A few days ago, I posted about the history of redlining,  the system of government-operated housing discrimination which both manufactured ghettos and systematically robbed the black community from 1937 until 1968. (Unofficial redlining continued well beyond that)
I mentioned in that article that this had been part of a system of extracting wealth from the nascent black middle class and moving it to white hands, but the full scale of the effect may not have been obvious. Fortunately, just yesterday Rakesh Kochar and Richard Fry of the Pew Research Center published a short data analysis of wealth inequality by race  which helps see what happened here.
The most important thing to remember is that income
is not the same as wealth.
Wealth is important, because it represents money you have access to in case of an emergency: you could, for example, take out a HELOC or mortgage against your house. As many threads have discussed before, "how much money you could get in an emergency" is probably the single best way to measure people's real financial security, because it's the metric most closely tied to the way in which money affects your life.
For example, if you have an unexpected car breakdown that costs $500 to fix, and you have $500, it's a pain in the ass but ultimately not a big deal. If you don't have $500, then you have to take out some kind of loan which imposes a long-term cost – i.e., not a loan you can pay off quickly. That long-term cost, in turn, reduces your ability to handle the next
emergency, so you'll end up needing more such loans, etc., etc. Generally, a measure of how financially stable you are is how big of a financial shock you could take before getting dragged into such a death spiral.
For those of you who have been following the Ferguson case, or similar ones across the country, one of the key ways in which the local police routinely extorted the population was to set up traffic fines or similar "minor" costs which had various extra fees attached, up to the point where the average black resident could no longer afford them – which would drag people into debt spirals. 
This has a far more sinister history: after the end of the Civil War, a "New Slavery" system was instituted based on that interesting loophole in the 13th Amendment which allowed servitude "as punishment for a crime." Black people would be arrested on trivial charges or on none at all ("Vagrancy" was a particularly popular one for years), and fined a nominal amount ($5) plus decidedly non
-nominal "court fees." These would tend to total between $80 and $120, far more than a typical black farmer or worker would have on hand in the 1870's. (These numbers increased with inflation, of course) If the person couldn't pay on the spot, a (white) business representative would then show up and offer to pay off their debt in exchange for them working at their factory, farm, or mine, and the victim would be packaged off and sent to hard labor. Often, this entire negotiation happened without the people arrested even being involved; they were simply rounded up, told they had been convicted, and sent off to mines, where the conditions were indistinguishable from slavery, down to the whips, the lack of food, and the overseers.
This was no coincidence: it became a major economic engine of not just the South, but soon afterwards the North as well, in the period from 1865 to 1943. 
In 1934, one of the key programs of the New Deal, the Federal Housing Administration, was created to help Americans buy and keep homes, and in doing so build a reserve of personal wealth. Its charter explicitly included racial segregation: its job was to help white
Americans, and white Americans only. The redlining system was created and operated by the FHA.
But it was far from the only such program. The key programs which created the great American middle class in the postwar period – the GI Bill, the various housing programs, and the mortgage tax exemption that went with them  – explicitly excluded nonwhite citizens. So the entire boom that moved people from Depression-era poverty to 1950's era comfort, to the wealth of the present day, was segregated and available only to some Americans.
And what's the result? The graph below shows the median net worth of American households over the past 30 years. The line at the top shows white families; you can see a steady rise over the 1990's and early 2000's, and a major hit in 2008 thanks to the Great Recession.
The two tiny lines at the bottom are black and Hispanic families. Not only do these curves show none of that big rise over the 1990's, they are literally an order of magnitude
lower. That difference is hugely tied to differences in home ownership rates, but quite generally, the fraction of white Americans who have some sort of money in the bank is hugely greater than the fraction for non-white Americans.
That's the impact of not just 300 years of slavery (which created a large population that was freed with nothing but the clothes on their backs), but 150 years of systematic plunder afterwards: each financial gain made by black families was quickly taken away, through everything from the shady loans created in the shadow of redlining to systems like Ferguson's "traffic courts."
Remember the basic rule: it takes money to make money. That's not just true when you're investing millions of dollars; if you don't have the money to get a nice set of clothes, you can't interview as well. If you're homeless, and can't even receive mail? You're screwed. If you can get a job, but the first car breakdown or failure of day care gives you a choice between losing the job and going into a debt spiral, then what happens with the second one?
That's the underlying mechanism behind the graph below. The trick is that, when people are already poor, taking a small amount from them has a much bigger impact, and keeps them from ceasing to be poor.
One way to think of it is this: Imagine we have two people, Alice and Bob, who each get the same income – say, $100 per month after expenses. Alice has $100 in the bank; Bob has $1,000. Each month, there's a 50% chance of an unexpected $250 expense.
Say that in the first month, they both crap out and get hit with an expense. Alice is now $50 in debt ($100 + $100 - $250), and Bob has $850. The next month everything goes fine, but Alice has to pay $5 in interest, so she ends the month with $45, while Bob has $950. The third month, they're hit again; Alice now has $105 in debt, and Bob has $800. Now we have another good month; Alice is now $15 in debt ($10 interest) and Bob is at $900.
But this isn't really realistic: the problem isn't interest so much as foregone opportunity. Say that this next month, they both have a chance to invest: maybe to get an apartment with better rent but an up-front deposit. They would each have the opportunity to pay $200 now, and then get an extra $50 per month. What a deal!
Except Alice needs the money to buy food, because she doesn't have access to unlimited credit. (Or you can think of their wealth as including access to credit, instead of just dollars in the bank) So at the end of this month, Alice is $45 in debt and makes $100 per month, while Bob has $700 and makes $150 per month.
You can guess what's going to happen next: that extra $150 per month will add up, and Bob will have opportunities to grab many other investments that Alice will have to pass up on by spending some of his buffer. Each investment increases Bob's average income, which in turn gives him more buffer, while Alice's buffer never grows.
Fast-forward that 50 years, and you have the graph below.
. My thanks to Richard Fry for sending me the raw data, so that I could plot this on linear axes and show you three different racial groups stacked up together. The data here comes from the Federal Reserve's Survey of Consumer Finances, http://www.federalreserve.gov/econresdata/scf/scfindex.htm
 As detailed in the Justice Department's report, http://www.justice.gov/sites/default/files/opa/press-releases/attachments/2015/03/04/ferguson_police_department_report.pdf
; see pages 42ff in particular.
 I highly encourage you to read more about this. The best book (by far) on the subject is Douglas Blackmon's Slavery By Another Name, https://books.google.com/books/about/Slavery_by_Another_Name.html?id=AvTtagpo8QgC
, which is probably the single most useful book for understanding 20th-century American politics and economics out there.
for its history. This was actually created in the 1890's, but didn't affect a majority of Americans until programs like the FHA made home ownership and mortgages common in the 1950's.