Let's have a little talk about the median and how to lie...¶
Let's have a little talk about the median and how to lie with statistics!
Median is commonly cited as a useful measure of general economic prosperity, because it is very simple: the figure smack in the middle of the list, unlike the mean, which averages everything together. https://t.co/UaRST6IAI2 For example, if you have an economy where the top 10% are ASTONISHINGLY wealthy compared to everyone else, their presence skews the mean:
10 12 13 14 16 17 84
The mean in this example would be 24, which is way higher than everyone except the outlier. Not helpful. Looking at the median gives us a much better approximation of where most people are. In the above example, the median is 14, which is in the middle of where the majority of the samples are. More helpful.
(In this case, it's actually the mean of the first six figures.) Let's look at a different example:
7 9 10 15 18 20 96
Hey, look at that! The median went up from 14 to 15! So did the mean, which went from 24 to 25. The economy is growing by every metric!
Oh wait.
Take a closer look at those first three numbers. They all went down. In fact, if we take away that huge outlier, the mean of the "normal people" actually went down, to about 13.
This is the problem with using median as a measure of a dynamic system, especially one with a high degree of internal variation and no exclusion of outliers. In fact, those numbers above the median could skyrocket to mind-boggling heights, and the bottom percentiles could plummet to 0, and that movement would do nothing to affect the median:
0 0 0 15 18 20 151
Median: 15 Mean: 51 People with absolutely nothing: 43% In the same report cited by OP, we see a revealing table (Table B.1). They tease apart a little bit from these aggregate figures, and show that, at the 25th percentile, net worth was 27,100 in 2022. The median net worth was over SEVEN TIMES that.
The mean? OVER A MILLION DOLLARS The report keeps citing "some increases in inequality," which is both putting it mildly (growth was 3x higher in the top 20% than in the bottom 20%) and ignores the sheer scale of the existing inequality.
Year after year, already-high inequality ticks higher and higher. The economic benchmark that matters to people in the lower 40% -- below the median, but still a HUGE chunk of the population -- is not the growth of the median, the mean, or even their own "growth," relative to abstract figures. It's the steady bleed of their economic position. It cannot be captured in snapshots of static figures, but must be tracked dynamically, with the full context of both time and the broader economy: "how the other half lives."
You will often see figures showing growth across all percentiles, adjusted for inflation... ... but the metrics used to "adjust for inflation" AGAIN heavily weight the consumption habits of those higher percentiles.
We can illustrate this with another simplified example, comparing expenses and income: Expenses Income 10 -> 14 11 -> 13 15 -> 17 16 -> 19 20 -> 23 24 -> 31
We can clearly see that the bottom third in this example went from a slight surplus to a net deficit. Meanwhile, the median expenses rose 13%, while the median income rose 18.8%.
The mean expenses went up 20%, while mean income went up 23.5%
When looking at both median and mean, income rose faster than inflation.
But for a full THIRD of the population, the opposite is true. This is an exaggerated, oversimplified example to help illustrate the way statistical measures don't grasp the full picture.
In real life, a third of the population hasn't suddenly gone from being able to save a little each month to not being able to pay their bills at all. But many HAVE gone from having a moderate surplus to very little, or none, or even started going into deficit. Sometimes temporarily, sometimes long-term. Even grouping into quintiles washes out the granularity within that bottom 20 or 40%. There's constant movement up and down. There is no existing mainstream economic measure of these dynamics, or the way in which individuals or entire cohorts can see some movement upward without that change compensating for longstanding downward trajectories. Relative to inflation, many are doing better. For some, that increase in prospects has even begun to overcome past downward mobility.
And yet, for a solid chunk of the population, prospects have either not improved enough to dig them out of the hole, or have gotten even worse. If the top third is prosperous, the middle third is starting to stabilize, and the bottom third is in freefall, the economy is not "improving."
Even an economic system predicated on pushing an entire segment into abject poverty cannot sustain that segment being 30-40%. This is especially true when that system is built on the back of imperial extraction, where the vast majority of the exploitation is outsourced. The entire project relies on keeping the consequences of the inherent contradictions of capitalism as far from the center as possible. And yet they are failing at that. The economic sanctity of the core is crumbling, even as the periphery works to liberate itself. Attempts to compensate by ramping up oppression -- both through military and economic means -- only accelerate the organization of resistance. Protect the median all you want: the bottom grows more desperate, more hungry, and more organized all the same.