The experience of learning to be a manager in the new economy is full of surprises. Sometimes it's humbling; other times it's exhilarating

I was in college when I first realized there's a big difference between delegating responsibility and relinquishing control. I was arranging a three-day outing for 1,500 students, and it was a bigger job than one person could pull off. So I recruited 30 classmates and gave each of them total freedom and control over particular activities. It worked like magic. On their initiative -- not mine -- we had computer models for transporting students, new campsites to accommodate more hikers, and some very creative menus.

After getting an M.B.A. and doing a stint on Wall Street, I became assistant general manager of a traditionally organized manufacturing company. When that company was growing, the management style wasn't so important. But when sales sloughed off, the traditional structure just didn't work. Rather than joining together to solve problems, people pointed fingers up and down the hierarchy. It was hard to tell who was responsible for mistakes and how to prevent those mistakes from happening again. It was also clear that employees were prevented from solving problems because they simply didn't have enough information.

So I became a believer in giving employees as much information and responsibility as possible. When I was hired as president of Merlin Metalworks Inc., I had a chance to try out my theories. There I discovered that theory can bump heads with practice when you're trying to manage a young, fast-growing, shorthanded company.

When I arrived at Merlin, in 1989, there were five people in the basement of an old mill building, using pre-World War II machines to produce bicycles from an alloy recently developed by the aerospace industry. Since then Merlin has grown 10-fold both in revenues and in employees, many of whom had been laid off by large corporations like GE and General Dynamics.

One day early on in my career there, two welders hovered outside my door, waiting for me to get off the phone. A welder's time is not cheap, but the person on the phone was a potential investor, and I didn't want to put him on hold. When I finally ended the conversation, I learned that one of the welders wanted to let me know that he had a dentist appointment the next week, and the other needed to leave early to pick up an anniversary present for his wife. They wanted to know if they should fill out some type of form to get approval for taking time off. The scene just didn't make sense. It was at that point I realized that employees are grown-ups who have things to do that I really don't want to take time out to think about, much less create, distribute, and file documents for. That was the moment flextime hit Merlin.

Despite my resolve to push power and information down the organization as far as possible, I found the learning curve pretty steep. One morning as I walked in, I glanced at the shipping log for the previous day and found, to my dismay, that nothing had been shipped. Zippo. I began asking around and learned that a customer had called about a problem bottom bracket -- the place where the pedals attach -- which made that customer's $4,000 bike useless. The customer-service department has the power to stop everything to solve an isolated problem for one customer. In this case, that meant turning off the final threading machine for a whole day, which brought shipments to a halt. Once people understood the blip they had caused in cash flow, they realized that we needed to adjust our service goals. An expensive lesson, but the mistake never happened again.

It's easy for a manager to get so caught up in the big picture that little details that affect employee morale can be ignored. That's especially damaging in a company that's working to involve employees in every aspect of the business. I learned about that one stressful week when the finishing department reached its most aggressive goal to date. Bikes were practically flying out the door. The finishers came by individually to tell me the department had made the goal. I was polite, but I was thinking, "Yeah, I already know about that." I was fixated at the time on a large wire transfer from one of our European distributors that was late. The next time the finishers made such an important benchmark, they ordered an ice-cream cake in celebration. No one saved me a piece.

At a company like Merlin, everybody has ideas about how to make things better. Once you've decided how you want to solve a given problem, it can be infuriating when someone questions your choice. For me, research-and-development decisions were the worst. I was forced to choose among options with very little data to go on. At one point we had decided to redesign the rear brake bridge on the road bikes. I was certain the new design would be cheaper, because it didn't involve welding and we could use a subcontractor during a period in which we were really busy. I gave the go-ahead to start production.

Then, shortly after the first order was placed, the person in charge of purchasing insisted on rehashing the decision. I agreed to go over it, although I don't think I was very successful at disguising my impatience. It turned out that the new design was going to lead to a series of new expenses, which added up to a lot more than a little extra welding time. I had to admit I'd been just plain wrong. That employee's persistence saved the company a lot of money.

Letting go often led to results that were different from what they might have been had I been involved. But so many things simply would not have happened at all had I not let go. One of the most satisfying things you can do as a manager is to point someone in a direction, give that person the resources he or she needs, let go, and watch him or her rise to the occasion. When we embarked on a short-term project to produce titanium racing wheelchairs, I needed someone to be in charge of the operation. I asked one of the machinists if she was up to the challenge. She was afraid she might fail and wondered what that would mean for her job. We agreed that if after a few months she felt as if she were banging her head against the wall, she could go back to her old job, and I would find someone else.

The results were astounding. There was, of course, some hair tearing (but it wasn't mine). In the end there were some 20 wheelchair racers from around the world in our titanium chairs, racking up results in all kinds of marathons and road races. And I had barely lifted a finger.

Under pressure, though, it's really easy to slip back toward a more traditional organization of work. We were preaching about teams, but in at least one case, we'd set things up in a way that forced employees to wait for commands and made it almost impossible for them to act on their own.

The problem was in the machining department. We were approaching a six-month backlog, and in a seasonal business, that can be suicidal. ("We realize it snows in Colorado in November, but is it OK if we deliver the bike you ordered last May on Halloween?") We kept increasing the size of the runs. (There are more than 35 operations in any given run of bikes, some of which proceed simultaneously, some of which must be linear, and a few of which are iterative.) The machinists would roll their eyes every time I added another 50 frames. "I know you made 200 57-centimeter road bikes last week, but this week we need 250 58-centimeter bikes." And then when we ran out of a critical size, I would ask the machinists to drop everything and make a minirun to accommodate an important customer. The machine coordinator became a drill sergeant. He had no choice.

He finally came to me and said we needed to make all the runs the same volume. "Right," I thought. "We'll just make all the customers ride the same size bike. So what if they can't reach the pedals." He explained that if the runs were small enough they could be completed faster, allowing us to do more runs a month. That way we could make only one small run of a size that sold in small quantities and two or three small runs of sizes that sold in larger quantities. If all the runs were the same size, similar operations would take the same amount of time, and the machinists would be able to keep up with the schedule and plan for themselves what to do next -- to work as a team. That new system eliminated both the need for panic-induced miniruns and the job of drill sergeant.

Finally, what manager hasn't had this experience: You're walking around the company and you come upon a group trying to solve a problem. And as you walk toward those people, you begin to feel some of the responsibility for solving the problem moving from their shoulders to yours. Sound familiar? Well, I now try to keep walking, leaving them with the responsibility until they call for help.

In the end relinquishing control can be full of surprises. But for employees and managers, the pluses unquestionably outweigh the minuses.

Ashley Korenblat is on the board of Business for Social Responsibility and has worked with the Office of the American Workplace, a new agency within the U.S. Department of Labor. She is looking for her next entrepreneurial adventure.