[part 1]

Tried getting this going with google gemini AI. It had an error right off the hop, I guess trying to do charts from data it would find. I don't know. I then tried with chat GPT because it was working not too bad with stuff the other week. But it taps out pretty fast where they want you to pay. Whatever.. got some data here. Any pics of charts, I'll upload them all to an imgur album. ok here you go.. the charts it doodled. You got 7 charts as we were going along.

https://imgur.com/a/NWOb3V4

This was the chat process, how I wound up with this stuff.

==============================

Can you get a graph of the cost of an average house with a garage, since like, I don't know.. as far back as you can go. Probably be around like after WW2. Then do inflation adjusted so you get an idea how much it costs compared to what an average person would make. You also have locations, coasts seems to be more. So I don't know how you can do an average.. the coasts seems ridiculous. It's been going up though a fair bit in central areas. See what you can do for charts.

Thought for 43s

Thanks — this is a great (big) project and I can do it — quick question before I dive in so I don’t waste time: do you want the analysis for the United States (national + coasts vs. interior) or for Canada (or both)?

If you don’t care / don’t reply, I'll assume United States and will:

Pull long-run national house-price series (Census median home values for 1940–2000 decennial, Census/HUD/NAR and FRED series for 1963→present) and an all-transactions house-price index (FHFA or Case-Shiller) to extend coverage.

Pull median household income (Census / FRED) for the same periods.

Pull CPI (CPI-U) so I can inflation-adjust all prices to real (2024) dollars.

Compute and plot:

Nominal median house price (long-run) and inflation-adjusted series.

Median house price ÷ median household income (price-to-income ratio) over time — this shows how many years of median income it takes to buy a median house.

A comparable chart by broad region: coasts (Northeast & West) vs. Midwest/South (using Census regions or FHFA regional indices) so you can see the coastal divergence you mentioned.

Short notes on data definitions/limitations (new vs. existing homes, owner-occupied value vs. sale price, garage presence caveat), and citations to the data sources (Census, FRED, FHFA, NAR).

Reply “US” to confirm that’s fine, or say “Canada” (or “Both”) and I’ll proceed. If I get no reply, I’ll proceed with the U.S. and produce the charts and dataset. Which do you want?

We'll stick with the US for now.

Done — I pulled public series from FRED and the Census and made three charts plus a small summary table.

What I used (sources):

Median sales price of houses sold (MSPUS) — quarterly, U.S. Census/HUD via FRED (series MSPUS). FRED +1

Consumer Price Index, all items, U.S. city average (CPIAUCNS) — monthly CPI via FRED (used Jan-of-year values to inflation-adjust to 2024 dollars). FRED

Median household income (MEHOINUSA646N) — annual, Census via FRED (1984–2024). FRED

Census decennial median home values (unadjusted) for 1940–2000 (decennial datapoints) from the Census historical table. Census

Notes on methodology and limitations (important):

You asked about “houses with a garage” specifically — there is no reliable long-run national time series that isolates only houses with garages. I used median home prices / median sales prices as a close proxy for “typical house” nationally. If you want a garage-specific series we’d need a specialized dataset (not available historically at the national level).

MSPUS is median sales price of new houses sold (quarterly), while the Census decennial table is median value of owner-occupied housing. They measure different things (sales price vs. stock value) and use different populations (new sales vs. all owner-occupied). I combined them carefully but flagged the difference in the chart and in the write-up.

Inflation adjustment: I converted nominal dollars to 2024 dollars using Jan CPI values from the CPI series (so all “real” numbers are expressed in 2024 USD). That makes house prices comparable across decades.

Income: median household income series from FRED (annual) was converted to 2024 USD using the same CPI basis so the comparisons (price vs income) are apples-to-apples.

For annual house prices I used the first-quarter (Q1) MSPUS observation for each year when available; for years prior to 1963 I used Census decennial medians (1940/50/60) and interpolated between those decennial points to create a continuous visual — the interpolation is only to help visualize trends and is not a claim of exact annual values in 1941–1959.

The affordability ratio plotted is (median house price in 2024 USD) / (median household income in 2024 USD). Higher ratio = houses cost more years of the median household income (less affordable).

What I produced (files & visuals):

Three charts shown inline:

Median house price (nominal series) and the same series inflation-adjusted to 2024 USD (with census decennial points highlighted).

Real (2024 USD) median house price vs. real median household income (2024 USD) on the same axes.

Affordability ratio: median house price / median household income (both in 2024 USD) by year, with a reference ratio = 3 shown.

A small summary table with key years (1940, 1950, 1960, 1970, 1980, 1990, 2000, 2010, 2020, 2024) giving nominal median house price, median house price in 2024 USD, median household income in 2024 USD, and the house/income ratio for those years (displayed for you as an interactive table).

A combined PNG chart file saved: Download the combined chart

Quick takeaways (from the charts):

Real (inflation-adjusted to 2024 dollars) median house prices have risen substantially from mid-20th century levels. There are clear spikes around the mid-2000s (housing boom) and a strong rise again in the 2010s–2020s.

Median household income (real, 2024 USD) has risen much more modestly compared with house prices — meaning houses cost more multiples of median income now than in earlier decades.

The affordability ratio (house price ÷ median income) is visibly higher in recent decades than in much of the post-WWII period, illustrating the point you made about coastal/metro prices outpacing central areas — though the national median hides a lot of geographic variation.

If you want any of the following next, I can do it now (no waiting):

Replace MSPUS with FHFA All-Transactions HPI or Case-Shiller-based implied median prices (helps if you want an index that covers repeat-sales rather than new-home sales). FRED +1

Produce separate coast vs interior regional charts (Case-Shiller has city composites and FHFA has metro-level series — I can build coast vs midwest/south comparisons). FRED +1

Produce a downloadable CSV of the combined annual series (nominal & real house prices, income, CPI used).

Swap the inflation base year (e.g., convert to 2020 USD instead of 2024).

Which of those would you like me to do now? Or do you want refinements (e.g., show the house-price-to-income ratio for specific metros like Los Angeles, New York, Dallas)?

These are all inflation adjusted?

Good question — not all the lines are inflation-adjusted, but I’ll spell it out clearly for you:

✅ Chart 1 (two lines)

Blue (nominal): raw median sale prices (in their original year’s dollars, not adjusted).

Orange (real / inflation-adjusted): the same prices converted into 2024 dollars using the Consumer Price Index (CPI). So that chart lets you see both the raw rise and the inflation-corrected rise.

✅ Chart 2

Both lines (house prices and household incomes) are inflation-adjusted to 2024 USD. That way you can compare what’s really changed in purchasing power — it’s a “real-terms” comparison.

✅ Chart 3 (affordability ratio)

The ratio uses both series in 2024 USD, so inflation cancels out. That line shows how many years of median income it takes to buy a median-priced house, with inflation already removed.

✅ Summary table

The “MedianHouse_2024USD” and “MedianIncome_2024USD” columns are both inflation-adjusted.

The “MedianHouse_Nominal” column is the original, unadjusted value.

So bottom line:

Everything that’s being compared (house vs. income, or ratio) is inflation-adjusted to 2024 dollars. Only the “nominal” line in the first chart shows unadjusted prices.

Would you like me to add one extra chart that shows the difference visually — e.g., nominal vs. real side by side with inflation levels shaded? It’s a good way to see how much of the rise is just inflation vs. real appreciation.

You said "Would you like me to add one extra chart that shows the difference visually — e.g., nominal vs. real side by side with inflation levels shaded? It’s a good way to see how much of the rise is just inflation vs. real appreciation." Yeah let's check that out.

Here’s the extra chart you asked for — it compares nominal vs. inflation-adjusted (2024 USD) median house prices, with the gray shaded band showing how much inflation had to be applied to convert each year’s dollars into 2024 dollars.

You can read it like this:

The blue line (nominal) shows what people actually paid in that year’s dollars.

The orange line (real) shows what that same price is worth today after accounting for inflation.

The gray shaded area shows how big an inflation adjustment was needed — e.g., a multiplier of about 15× for the 1950s, 10× for the 1970s, etc.

So the vertical distance between the blue and orange lines in early years is almost entirely inflation; the remaining upward movement in the orange line is true real appreciation.

Would you like me to overlay U.S. median income (real) on this same plot so you can directly see how home prices diverged from income over time, with inflation shaded behind it too?

You got anything in all this, of what an average person would get paid? Nevermind your "gross" because it's like a third they take in taxes. So all that matters is how much you get paid.

Perfect — you’re asking the real question: “What did people actually take home — not gross income — compared to house prices?”

That’s a super important distinction, because the “median household income” numbers I used earlier are gross before-tax income (from the Census). As you said, what matters is after-tax disposable income — what people could actually spend on housing.

Here’s how we can get there:

🧾 Step 1. Available data on after-tax (disposable) income

The U.S. Bureau of Economic Analysis (BEA) publishes a long-term series called “Real Disposable Personal Income per Capita” (A229RX0).

“Disposable” means after federal, state, and local taxes.

“Real” means already adjusted for inflation (to chained 2017 dollars, but we can scale it to 2024 USD to match the other charts).

It’s per person, not per household — but you can roughly multiply by ~2.5 for an average household to approximate a “household disposable income.”

🏠 Step 2. Compare that to inflation-adjusted home prices

We can compute:

Median house price / real disposable income per capita and also

Median house price / real disposable income per household (≈ per capita × 2.5)

That ratio will be lower than the one we saw before (since disposable income < gross income), but it gives a truer sense of affordability.

Would you like me to:

Pull the BEA series for Real Disposable Personal Income per Capita and add it to the dataset,

Adjust it to 2024 dollars,

Plot house price vs. after-tax disposable income (and a ratio line like before)?

That’ll show how many after-tax years it takes to buy a median house — which is probably the most realistic affordability measure.

Want me to go ahead and do that?

When you use the term "nominal".. that's like cheaper houses?

Excellent question — and it’s easy to mix up because “nominal” sounds like “small” or “cheap,” but it doesn’t mean that.

Let’s break it down simply:

🟦 Nominal = the actual dollar amount at the time

It’s what people literally paid in that year’s money.

It doesn’t account for inflation, so older “nominal” prices look cheaper only because the dollar used to buy more.

Example:

In 1955, the median U.S. house cost $11,000 (nominal).

That sounds cheap — but $11,000 back then bought about what $125,000–$130,000 buys today.

So “nominal” isn’t cheaper houses — it’s unadjusted money values that look smaller because of inflation.

🟧 Real (inflation-adjusted) = true purchasing power

Adjusts nominal prices for inflation (e.g., using the Consumer Price Index).

It tells you how expensive something really was in today’s money.

Example:

That same $11,000 house in 1955 ≈ $125,000 in 2024 dollars → that’s the real price.

So if you ever see “nominal vs. real”:

Nominal = what it cost then

Real = what that’s worth in today’s dollars

That’s why the nominal line in the chart looks lower — not because houses were literally “cheaper” in real terms, but because the dollar had a different value.

Would you like me to show a quick side-by-side table so you can see a few decades of nominal vs. real (2024 USD) house prices to make it intuitive?

[continued in part 2]