> It was dropping for most of the 1990s, and most of the 2010s.
It was dropping in the 90s because of Clinton's enormous budget cuts and firing half a million federal employees which is clearly an outlier and something that now seems to be far more partisan and difficult to achieve. And in 2010s it was dropping relative to enormous stimulus spending and corporate bailouts aimed at helping the economy after the housing crisis.
> So it's been rising steadily over the past 45 years...except for about two decades of that time? This isn't what most people call a steady rise.
You can quote me correctly instead of making up a weak strawman. The entire thread is here for everybody to read, you're not going to be able to fabricate some "gotcha". I said it has been rising steadily for a century, and that recent trend is more shaky but probably still trending upward.
The comment I replied to which is easy to find is this:
> 1. There are claims that federal spending is out of control. How do you square that with the fact that spending as a percentage of GDP is only slightly elevated compared to the historical average going back to at least the 1970s, with the main deviation in the past few years coming from the after-effects of the pandemic? [1]
Most certainly since 1970 the trend (using a linear regression) has been very significantly upwards, which contradicts that claim.
Since the 1980 the trend has been upwards.
Since 1990 the trend has been upwards.
Since 2000 the trend has been upwards.
Since 2010 the trend has been upwards.
Now just give it a rest for god's sake. Don't keep doing the reddit internet arguing thing.
If the list of data points was 10 98 times, then a 50, then another 10, it would give you the same results with those linear regressions. You'd see a "trend" going up when you measure from anywhere in the past to the present. But that trend would not represent anything real about the first 98 years. That is not the correct test to use.
> Excuse me, I didn't latch onto 1980, crazygringgo nominated that date.
I said "latched onto" specifically because someone else said it first. Though they said "about 1980".
> Sure unless you choose a different end point, right?
It's only different if you choose the two covid years. There are several years before and one year after that are all lower than the 1990-ish zone.
> A linear regression from 1990 to 2022 has it trending upward too.
Because of 2020 and 2021. Unless you expect another anomaly like that to happen in the next handful of years, it's obscuring the real trend.
The numbers are slowly climbing over the last decade, but after a big period of decline in the first part of the last 40 years, and we're still below where it used to be.
> I said "latched onto" specifically because someone else said it first. Though they said "about 1980".
Wrong, try actually comprehending the thread. I was being generous, they said "mostly until 1970". 1980 is higher than most of the 70s. I could have taken an average of every year in the 70s, and 1980 is 1.5% higher than that. So if I did that it would have shown nearly 2x the rise.
> It's only different if you choose the two covid years. There are several years before and one year after that are all lower than the 1990-ish zone.
So it's different if you choose something different as you did which is okay, but apparently if I do it it's wrong.
> Because of 2020 and 2021. Unless you expect another anomaly like that to happen in the next handful of years, it's obscuring the real trend.
And the 90s were low because of Clinton's historic expenditure reduction, which is an outlier too. You want me to remove my outlier so your outlier can look better by comparison?
> The numbers are slowly climbing over the last decade, but after a big period of decline in the first part of the last 40 years, and we're still below where it used to be.
The numbers have steadily been increasing over a century, and even in the past 45 years, certainly from the 1970s, they have been rising. Exactly as I said in my original post.
They were going down for an entire decade. You can't remove an entire decade and then say "steadily rising" even if you have good evidence it was an outlier.
The numbers also went down 6 years in a row in the wake of the 2008 spike.
And if you don't treat covid as an outlier, then the numbers went down massively from 2021 to 2022! Treating 2020-2021 as normal years works against your argument more than it works against mine.
> They were going down for an entire decade. You can't remove an entire decade and then say "steadily rising".
I certainly can, over a century.
> The numbers also went down 5 years in a row in the wake of the 2008 spike.
Yes, that's how long it took to reduce spending and it still didn't bottom out at pre-2008 levels. This all contributes to the trend of increasing expenditure relative to GDP.
> And if you don't treat covid as an outlier, then the numbers went down massively from 2021 to 2022! Treating 2020-2021 as normal years works against your argument more than it works against mine.
No it doesn't. It still didn't drop to pre-covid levels.
> Government expenditure as a proportion of GDP has been rising steadily for a century and is now approaching 40%. From IMF: https://www.imf.org/external/datamapper/exp@FPP/USA. The recent trend is shaky due to Covid, housing crash, maybe dot com crash, etc., but it looks like it's probably still trending upward.
Is true.
> The two long series of dropping values make up more than a third of the last 40 years.
The trend over the past 40 years is rising.
> Good thing nobody claimed that.
You tried to because you disingenuously tried to say that the drop would "disprove" my argument. You don't even understand what my argument is, and fixating on nitpicking the definition of "steadily" is just hilariously insane. Like, we can see WWI and WWII in those figures, I'm not trying some weasel word sleight of hand to fool people into believing I meant monotonically increasing, as though the actual increase doesn't prove my point in substance.
The point I replied to said this:
> 1. There are claims that federal spending is out of control. How do you square that with the fact that spending as a percentage of GDP is only slightly elevated compared to the historical average going back to at least the 1970s, with the main deviation in the past few years coming from the after-effects of the pandemic?
At least the 1970s. Every one of the full 4 decades after the 1970s was substantially higher than the 1970s, taken as an average (which is slightly weird but it'll do). And not just slightly elevated, the next lowest decade is more than 6.25% above the 70s, and the rest are 10-11% above.
Getting angry about the numbers is really just clutching at straws because you don't seem to like the conclusion. Try arguing with some facts you actually have on your side if you don't like it so much.
> You tried to because you disingenuously tried to say that the drop would "disprove" my argument.
"It went down 19%" disproves your argument if you don't treat it as an outlier.
It doesn't matter how that compares to 2019 in particular.
I don't understand why you think there's anything disingenuous about that.
> Like, we can see WWI and WWII in those figures, I'm not trying some weasel word sleight of hand to fool people into believing I meant monotonically increasing, as though the actual increase doesn't prove my point in substance.
I assumed we were treating those as outliers. Hence my confusion at not treating covid as an outlier.
> Every one of the full 4 decades after the 1970s was substantially higher than the 1970s, taken as an average (which is slightly weird but it'll do). And not just slightly elevated, the next lowest decade is more than 6.25% above the 70s, and the rest are 10-11% above.
But a lot of the interesting points don't align well with decades.
If we offset by exactly half a decade, to get a different view but in the least cherry-picky way possible, 1975-1984 is 34.6 average, 1985-1994 is 37.3 average, and 1995-2004 is 34.6 again.
Then 2005-2014 goes up again, but only to 36.8, less than the peak.
Seems to me, the person is claiming that a linear trendline increasing causes "steadily" to be an appropriate proxy for "values went up and down some, but overall went up".
If we can agree that the trendline is linear and that it goes up, then yes, we can agree on "steadily".
However, if we observe that the actual data goes both up and down over a specified period, then no, the data cannot be characterised as "steady".
For those interested, perhaps I could suggest drawing a process capability chart? The sigma variation lines would seem to clearly demarc outliers.
It really is.
> It's funny that you latched onto the mention of the year 1980,
Excuse me, I didn't latch onto 1980, crazygringgo nominated that date.
> because if you used 1982 (or 1983, 1984, 1985...) you'd see that it has gone down.
Sure unless you choose a different end point, right?
> The most recent measurement is lower than it was in the years around 1990, and the measurements before covid hit were also lower than those years.
A linear regression from 1990 to 2022 has it trending upward too.