In the first example, here’s the original language from the CAP study:

According to the National Foundation for American Policy, immigrants will add a net of $611 billion to the Social Security system over the next 75 years. Immigrants are a key driver of keeping the Social Security Trust Fund solvent, and Stuart Anderson of the National Foundation for American Policy finds that cutting off immigration to the country would increase the size of the Social Security deficit by 31 percent over 50 years.

Here’s what Williams wrote in the original published version of his February 18 column, as found through a Google cache search:

According to the independent National Foundation for American Policy (NFAP), immigrants will contribute $611 billion to the Social Security system over the next 75 years. Indeed, immigrants are a key force in keeping the Social Security trust fund solvent for older Americans who are at or near retirement. NFAP also found that halting all immigration into the United Size [sic] would explode the size of the Social Security deficit by at least 31 percent over 50 years.

To the first sentence, Williams added the word “independent.” In the second sentence, he added “indeed” and turned “key driver” into “key force.” He split CAP’s second sentence into two, but the language is strikingly similar — “cutting off” becomes “halting” while “increase” becomes “explode.”

But that section of the story was dramatically rewritten and republished on March 2, with the editor’s note. Here’s how that paragraph is cast in the version of the column that appears online at The Hill now:

The CAP report makes the point. It cites a study on the impact of immigration reform by the independent National Foundation for American Policy (NFAP). That study finds immigrants will contribute $611 billion to the Social Security system in the next 75 years. Indeed, the arrival of newcomers is key to funding Social Security for older Americans who are at or near retirement. NFAP found that halting all immigration into the United States would explode the size of the Social Security deficit. They estimate an increase of at least 31 percent in that deficit in the next 50 years without continued immigration.

In this second example, we’ve italicized the borrowed wording:

CAP writes:

These big gains occur because legalized workers earn higher wages than undocumented workers, and they use those wages to buy things such as houses, cars, phones, and clothing. … Hinojosa-Ojeda found that the tax benefits alone from legalization would be between $4.5 billion and $5.4 billion in the first three years.

Here’s how the cached version of Williams’ column reads:

These big gains occur because legalized workers earn higher wages than undocumented workers, and they use those wages to buy things and stimulate the economy through commerce. Professor Hinojosa-Ojeda also calculated that the tax benefits alone from legalization would be between $4.5 billion and $5.4 billion in the first three years.

The revised version not only attributes that to the CAP reports, but the paragraph is rewritten to put the borrowed language in quotation marks. Again, the editor’s note mentions only the attribution problem and not the word-for-word echo that now appears in quotes.

In citing that study, CAP attributes the increase in GDP to gains that “occur because legalized workers earn higher wages than undocumented workers and they use those wages to buy” merchandise. That stimulates the economy through commerce.

A third section of the story also appears to have been borrowed and then changed.

In the CAP report:

Demographers Randy Capps, Michael Fix, and Everett Henderson, for example, compared welfare-participation rates of legal permanent resident immigrants to the native born. When controlling for income, they found that immigrants had similar—if not lower—participation rates than natives in the three main social programs: welfare, food stamps, and Medicaid. At the 200 percent poverty line—a common threshold for low-income households—32 percent of native families received food stamps, compared to 22 percent of naturalized-citizen immigrant families.

In Williams’ original column:

Jeffrey S. Passel and Michael E. Fix, two respected demographers, recently compared the welfare participation rates of legal immigrants to native-born American citizens. Controlling for income, they found that immigrants had nearly identical — and in some cases lower — participation rates than citizens in the three big social programs: welfare, food stamps, and Medicaid. They found that 32 percent of natural-born citizen families at the 200 percent poverty line received food stamps, compared to only 22 percent of immigrant families.

And finally the rewritten column, which again adds an attribution to CAP and quotation marks to the borrowed language, but retains many echoes:

 The CAP study also included reference to a report by Randy Capps, Everett Henderson and Michael E. Fix, respected demographers, comparing the welfare participation rates of legal immigrants to native-born American citizens. The Center’s conclusion is that the three demographers show that “controlling for income… immigrants had similar —if not lower — participation rates” than American-born citizens who get benefits from three safety net programs: welfare, food stamps, and Medicaid. They found that 32 percent of natural-born citizen families at the 200 percent poverty line received food stamps, compared to only 22 percent of naturalized immigrant families.

Born in Panama, Williams graduated high school in Poughkeepsie, N.Y., before receiving his bachelor’s degree from Haverford College. He went on to write for the Washington Post for 23 years, and was a host at National Public Radio for 10. After being dismissed by NPR for comments “inconsistent with [its] editorial standards and practices,” he was signed by Fox News to a contract worth an estimated $2 million. He is the author of the acclaimed 1988 book “Eyes on the Prize: America’s Civil Rights Years, 1954-1965,” and won an Emmy award for television documentary writing.

He has also been awarded numerous honorary degrees.

UPDATE: A 2012 Vanity Fair piece about NPR provides useful context, suggesting that Williams might have been overextended between his various gigs, characterizing his work as “spotty,” and asserting that he “missed meetings” and didn’t “do his homework.”