> It was odd to me, because it hadn't really occurred to me before that, given infinite memory (and within a mathematical framework), there's fundamentally not necessarily a difference between a "list" and a "function".
You don't even need infinite memory. If your function is over a limited domain like bool or u8 or an enum, very limited memory is enough.
However the big difference (in most languages) is that functions can take arbitrarily long. Array access either succeeds or fails quickly.
Check out https://en.wikipedia.org/wiki/The_Discovery_of_France to see how much better the Brits had it than the French. Many of the accounts in the book are drawn from British working class _tourists_. Ordinary people in Britain had enough disposable income to visit France.
Your graph is rather misleading: you adjust for inflation in two different ways to get 'real' data. One series uses CPI, the other uses the implicit price deflator for gross domestic product.
You can avoid this problem, by plotting nominal values and looking at their ratios. The price level will naturally cancel out.
This is one reason I graphed them as indexed values. You're not comparing the 'real' inflation adjusted values, but the independent indexed value, relative to a fixed point in time, for both.
That would only save your graph, if your two measures of inflation had a fixed ratio over time. They don't, as you can see on https://fred.stlouisfed.org/graph/?g=1QI7c . The CPI tends to grow faster than the GDP deflator, but not reliably so.
The growth of this ratio explains some of the widening gap between your two graphs over time.
Btw, https://fred.stlouisfed.org/series/LABSHPUSA156NRUG# is what you might actually be interested in: it's the labour share of GDP. As you can see it's rather stable. Most of the action is in the difference between mean vs median labour compensation.
Again these values are independent so what really matters is self consistency. They vary because they measure different things. For instance the CPI focuses on consumer goods and includes imported goods, which is extremely informative for wages. By contrast the GDP excludes imports which makes it informative for the overall size of a domestic economy. They cannot be directly compared, but their relative growth can be directly compared and is quite informative.
The reason your graph of labor compensation is stable is because it includes all forms of compensation for all 'workers'. So for instance Elon hitting his milestones will result in direct stock compensation in the ballpark of $1 trillion - those 22 million years of median income (not even wages) are then counted as 'labor compensation.' That's obviously an extreme example, but there are countless smaller scale examples that collectively distort the figure even more than Elon. Compensation, and even direct wages, at the top have skyrocketed while the majority of society has stagnated or declined. So we are set to become the first country with a trillionaire, and a country with 42 million people who can't afford food without government handouts. A tale of two cities?
In any case this is a big part of the reason why I think lifestyle descriptions are so much more useful than numbers. The stories of the boomer generation sound like fairy tales now a days - somebody putting themselves through college, buying the first car, and even having enough squirreled away for the down payment on their first house - all on the back of a part time job that didn't even require a college degree. It simply doesn't sound real, yet you can verify that it was fully viable by looking at wage and cost data at the time - it's real. But the fact that it doesn't sound real, even from a relatively very recent perspective - our own, rather emphasizes the point of stories being more useful than numbers.
> The primary thing that matters with indexed values is self consistency - their values are independent. The primary reason that these values are drifting is because median wages are in no way whatsoever keeping up with GDP growth. You yourself just, perhaps inadvertently, demonstrated such with the labor share of GDP graph.
> The reason that graph is so stable is because labor compensation includes all forms of compensation for all 'workers'. So for instance Elon hitting his milestones will result in direct stock compensation in the ballpark of $1 trillion - those 22 million years of median income (not even wages) are then counted as 'labor compensation.' That's obviously an extreme example, but there are countless smaller scale examples that collectively distort the figure even more than Elon.
> This is a big part of the reason why I think lifestyle descriptions are so much more useful than numbers. The stories of the boomer generation sound like fairy tales now a days - somebody putting themselves through college, buying the first car, and even having enough squirreled away for the down payment on their first house - all on the back of a part time job that didn't even require a college degree. It sounds impossible, but you can indeed work out the math from wage and cost data at the time (okay kind of backtracking on the numbers don't matter aren't I?). It's real. But the fact it sounds impossible goes a long way towards demonstrating the soundness of the data I've shown.
If you use price levels measures that grow at different rates, indexing to set them equal at one point in time doesn't save you.
You are weakening your own point by screwing up the stats: even after we correct for the difference in inflation measures, there _is_ a widening gap between mean and median worker income. (Where, yes, worker includes Elon Musk as CEO of Tesla.)
> This is a big part of the reason why I think lifestyle descriptions are so much more useful than numbers. The stories of the boomer generation sound like fairy tales now a days - somebody putting themselves through college, buying the first car, and even having enough squirreled away for the down payment on their first house - all on the back of a part time job that didn't even require a college degree.
Of course, the kind of crappy cars that boomers were driving aren't legal to buy anymore. Their houses were also much smaller, etc. It's a separate discussion of whether we should legalise crappy cars and small houses again. (I'm against the former but in favour of the latter.)
You'd want to correct for these quality differences to make your point stronger.
Also keep in mind that the boomer's "Golden Age" was the pinnacle of inequality. In the decades since, inequality has drastically shrunken. Mostly thanks to people in India and China moving from dirt poor and starving to merely poor (for India) and medium income (for China).
I would not want to go back to that supposed Golden Age, just because super-rich Americans were slightly less well off comparatively than today. Median and average Americans are still better off in absolute terms; and approximately everyone else on the globe is massively better off in absolute terms today.
You don't even need infinite memory. If your function is over a limited domain like bool or u8 or an enum, very limited memory is enough.
However the big difference (in most languages) is that functions can take arbitrarily long. Array access either succeeds or fails quickly.
reply