Two years ago, the Federal Reserve required banks to stop routinely allowing customers to overdraft their checking accounts via declinable ATM withdrawals or debit-card purchases - source of the infamous $40 hoagie, a $5 sandwich served with a $35 overdraft fee.  The Fed said banks could do so only if customers had agreed ahead of time to the deal: that a bank would approve point-of-sale or ATM overdrafts in return for collecting such fees.

Consumer advocates had long argued that the fees were in essence short-term, high-cost loans, and that many banks were manipulating their accounting practices to multiply them. A common procedure is to clear checks and debits together from largest to smallest, rather than chronologically, so that a customer who continues to make small purchases while unaware of a negative balance can get socked for hundreds of dollars in fees.

The new Consumer Financial Protection Bureau, which last year took over the Fed's regulatory duties governing consumer financial products, is now taking a closer look at those practices through examinations at nine banks, according to Bloomberg News reporters Carter Dougherty and Margaret Collins. A key focus: how banks have been persuading customers to opt into the programs even when cheaper alternatives are available.

The Bloomberg story today says the examination focuses on practices at JPMorgan Chase & Co., Wells Fargo & Co., PNC and other large national and regional banks, including at least one, Bank of America, that has stopped the practice of allowing debit overdrafts.  Instead, BoA customers' purchases are declined if there are insufficient funds.

Bloomberg, quoting four unidentified people briefed on the examination, says:

The inquiry focuses on how financial institutions persuade customers to enroll in what they call overdraft protection programs. Examiners are looking at online and mailed marketing material as well as scripts used by the banks’ customer-service representatives to determine whether they could be confusing to consumers, said the people.

Bureau examiners have conveyed “a tone of skepticism that this is really a good product for borrowers,” said Jo Ann Barefoot of Washington-based Treliant Risk Advisors, who counsels banks on dealing with federal supervisors.

While tighter rules could help U.S. consumers, they also could threaten a major revenue stream for banks already struggling to replace income pinched by new regulations including a cap on debit-card “swipe” fees. Last year bank customers paid $31.6 billion in overdraft fees, down from $33.1 billion in 2010, according to Moebs Services, a Lake Bluff, Illinois-based research firm. About 15 million Americans overdraw their accounts 10 or more times a year, the firm said.

The bureau’s examiners also are reviewing the banks’ justifications for the size of overdraft fees, two of the people said. Large banks charge an average $35 per overdraft, compared with $25 at community banks and credit unions, Moebs reports.

Consumer confusion is widespread in this market, and the result is that the least-sophisticated customers - most likely the young and the poor - bear the brunt of the cost. An FDIC study several years ago, focused on larger banks with automated overdraft-protection programs, showed that just 9 percent of depositors generated 84 percent of overdraft fees.

Findings by the Pew Safe Checking Project suggested that bank customers might simply be missing the disclosures, since depositor agreements at the nation's 10 largest banks averaged an astounding 111 pages.  But Pew, via focus groups, has also found evidence that banks may be confusing people more directily, as I discussed in a February column.  Pew found that some customers were getting the significance of opting in exactly backwards - perhaps because they assumed the word "protection" means what it says.

"We had one woman say, literally, 'I opted in so I could avoid the $35 fee,' " Pew's Susan K. Weinstock, the project's director, told me.

When the CFPB's new director, Richard Cordray, announced in February that the agency was looking into overdraft fees, he warned that the agency considered it "wrong to confuse consumers deliberately for financial gain."

If that's what the evidence shows on overdraft protection - as organizations such as the Consumer Federation of America have long contended - it deserves a stronger crackdown.

Wireless bill shock is one of those know-it-when-you-see-it things.  It happens when you get socked with charges that are hundreds or even of thousands of dollars higher than your normal monthly bill. See some local cases here ($17,500 to watch a movie in the Dominican Republic) and here (unexpected data roaming charges in Europe).

The FCC says bill shock has happened to about 1 in 6 wireless customers - about 30 million people. The agency defines it as "a sudden and unexpected increase in monthly wireless bills" when consumers don't realize they've blown past plan limits for voice, data and text, or get hit with unexpected international roaming charges.

At least for now, the FCC is taking a voluntary approach to addressing the problem, which coincidentally makes somebody a lot of money - though it's not entirely clear how the U.S. carriers share in proceeds collected on behalf of overseas roaming partners.

The voluntary approach relies on alerts to people at risk of overage charges, and carries an October 2012 deadline.  To encourage quicker compliance, the agency is keeping public track of how the carriers are doing. So far, only T-Mobile has hit the trifecta, offering alerts for voice, data, and international-roaming overages.

You can find the chart, which includes links to the carriers' policies, by clicking here.

T-Mobile, which was kept alive last year when the Justice Department blocked AT&T's attempt to swallow it, is also busy trying to remake its image with a racy new commercial featuring its familiar spokeswoman on a motorbike - to signify speed, of course, not sex. Whatever you think of that approach (you can see the commercial below), give the fourth-place carrier some credit for focusing on substance as well as style.  

It's always fascinating to see yourself through other people's eyes. Courtesy of Austin Frakt at the Incidental Economist blog, here's a link to the latest report card by the Conference Board of Canada on the health status of our neighbor to the north and 16 of its peer countries, based on data collected by the OECD.

"They score themselves a B," Frakt says of Canada. "Guess which country is at the bottom of the list?"

You can find details here from the report, which tries to address the question of why Canada, which is immensely proud of its health-care system, only gets a B and lands in the middle of the pack on this ranking of wealthy nations. Yes, "wait times for some health care diagnostics and treatments" is cited as one potential source of drag - an acknowledgment sure to cheer diehard opponents of changing the U.S. system, in which the well-insured and affluent can get costly diagnostic tests and treatments more swiftly. On the other hand, the United States leads all nations in what we pay for our health care, and we earn a D, with especially low marks for infant mortality and life expectancy. If only we could remove the poor and uninsured from the data....

One common thread: It appears both countries have a touch of Lake Woebegone Syndrome. We may not be the healthiest countries around, but we both earn A's for "self-reported health status."

The Fair Labor Association was asked last month by Apple to audit three of its Foxconn final-assembly manufacturing plants in China - the massive, city-like factories where iPhones and iPads are produced for the world market.

The nonprofit group said yesterday that its inquiry, which included a survey of more than 35,000 randomly selected workers, had found "serious and pressing noncompliances with FLA’s Workplace Code of Conduct, as well as Chinese labor law."

This morning, PC Magazine says, Foxconn promised changes:

FLA said its assessors logged more than 3,000 staff hours - observing conditions inside the plants, reviewing policies and procedures, examining documents such as payroll records and production schedules, and interviewing hundreds of Foxconn workers and managers both on- and off-site. Summing up its finding, it said: "FLA found excessive overtime and problems with overtime compensation; several health and safety risks; and crucial communication gaps that have led to a widespread sense of unsafe working conditions among workers."

You can read the whole report here, and find specific documentation here.  The audit marks an important step in the right direction for Apple, for its Chinese contract workers, and for customers of Apple products concerned about reports - some apparently fabricated, but not all - of working conditions that many Americans would consider inhumane.

Business groups bitterly opposed the Consumer Product Safety Commission's new publicly accessible product-injury database, SaferProducts.gov.  It was designed to enable consumers to quickly report injuries linked to products such as toys, cribs, and household appliances - things that are usually safe but that occasionally come with dangerous defects - and to enable other consumers to search for such reports if they have particular concerns about an item.

It's too soon to tell if the new database, which posted its first reports on April 2, 2011, will significantly reduce the lag time between discovery of a defect and a product's recall, or encourage manufacturers to monitor  their own products more aggressively. But a report today from the Consumer Federation of America, Kids In Danger and Consumers Union suggests that the database is working largely as intended.

You can find the report here. Rachel Weintraub, senior counsel at the CFA, says that the vast majority of 6,000 reports in the first 10 months involve newer, well-identified products and reports from consumers themselves, in contrast to opponents' warnings that third-party advocates would clog the database and that manufacturers would be harmed by vague gripes involving older products. 

"It’s not someone talking about their 30-year-old refrigerator in the basement," says Weintraub, who notes that 84 percent of the injuries are linked to products identified by model names or serial numbers. "It’s definitely providing more information to the public, it's providing more information to the CPSC, and it's providing more information to manufacturers to evaluate the real-world use of their products."

About 40 reports came from medical professionals, medical examiners or coroners, and about 60 from other public-safety professionals.

I blogged here about plans for the database, and why advocates such as Weintraub believed it would be valuable, and here about efforts to kill it before it got off the ground. Those efforts haven't stopped - as I wrote in a column last fall, the new database remains a target, and has even been attacked in an unusual "Company Doe" lawsuit.

But for now, SaferProducts.gov is a resource for consumers who want to warn about product dangers, or to see if anyone else has raised concerns about something they were given or plan to buy. 

Kids In Danger just released another new report, A Measure of Safety, with this hopeful statistic: "Children’s product recalls announced by CPSC in 2011 dropped by 24% to 121, the lowest number since 2006."

For parents (or anyone else buying or passing along stuff for young children), Kids In Danger is another great resource.  One of its key messages, based on the senseless tragedy that led to its founding, is that no recall ever succeeds 100 percent.

With all eyes today on the Supreme Court and the future of Obamacare - or ObamaRomneyCare, if you prefer - it's easy to overlook other news from Washington. But here's one that will matter to anyone who worries about Internet data tracking.

The Federal Trade Commission has issued its long-awaited final report on "best practice" for protecting American consumers in an era of ubiquitous digital data. And FTC Chairman Jon Leibowitz says he's hopeful that it's all the nudge industry will need - with the threat of congressional action if self-enforcement fails.

“We are confident that consumers will have an easy to use and effective Do Not Track option by the end of the year because companies are moving forward expeditiously to make it happen and because lawmakers will want to enact legislation if they don’t,” Leibowitz said in a statement.

The FTC says its goal is to make "privacy the 'default setting' for commercial data practices - a huge and complex challenge, as the report acknowledges, in an era of "smart phones, smart grids, and smart cars [when] companies are collecting, storing, and sharing more information about consumers than ever before."  So let's hope the optimism is warranted.

I've been out of the office, and will be for another week, but this merits mention: Mike Daisey's widely quoted account of labor conditions at an Apple contractor's plant in China has been retracted by This American Life, the radio show that broadcast it, for what host Ira Glass calls "significant fabrications."

I cited Daisey's report in a column two weeks ago about Apple's China problems, though I also noted that Apple was quietly questioning some of Daisey's most dramatic assertions, including the presence of underage workers at a Foxconn final-assembly plant.

In a blog post, Glass says the fabrications included some of those assertions:

You can listen to the retraction, or find a transcript, by clicking here.  Despite its embarrassment, This American Life wisely tries to focus its latest account on the problems that Apple has publicly acknowledged at plants in its supply chain. For instance, as I wrote two weeks ago, Apple's "Supplier Code of Conduct" limits workers to 60 hours per week, and requires at least one day off out of seven. Yet Apple own 2012 "supplier responsibility" report said that more than half of the workers at each of 93 facilities had labored more than 60 hours during at least one week in a 12-week sample. And more than half the employees at 90 plants had worked more than six days in a row at least once per month. (You can find Apple's reports here.)

Apple deserves credit for its transparency, and genuinely seems focused on improving conditions for those who make its dazzling products.  But Daisey's journalistic sins shouldn't obscure the fact that conditions at plants that make products for Apple - and for virtually all its high-tech competitors - would violate norms throughout the West, and sometimes pose threats to workers' health and safety.

Credit unions added more than 1.3 million customers last year, more than twice as many as in 2010, according to this story in the Los Angeles Times.  That brings the number of credit union members to a record 91.8 million, the National Credit Union Administration says.

This chart from MyBankTracker.com may help explain why:

AT&T Mobility has now clarified which customers are subject to the wireless-data throttling I described last week, though it still isn't explaining exactly what happens to them, or give a good reason why - other than that they have the temerity to think that "unlimited data" means what it says.

The bottom line: If you use less than 3 gigabytes of data per month on your iPhone or a similar device, or less than 5 gigabytes on an Android that runs on the speedier LTE network, there are no limits on your unlimited data.

But go past those limits, and you'll be throttled until the start of your next billing cycle.

How severely? AT&T still isn't saying - though it's also still not disputing reports of a 99 percent drop in speeds.  That may sound like an enormous reduction, but it's consistent with what T-Mobile has acknowledged doing to customers who pass the stated caps on its data plans: They still get data, but essentially at 2G speeds, not 4G. AT&T says its throttled customers will "still be able to email and surf the web."  Presumably you'll recognize them as they stare blankly at nearly blank screens, or mutter curses at a mute Siri.

T-Mobile says it wants its throttled customers to consider moving to a plan with higher limits.  AT&T implies that it wants them to move to tiered plans, and also that they're data hogs: In January, it says, "the top 5 percent of our unlimited data plan customers used an average of over 50 percent more data than the top 5 percent of customers on tiered plans."

Interestingly, that "50 percent more" statistic appears to conflict with 2011 data that I and others reported last week from Validas, a Texas company that reviews wireless bills for customers. But it's more likely just an example of how statistics can be spun.

Validas, which sampled more than 55,000 cell-phone bills, found that AT&T's unlimited-data customers used about 25 percent more data, on average, than customers on tiered plans.

So is usage really soaring? Maybe not.  Validas' sample examined the top 5 percent of all AT&T data customers - about 41.5 percent of them unlimited-data customers.  For its purposes, AT&T is comparing the top 5 percent of unlimited-data customers with the top 5 percent of customers on tiered plans - customers who face overage charges above prescribed limits.

However you slice the data, there's limited evidence that many people are grossly abusing the unlimited-data plans. Validas says the top 5 percent of AT&T's unlimited-data customers used an average of less than 4 gigabytes of data per month. Validas says just 0.7 percent of all AT&T smartphone customers used more than 5 gigabytes per month.

Data hogs? Perhaps they're just people who believe the word unlimited means what it usually does.

Here's AT&T's statement, provided via email: