According to an AP story
The lack of a helmet requirement in New York’s month-old Citi Bike program...has some experts predicting a disastrous increase in injuries and fatalities.
so far, through more than 500,000 rides, such predictions haven’t come true, with just three minor accidents and nothing more serious than some scrapes and bruises.
John Pucher is the only expert they cite [unless you count comedian (and former roommate of my former boss) Jon Stewart or the AAA-New York spokesman as experts], and while I certainly think he's a knowledgeable guy, they don't actually have a quote from him and they lump helmets in with other possible causes.
A lack of a helmet requirement, blocked bike lanes and inexperienced cyclists hitting the road are all reasons Rutgers University public policy professor John Pucher cited in predicting that the number of cyclist injuries and fatalities will double or even triple in Citi Bike’s first year.
And near the bottom of the article they have a study that seems to contradict this.
The Mineta Transportation Institute in California studied 14 bike-share programs and found relatively low accident rates, averaging 1.36 serious or fatal accidents in 2011. In the same year, there were 22 fatalities for bicyclists in New York.
The fact that DC and Boston have been relatively safe isn't even really mentioned. They do pull out the discredited claim that
in 97 percent of bike fatalities from 1996 to 2005, the rider wasn’t wearing a helmet.
The statistic itself is somewhat daming in isolation, but no more so than pointing out that 100% of bike fatalities in the same period were not wearing rainbow colored suspenders. They don't give a citation, but it is likely this report which also notes that only about 13% of New York cyclist were even wearing helmets during that time. Add in the fact that often the helmet use was unknown and that the reporting of helmet use in fatalties is frequently reported wrong and this statistic is just above junk status. That's not to say that helmets aren't useful, only that this statistic makes them look more useful than they probably are.
The Pucher reference most likely comes from this article, but they wildly misread it. Pucher does mention that having more cyclists on the road increases total risk - just as the number of Segway related deaths has gone way up since the 1980's when there were none. But he does not in any way attribute that to helmets
"The problem with requiring helmets with bikesharing systems is they're generally meant to be for very short trips," says Pucher. "Let's say you get off the subway or Metro, and you just need [a bike] to cover maybe, say, ten blocks. It's unlikely you're going to carry with you a helmet to use for that ten-block ride."
Alternatively, he says, bikesharing programs would have to take on the difficult task of figuring out how to dispense one-size-fits-all helmets at every station.
Short trips also generally make for less risk than longer trips. And while they may not do so for safety purposes, bikeshare systems encourage shorter trips within their pricing systems. Capital Bikeshare, for example, ramps up the price after 30 minutes of riding.
While there is an argument that helmet laws are unnecessary, Pucher points out that more riders on the roads, particularly when new bikeshare systems open for business (like New York's Citi Bike, which launches in July), naturally leads to greater risk.
Then the money line - which isn't even really a prediction, it's a fear.
"All of a sudden, you've got these 7,000 bikes, which anyone with a credit card can use. ... My guess is the people using those bikes are far less likely to be experienced cyclists," he says of New York's new system. "What I fear is you're going to have indeed a spiking--and it could be a doubling or a tripling--of injuries and fatalities, both of cyclists and pedestrians."
So he's talking about inexperience cyclists - not helmets. He's also talking about pedestrian injuries spiking.
I'd love to hear what he thinks of the AP story.