(Prashanth Vishwanathan/Bloomberg)

Once upon a time, if you emerged from college with zero marketable skills but dreams of a safe career path, you went to law school. For three years and $150,000.

Today, it seems, you go to a computer coding boot camp. For just three months and just $15,000.

In the past few years, dozens of these sexy new “coding conservatories” have cropped up around the country. They promise to teach in only a few months the hard skills required to land a stable, high-paying, lifelong career. Currently these programs are usually ineligible for federal student aid, but policymakers have begun eyeing them as potential successors to more traditional retraining programs, which have checkered job-placement track records. This month, for example, Commerce Secretary Penny Pritzker visited a New York-based coding conservatory, the Flatiron School, to see whether its training model might be worth replicating elsewhere around the country.

Employers, students and policymakers seem to view such programs as the magic bullet for putting unemployed Americans back to work and closing high-tech skill shortages, and investors see them as cash cows. But many of the same kinds of overpromises common to the recently popped law school bubble may lurk in these coding academies, too. Policymakers should proceed with caution.

For now, though, there are some impressive success stories.

Take Tiffany Peón, who graduated from Berklee College of Music in 2009.

Peón spent much of her first year out of college waiting tables, living with her parents and studying for (what else?) the Law School Admission Test. But a move to New York changed her career path; while staffing a customer-service line at a startup, she often fielded calls from recruiters trying to poach her firm’s developers. What those developers did looked pretty fun. She began brainstorming ways to break into the field.

She contemplated teaching herself, as generations of autodidactic computer nerds have done before her, and as her brother (a senior engineer at a major tech firm) recommended. But she doesn’t learn well in isolation, she said. She also considered doing post-baccalaureate and master’s work at New York University, but that meant multiple years of study and deep student loan debt.

Then she heard about the Flatiron School.

Flatiron is not yet three years old. But it boasts millions in funding from venture capital firms, a sizable job-training contract with New York City, and, most importantly, an enviable job-placement rate. An independent auditor found that 95 percent of job-seeking graduates secured work within 120 days; the average salary of those accepting full-time salaried positions was $74,000.

Upon completing Flatiron’s 12-week Web development program, Peón landed a job almost immediately. She’s now an associate software engineer at Constant Contact, an online marketing company, and finds herself fending off companies trying to hire her.

Peón says she thinks her work is similar to that of people with bachelor’s degrees in computer science. Employers I spoke with are not so sure. In one entrepreneur’s analogy, coding boot camps equip people to be carpenters who build things from other people’s plans, but not the higher-paid architects or engineers who create the designs and make sure they’re structurally sound. Maybe that’s okay; the tech economy needs carpenters, too.

The bigger challenge, I fear, is how well these kinds of programs will scale.

Unlike most training programs, Flatiron is extraordinarily selective. Its admissions rate of 6 percent rivals Harvard’s. All admits must go through interviews with both co-founders and jump through other hoops such as coding a tic-tac-toe game (even if they have no background in programming). It’s no wonder, then, that employers return again and again to Flatiron for high-quality hires: Flatiron has not only trained these students, but has also pre-screened them to make sure it ends up offering only the most perseverant, passionate, marketable workers around.

This is of course not so dissimilar to the law school model. Talk to any hiring partner at a law firm, and they’ll tell you the reason they recruit at Harvard Law has less to do with what graduates learn there and more with how useful the school’s admissions process is as a screening device.

I don’t blame Flatiron for cherry-picking its students; this makes sense, especially when it is capacity constrained. But the real test for schools such as Flatiron, and the many copycats that already have worse job-placement rates, will be whether they can take any shmoe off the street and turn her into a star programmer, as policymakers and employers are apparently counting on them to do.