Because the National Compensation Survey data is a model based on sampling, not actual collected data, like the census. That's why they call their database "Modeled Wage Estimates". Estimates and models are not useful in actual analysis.
I have a feeling you got this from a news broadcast, or a Bernie Sanders talking point...so I feel it is my duty to actually do the math to show you how stupid (and incorrect) this argument is. Let's look at US Census Bureau figures, shall we?
The middle 20% of wage earners earns between $45,000 and $69,999. This places them in the 25% lowered to 22% tax bracket, an average annual tax savings of $1350. A weekly tax break of $25.96.
The top 20% earns between $115,000 and $500,000+ (the highest bracket). This places them in 4 different tax brackets:
$115,000-$157,500 at 28% dropped to 24%, $4600 tax break ($88 per week)
$157,500-$200,000 at 33% dropped to 32%, $1575 tax break ($30 per week)
$200,000-$500,000 at 33-35% leveled off at 35%, $0 tax break
$500,000+ 39% and is now 37%, $10000 tax break ($192 per week)
Congratulations, every single figure in your meme was wrong. What else can we expect from the left leaning Tax Policy Center (a joint project of the left-leaning Brookings Institution and the Urban Institute). Neither your meme nor the Tax Policy Center acknowledge the complexity of the tax code and that the percentages given do not reflect a single tax bracket. Why not make some statement about the lowest 20% getting no tax break at all? The reason is obvious though, because of earned income credits and other tax incentives, the lowest bracket received more in refunds than they pay in taxes, and that would mess up your cute little jab.
Those at the top pay 1.5 times more in taxes every month, than the middle 20% pays all year.
Do you know who should be pissed? Those in the $200,000-500,000 tax bracket saw a 2% tax INCREASE, and they're part of your 20% top wage earners.
More proof that the left can't meme.....or math, apparently.