Excerpted from this article in  ProPublica:

For HAMP’s first two years, the government offered very little public detail about its oversight efforts. It was virtually impossible for the public – or even Congress – to know how well the banks and mortgage servicers were complying with the government’s effort to prevent struggling homeowners from losing their homes. Those years were crucial, because that’s when the vast majority of homeowners eligible for a modification – about three million – were evaluated by servicers.

The documents obtained by ProPublica show auditors finding serious problems at a major servicer during that time. Instead of publicly revealing the findings, Treasury chose to privately request that GMAC fix the problems.

“For two years, they’ve known how abysmal servicers were performing and decided to do nothing,” said Neil Barofsky, the former special inspector general for the Troubled Asset Relief Program, better known as TARP or the bank bailout, which provided the money for HAMP.

“It demonstrates that if you have a set of rules for which compliance is completely voluntary and no meaningful consequences for those who violate them, having all the audits and reviews in the world are not going to make a bit of difference,” he continued. “It’s why the program has been a colossal failure.”

Treasury continued to release few details about its audits until this June, when it began publishing quarterly reports based on the audits’ results. The public report showed what Treasury called “substantial” problems at four of the ten largest servicers – Bank of America, JPMorgan Chase, Wells Fargo, and Ocwen – and Treasury for the first time withheld taxpayer subsidies from three of them.

Mortgage servicers that signed up for the program agreed to follow strict guidelines on how to evaluate struggling homeowners seeking a reduced mortgage payment. In exchange, they’d receive taxpayer subsidies. But as we’ve reported extensively, the largest servicers haven’t abided by the guidelines. Homeowners have often been foreclosed on in the midst of review for a modification or been denied due to the servicer’s error. For many homeowners, navigating what was supposed to have been a simple, straightforward program has proven a maddening ordeal.

Meanwhile, HAMP has fallen dramatically short of the administration’s initial goals to help three to four million homeowners. So far, fewer than 800,000 homeowners have received a loan modification through HAMP, less than one in four of those who applied.

Part of the $700 billion TARP, HAMP launched in early 2009 with a $50 billion budget to encourage loan modifications by paying subsidies to servicers, investors, and homeowners. But in another example of how the program has fallen short, only about $1.6 billion has gone out so far.

It wasn’t until July 2010, sixteen months after HAMP launched, that the unit performed their first major audit of GMAC. The review included a visit to GMAC’s offices and a detailed review of a sample of loans.

The report enumerated various rule violations, including in how GMAC evaluated homeowners for modifications. GMAC’s practice was to begin the foreclosure process too quickly: The program required the servicer to give the homeowner 30 days to respond to a trial modification offer, but GMAC’s procedure was to wait only 20.

GMAC’s Proia said no homeowners were “negatively impacted by this issue.”

Auditors also found that GMAC was regularly miscalculating the homeowner’s income. In a review of 25 loan files of homeowners who had received a modification, the auditors said 21, or 84 percent, involved a miscalculation of income. Since the borrower’s income is a key factor in whether the homeowner qualifies for a modification, the high error rate raises obvious questions about whether GMAC was accurately evaluating homeowners’ applications.

Asked about this the frequent income miscalculations, GMAC’s Proia said that the “issue was identified in the early stages of the program,” that calculating the borrower’s income is a “complicated process,” and that GMAC has improved since the mid-2010 review – an assertion backed up by recent audit results published by the Treasury.

The July 2010 review also found that GMAC had been aware of certain problems such as “incorrect income and expense calculations,” but had not fixed them. Proia said the company does its best to fix problems when it becomes aware of them.

Typical of the Treasury’s oversight of the program, GMAC was never penalized for any of the rule violations. For the first two years of the program, Treasury officials publicly threatened servicers with the possibility of penalties, but instead followed a cooperative approach. When auditors found problems, servicers were asked to fix them.

The documents illustrate that back and forth. In response to the auditors’ findings, GMAC was required to develop an “action plan.” GMAC refused to provide the action plan to ProPublica and recommended seeking it and other similar documents by filing a Freedom of Information Act request with the Treasury.

Since June, when Treasury first announced it would be withholding incentives, Bank of America has received $2.5 million in taxpayer incentives. While that’s a steep reduction from the roughly $7.5 million it had been receiving monthly, the bank is supposed to be receiving nothing. Chase received $404,000 during that same time.

Treasury responded that it has programs to encourage modifications on both first and second mortgages, and that the payments Bank of America and Chase received were related to second mortgages. “Current system limitations” meant the Treasury couldn’t withhold these payments, according to the Treasury spokeswoman. Treasury is working to fix the problem, she said.