High-Tech Hiring

obert Hosenfeld, personnel officer for the U.S. Geological Survey, has grand ambitions. He envisions a day when hiring someone to fill a USGS job will be almost entirely in a manager's control. With a few clicks of a computer mouse, managers will be able to classify jobs, post vacancy announcements, collect resumes and applications, and get a list of qualified candidates who have been rated and ranked according to their skills whom they can then interview. The whole process will be finished in a matter of days.

To managers who know firsthand the frustrations of navigating the slow, bureaucratic federal personnel system, this scenario may seem too good to believe. But under Hosenfeld's leadership, the USGS has developed an automated recruiting and hiring system that nearly accomplishes that goal. Human resources specialists are still integral parts of the Online Automated Recruitment System (OARS), which was launched in February, but assuming OARS proves successful, it won't be long before human resources staffers are largely out of the equation. And although the USGS still must abide by cumbersome personnel regulations&:151;for example, the Rule of Three, which requires managers to select one of the top three candidates listed on a hiring certificate&:151;managers say automation has made these rules decidedly more palatable.

Other agencies, such as the Defense and Commerce departments and the Federal Aviation Administration, also have been shifting their hiring processes from the paper- and time-intensive manual mode to an automated one. Their goal: to save time and money and to boost productivity by leaving positions vacant for shorter periods of time. The efficiency benefits are clear, but the trend toward automation raises questions of fairness. Should government be entrusting computers with such an inherently human task&:151;that is, evaluating a human's skills and abilities? Then there are quality concerns. Can computers be programmed to pick up on the subtleties that so often are important in distinguishing top candidates from average ones?

Those questions remain to be answered, but for the time being, the prospect of a simpler, faster system alone seems a worthy goal for managers, employees and human resources professionals. "There is a war for talent these days," and automation is helping agencies fare better in that war, says Richard Whitford, director of the Office of Personnel Management's Washington Service Center Employment Information Office. "You can't take months and months and months to fill jobs."

The move toward automation couldn't be coming at a better time: Agency human resources staffs are down as much as 40 percent since 1993, and there is a heightened need for them to focus on strategic workforce planning as federal workers age. Automation frees HR staff from processing personnel paperwork, allowing them to focus on crafting the future workforce.

A New Process

The new systems are taking various forms across government. DoD, for example, uses an "intelligent" search mechanism that electronically reviews resumes and rates and ranks applicants' qualifications based on the presence of keywords in their resumes. USGS and others use a questionnaire with multiple-choice or yes/no questions designed to determine applicants' precise skill levels. Although USGS is not the first to use this approach, OARS is without doubt one of the most comprehensive automated hiring systems in government, with virtually all hiring now going through the system.

Hosenfeld conceived of OARS a few years ago, when he'd reached the peak of his frustration with the labor-intensive federal hiring process. He knew automation would be part of the solution, but he did not just want to automate the existing processes. After seeing private Internet job sites such as Monster.com, which allows people to electronically post resumes that can be viewed by employers, he believed there had to be a better way for the USGS to recruit and hire. OARS "is not a bad process automated," he says. "This is a whole different way of doing business."

Here's how it works. People interested in applying for a USGS job first must register in OARS. As part of the registration, a prospective applicant must post a resume and answer 25 basic questions on subjects such as federal employment experience and veteran status. The computer then reviews the information and tells the registrant what types of jobs he or she can apply for. OARS will make it clear, for example, that someone who does not currently work for the USGS cannot apply for a position that is open only to internal candidates.

"We don't want people applying for jobs they don't qualify for," Hosenfeld says, noting that in the past, people were often unsure which jobs they qualified for and so wasted time and money submitting applications that could not be considered. Human resources staffers, in turn, wasted time reviewing these applications.

Once in the system, registrants can search available job listings and apply online. Applications consist of a series of questions&:151;the number ranges from 15 to more than 100 depending on the job&:151;about one's proficiency in various skills. For example, a question about oral communications skills might ask whether an individual has experience presenting technical information to external groups. Applicants would have to choose from five multiple-choice ratings ranging from "I have not had education, training or experience in performing this task" to "I have supervised performance of this task and/or I have trained others . . . and/or am normally consulted by others as an expert for assistance in performing this task." A traditional application would merely state that good oral communication skills are important for the job, leaving it to the applicant to describe experience he or she considered relevant.

Using the completed application, OARS tabulates a score for each person, much as peer review panels and human resources specialists do under the traditional system. For example, applicants get a certain number of points depending on their skill level and extra points if they are veterans. Almost instantaneously, the computer generates a hiring certificate that ranks applicants according to their scores. At a manager's request, individual questions can be weighted more or less depending on their importance.

Before OARS, it generally took between 32 days and 63 days after a vacancy announcement closed before a manager received a certificate, Hosenfeld notes. Today it takes about four days. The personnel office uses that time to do a quality control crosswalk between the ranking results and applicant resumes to make sure there are no obvious errors. For highly qualified candidates who are likely to be interviewed, human resources staffers also may request transcripts, a curriculum vitae or other paperwork. Once applications are complete and the hiring certificate is prepared, managers can use their computers to gain access to a candidate's questionnaire and resume.

From the applicant's perspective, the process is notably easier. First, it takes only an hour or two to complete a multiple-choice application. Thoughtful applicants have been known to spend days drafting answers to questions about their "knowledge, skills and abilities," or "KSAs," as they're known on traditional applications. Second, because resumes and applications are submitted electronically, there are no photocopying or postage expenses. OARS also will e-mail registrants about new job openings in their areas of interest. Under the old system, the best way to find out about available jobs was by reading bulletin board postings outside the personnel office, Hosenfeld says. Applicants can check the status of their applications online at any time.

Questions Are Key

OARS has seen a considerable amount of interest from job applicants. As of late June, there were more than 11,000 registered users and the number was increasing by about 100 per day. "Man, are we getting applications," Hosenfeld says. "It's like I died and went to heaven."

For a GS-5 hydraulic technician job in Montana, for example, USGS might have received about five to seven applications under the old system. The same job recently posted in OARS attracted 145 applicants. The same goes for hard-to-fill information technology jobs. Before OARS, USGS was lucky to get two or three applicants for a computer specialist position. Now the number is closer to 25. Hosenfeld attributes the high response rate mostly to word-of-mouth recommendations, because he hasn't done any marketing of the system beyond some internal agency presentations.

The numbers are especially impressive given that the agency is seeing candidates of high quality, perhaps even higher quality than before OARS, Hosenfeld says. He believes the process of answering specific job-related questions steers away people without the necessary skills. At the same time, those with high skill levels stand out, he adds.

Pinpointing top candidates, however, is dependent upon asking well-designed questions, OPM's Whitford notes. "The questions should be grounded in an occupational analysis." OPM, which offers automated hiring services to agencies for a fee through its USA Staffing system, pioneered the questionnaire approach.

"The bottom line is, if you ask great questions, you are going to differentiate the great candidates from the not great candidates," Hosenfeld adds. So far, human resources staffers have been the primary drafters of USGS' questions, but they encourage managers to help with the process. Applications that don't have manager input can't be as good, says Sally Lyberger, personnel management specialist at USGS. "This system puts the onus on the manager, and that's where it should be," she says.

Tim King, a research fishery biologist at USGS' Leetown Science Center in West Virginia, knows this firsthand. The first time he advertised an open position for a biologist at his laboratory, none of the five candidates on the OARS-generated certificate were qualified, King says. Most of them had never worked in a genetics laboratory, yet he needed someone to run a genetics laboratory. King hadn't been aware that he could write questions for the application and so had relied entirely on HR office-provided questions that proved too general for his needs. The second time he advertised the job, he helped draft very specific questions about laboratory experience. The resulting certificate yielded four excellent candidates, any of whom could have done the job, he says.

As they gain more experience filling jobs through OARS, USGS personnel staffers are assembling a library of questions from which managers will be able to draw. Eventually, managers will be able to assemble their own applications using proven questions. The human resources staff will be available to address problems and answer questions.

Is It Fair?

Pluses aside, the very idea of having computers assess a person's skills and determine whether he or she is worth interviewing for a job seems inherently inconsistent with what the human resources profession is all about&:151;people. Yet proponents of automated hiring insist the high-tech systems, because they've been designed by and are overseen by human resources professionals, are every bit as people focused as the traditional way. They also are more objective and therefore fairer, proponents say.

"We've not taken the human out of human resources," Hosenfeld insists. "We've actually changed the paradigm" of making hiring decisions. In the USGS case, OARS has made it possible to abolish the peer review panels that used to score applicants based on their answers to vague questions about their knowledge, skills and abilities. That process, done behind closed doors, was often criticized as arbitrary and unfair. "There's so much human error in that" process, Lyberger notes.

Human resources officials across government agree. At the Commerce Department, where merit promotions are automated and systems for hiring senior executives and external applicants are in the works, employees consider the new approach more transparent, says Debra Tomchek, director of human resources management at Commerce. Because the rating and ranking process happens through the automated system rather than review panels, there's less chance for subjective judgments, she says.

Automated review of job applications may be too rigid to pick up on every attribute of a job candidate, adds Mike Edwards, director of personnel for the Federal Aviation Administration. But supervisors can usually uncover that information during the interview stage, he says. The FAA has had an automated system in place for certain occupations since 1996.

There's also the matter of whether a list of multiple-choice or yes/no questions really can identify the best of a group of applicants, especially when they are assessing their own skill level. With OARS, the initial evaluation is based almost totally on the applicant's word, notes Laurel Bybell, associate chief scientist with the USGS Eastern Earth Surface Processes team in Reston, Va., who recently hired someone using OARS. "So I think the interview process becomes more important." Lyberger agrees. "People will always inflate," she says. "That's why you have the interview."

Even applicants see the potential for exaggeration. Robert Kratt, a computer assistant at the Upper Midwest Environmental Sciences Center in Onalaska, Wis., who was hired for his current position through OARS, says he was aware as he answered application questions that he was being ranked based on his answers. "Truth is a relative matter," he says. "I don't feel [the computer] made it more objective because it was the individual rating himself." Kratt knew he was qualified for the job, so if he hadn't gotten an interview, he would have thought it was because he hadn't rated himself highly enough, rather than because he wasn't adequately qualified for the job.

Unions have signed on to OARS, at least for now. "I'm trying to keep an open mind," says Glenn Holcomb, an offset lithographer and president of the American Federation of Government Employees Local 1309 at USGS headquarters in Reston, Va. Still, he's not comfortable with computers taking on a progressively larger role in hiring. A faster system has its merits, he says, but "I'm afraid you are going to lose some good candidates" without some human review. If that happens, Holcomb will be the first to complain, he says.

Payoffs

There remain several unknowns regarding how well automated hiring ultimately will serve agencies and their employees. For example, although OARS appears to be operating smoothly so far, as more people register and more jobs are posted, data management may become an issue, Hosenfeld says. Consistency and compatibility of systems also will need attention. At present, agencies are implementing systems largely without considering governmentwide needs.

As agencies gain more experience with automation, those matters likely will be resolved. In the meantime, the systems should yield a significant payoff, primarily because of saved labor costs. At the FAA, a collection of automated services that includes a rating and ranking system already is saving the agency nearly $2 million a year. Officials expect to save more than $220 million a year using DoD's Modern Defense Civilian Personnel Data System. It will include the resume search tool as well as personnel records and other human resources data and will be fully operational next summer. Hosenfeld says OARS&:151;which has cost about $500,000 to implement&:151;will pay for itself after the first year and generate savings from then on. The costs of recruiting and hiring should drop more than 30 percent, he says. These savings are critical, given the staff cuts human resources departments have seen in recent years.

The leaner staffs simply can't keep up with the paperwork required by a traditional hiring system. Besides, with baby boomer employees rapidly approaching retirement age and a highly competitive labor market, agency personnel shops need to redirect resources so they can play a strategic role in heading off a potential government brain drain. Thanks to automation, USGS staffers will be able to do this. Hosen feld says. "We're [no longer] concerned with the process. We're concerned with the quality of the human resource."