called a parameter entry box. So you can actually say, Whoops, I have $5,000. A surprise expense. Whoops, I have $10,000. I had a healthcare bill. I had this. Like you could just plug that in and say, we'll take the poverty line and add that amount of money to it. Now how far above. So like how much breathing room do I have? And what you find is if you put $15,000 in there, the whole country turns red. So it doesn't matter where you live. The, the whole country. The median income in those Cities. Using the 2024 data, which is what I was using at that time, they only have $15,000 of breathing room. And that's pre tax too. So it's probably even less than that. Before they are living under the poverty line to support a family of four. And so that's why they have to work two jobs. We can start understanding more about the lived experience of people in the median. Now go and switch that from median to 25th percentile or even to 10th percentile. Everything is blood red. I mean, to actually like, this is the thing that's beautiful about data for the people is like when I did that, like, I mean I, I felt it like when I switched it and I'm like, wait a second, 10% of all the people, 10%. This isn't like a small percentage of all the people living in the city make under like $16,000 a year. 17,000, 20,000. It's. It's just a, it's like a hole in my heart almost that it's like, well, what if that were me? Like, I mean, I can't even envision like, I mean like my wife that works, a good government job. And like we really were struggling, like to kind of over the months of not having no income until I got this fellowship gratefully to o' Shaughnessy Ventures, we were really, really, you know, budgeting aggressively and cutting back on things and just to make things meet on her very good salary. And yet there are people that are making $20,000 a year, $30,000 a year. Like, I can't even like fathom how that must feel. But I can start to, I can start to. When I look at this data and I really try to get myself out of my own privilege problems and start to see through the lens of other people. And that's what I'm really. Look, there's a lot of investing takeaways. So many if you still believe that the economy has anything to do with the market. Big qualifier, big asterisk there.

[B]

But I mean, take us here because I think this is a great place for us to put the bow in this conversation.

[A]

Yeah.

[B]

The work with data for the people.

[A]

Yeah.

[B]

About to be officially a non profit. Congratulations on making that move. Another reason, check out the site. You can support work like this because there are direct market implications of understanding these data sets. But I also know there's a deep personal connection that you feel. I feel this too, where you think about, you think about your kids. You think about your kids. Kids. And you say this is not a sustainable framework to build the next several generations on because we're not using the data that's at our disposal.

[A]

Yeah.

[B]

Talk to me about the mission for the company going forward.

[A]

Yeah. So, yeah, thanks for the shout out. I literally am just submitting the right now simply because it costs. I was in an LLC structure with a DBA doing business for data for the people. I now, you know, I'm getting, I'm filing for an Ohio Nonprofit corporation. I'm hoping that will be done next week. Transferring the name over or giving rights to use the name. And then, yeah, then we'll be going through the Form 1023 process, getting the 5, the 501C3 status. And, and yeah, and I will keep people updated on how that goes because this is not a substack model. There is no paywall. There never will be a paywall in our work. I want everybody to have access to this, and I don't want to have to keep pestering you about, you know, paying money for this. I refuse to do that. But ultimately, our work will be supported by those that see value in it and can afford to do it. Like, if you can afford $10 a month, it's totally up to you. I don't want to know your financials. But, like, that goes a long way to us. You know, if I, even if I had like 100 people that could afford $10 a month, I mean, it's a meaningful amount of money for, for us. And, and I understand that it's prohibitive right now because I'm not a 501C3,

[B]

but I will be.

[A]

So right now, if you look at the website, we're publishing at a, at a frenetic pace. Thanks.

[B]

It's a good word for it.

[A]

Frenetic is inaccurate. The magic of AI. You know, I have figured out how to play nice with AI and how to use it to do things that I never could have envisioned doing. Like the, the data visualization that we talked about to start. Impossible to do without AI not with my skill set. I'm not smart enough to do it. And that's because when you actually go into the raw database that lives underneath the unemployment data, there are at least 30 different dimensions with intersections of every single dimension. And many of those are blank. And so when I get the intersection of sex and age, that will be populated. But then if I try to intersect that again with another dimension, say, race, all of a sudden it's blank. I spent days and days trying to wrap my head around it until, like, you know, my head was spinning and I gave up. And so I, I, I focused a lot on nativity because I think there's amazing stories about immigrants versus native born. Look at the data visualization. I've written about this at length, but that's about all the scraping I was able to do on my own. And then I, you know, we have all this wonderful access to Anthropic's Fable for another couple weeks or another week, and I basically just went to Fable and I'm like, do this all for me. And it did. Of course it takes tremendous amount of editorial oversight to make sure it's doing it right and testing the code and everything, which fortunately I'm skilled and I know how to do this. But my gosh, the amount of crank turning that it did and the amount of testing and connections and then even specifying that I only want you to present these dimensions when they're available and be smart enough to know to default to the age group of 16 plus when you're looking at the combination of race and sex. And it just figured it all out. So that kind of stuff has allowed me to take what would have taken me about two weeks and do it in two hours. Now that's why I'm able to publish every single day a new original data research study. Yesterday or this morning's piece, I actually for the first time went down into the census microdata, which if you've heard about that, don't go anywhere near it until, unless you're using AI because it is, you know, you'll just be committed to a loony bin, impossible to decipher what's going on. But if you want the intersection of every single dimension, microdata, that's where you get it. And I was able to unearth some amazing findings that nobody had ever looked at that America is far older than we thought. You know, there was a latest, there was a, a change in the way that the models were measuring the age of people and we lost about 2 and a half million prime age workers. So that's a big deal. You know, it's a big deal for Social Security for all these things. I was able to figure that using an ad. So long story short, like we're doing a ton of data journalism right now, but I archive everything, I do every chat I have and everything that we're doing is being used to train an LLM on. It's not just me, there are other people that are doing this. There's Jonathan, there's other data fellows. All of our interactions is being used to train a special purpose built LLM to understand how to triage problem solving within data storytelling and data analysis and data visualization. So how do you basically create a digital twin of us then on the other side? Shout out to Amanda Sinton, who is a founding data architect. She is brilliant and she's building the entire architecture for how do we rescue data sources, how do we publish them on GitHub in a way that everybody can use for free. And then how do we integrate that into an AI intake agent that anyone can go to? And there are many reasons why AI. I call it the Everything Store AI, which is what all Claude is Gemini and everything. Many reasons why it can't do this. But for people that. And I really think local journalists, community advocates, just interested people that have problems that they want to solve, it could be investors too. If you really want a research assistant that is trained on the art of working with data and helping you kind of problem solve on how to get to the scope of something that actually is achievable to solve and then can it create charts and data visualizations in a story and all the narrative for then you to edit. I don't think this should be an AI only tool. There always needs to be the human editorial oversight because we are the source of creativity and storytelling. Right. But I think we can get to doing 80 to 90% of the work for you with an AI bot. And so that is all being built out. So right now we have Amanda working on one side building the infrastructure for this platform and me feverishly creating as much content as I possibly can do, solving every question or answering every question that I've ever had by using AI in this responsible way and then taking myself out of being the crank turner to being the, call it the idea generator. I'm always generating the idea. I'm almost always creating the stream of consciousness piece, the writing on it. Like I do the data work with Claude. With AI, I'm like, here's the story that I see. And then I am spending the vast majority of my time in an editorial role. Oversight Claude is really bad at. Sometimes it wants to go too far with the claim. Sometimes it doesn't go far enough. Sometimes it will make one little statement that you're like, you're going to torpedo the entire credibility of the entire piece based on that statement. So you have to be very careful on an editorial basis, which is the stuff that like I'm really learning looking to build that into the LLM so it doesn't go anywhere near that. Which if you just go create something right now, it will do that. And so it'll run into that wall. Yeah. So like we're, we're just learning. We're learning a lot about what it's really good at and where it is good. Oh my gosh, it is like, it is like universe changing good. And then where it's not, it can basically destroy all the work that it's good at. So it's like, how do you build that infrastructure that takes the best of it and then really filters it? So if you go and you ask a question that is clearly profit seeking, we're going to direct you to go to Gemini. That's not what we're for. The problem ultimately, and I'm not saying it's not for investors, but it's for people that care about building as robust and as strong of an America as possible for all of us. Investing can overlap with that. Sometimes it doesn't. And so we'll know how to ferret out the questions in which way you're going with it. And then we'll either choose to help you, or we'll be like, you know what? This is not what our free service is for. That's what we're looking to build.

[B]

Eric, insightful as always. One more time, just tell the people where to find it. Tell them where they can bug you on the Internet.

[A]

Yeah. So the best place to find me on my website, data for the people dot com. Number four, not F O R. We're moving up in all the search rankings. So you could just search for data for the people will come up. We should be number one on that list. So you can do that. If you forget, email me, it's ericataforthepeople.com and sign up for the distribution list. I'd love to hear feedback. I'm getting a lot more engagement, especially from this crowd. I love it, you know, when people tell me, good or bad, you know, be critical with what I'm writing. I love that. I want to get engaged on the debates that are coming out of this data. I put a lot of my own bias into my interpretation of what I'm seeing. I will not allow any of my bias into how the data is worked. All the methodology is free. You're free to recreate your tools and you're free to come up with completely different conclusions than me, I believe. I mean, our vision ultimately is to create a shared basis of reality for America and for the world. And data can do that if we choose to use it for good. But that's not to say that we can't have polar opposite viewpoints. We are allowed to have that. That's what makes America beautiful. But we can't have our own sets of data. We just can't get by with that because then we actually can never see each other's perspectives. So if we can be any part of building that foundation and if you can be any part with supporting us, especially once we're a 501C3 just following our work, telling people about it, right? I'm not on social media that much, although we're trying to automate the entire process to shove out a whole bunch of things so you'll see more posts coming from Data for the People. But please, if you value our work, let people know. You know, spread the word. It's all organic, it's all word of mouth. I'm really relying on you to do that.

[B]

Spread the Word data for the people.com Eric, thanks so much for joining me today.

[A]

Yeah, thanks Matt.

[B]

Excess returns in all the places. You know what to do. Like comment, subscribe, send Eric an email, all the things below and we're out. Thank you for tuning into this episode. If you found this discussion interesting and valuable, please subscribe on your favorite audio platform or on YouTube. You can also follow all the podcasts in the Excess returns [e-mail] if you have any feedback or questions, you can contact [e-mail] no information on this podcast should be construed as investment advice. Securities discussed in the podcast may be holdings of the firms of the hosts or their clients.