It came as an unpleasant surprise to Tony Gardner in March that he owed Metro's bus system about $1 million more than he thought he did.

It was not really his bill, but A.S. (Tony) Gardner is Arlington County's chief of fiscal analysis and, like all budget officers, he tends to speak possessively about the millions of dollars he moves around every year.

If one of those millions disappears unexpectedly, it is an unpleasant surprise. Metrobus is nothing but surprises.

That is because Metrobus is undergoing dramatic change in both form and function as it is redesigned, one piece at a time, to fit into the fabric of the ever-expanding subway system.

At the same time, because of things like Tony Gardner's surprise and wildcat strikes, there is the chance that Metrobus will lost its metropolitan thrust and be broken into a series of little bus operations. District of Columbia officials are concerned about that, although it is too early for panic.

City Council Chairman Sterling Tucker, a longtime Metro Board member who resigned from that body before officially starting his campaign for mayor, put it best the day he voted to permit Montgomery County's minibus Ride-On service to assume some routes previously run by Metrobus.

He did not want to vote against Montgomery County, Tucker said, but he was concerned that the Ride-On example was the beginning of a trend where "eveybody will have their own bus company and the District of Columbia will have what's left."

"If the suburban jurisdictions choose to operate the bus system themselves because they think they can do it more cheaply, they're going to set it up so that it serves primarily commuters," said D.C. Transportation Director Douglas N. Schneider. "That means there is not going to be any bus service for the guy who needs to get out to that suburban job."

One of the District government's main motives in supporting the full 100-mile subway system has been to bring suburbs jobs within quick reach of inner-city residents. But the subway will not go everywhere, and that concept necessarily assumes good bus service from out-lying subways stations as well as to them.

In the past 12 months more than 650 of Metro's 800 bus routes have been changed. Before the changes, most routes ran all the way into downtown Washington or through downtown Washington. Now, those bus routes have been curtailed at outlying subway stations.

The reason for the curtailments, like the reason for most things at Metro, is to save money. Shorter bus routes mean less driving time, and therefore mean that fewer buses and drivers are required. Metro's bus fleet has actually decreased by about 200 in the past year.

Many of the changes have inconvenienced transit users by forcing them to make a tranfer that was once unnecessary; many other changes, particularly in the southern and eastern sections of D.C. have also increased fares because riders must transfer to the subway and pay a second fare. Bus-to-bus transfers are free.

Although the subway has been enormously popular and, with the summer tourist influx, was averaging 200,000 riders a day before the strike, Metro-bus has been losing riders to higher fares, to the subway and to other bus systems.

Some of that loss is intentional. The subway, in the long run presumably cheaper to operate than buses, is supposed to replace much arterial bus service.

In January, the last month for which figures are available, the bus system carried about 50,000 fewer riders per day than it had on average in 1977, but total transit trips - including the subway - had increased 20,000 per day. Since that survey and before the strike, Metro's Silver Spring extension opened and the total transit picture looked even better, preliminary figures show.

But the loss of Metrobus riders to other bus systems was not entirely planned and raises the question of whether Metrobus is going to survive in its present form or even whether it should.

In the last 18 months, Metro has lost its contract to provide commuter buses to Reston. Fairfax City, concerned about long-term Metro deficits that it was being asked to share, contracted with an independent operator to provide rush-hour-only service to downtown Washington and back.

Montgomery County, where the political leadership has been unswervingly pro-Metro, has set up an efficient, popular, successful minibus network centered on the Silver Spring Metro station. That bus network is independent of Metrobus.

In Fairfax County, Board of Supervisors Chairman John F. Herrity talks openly about starting his bus network and abandoning Metrobus. He does not have the support of the board, but that could come in time.

Independent bus operations, run or contracted for by each local government, appear to have two advantages over the metropolitan concept:

Costs, routes, fares and the amount of service provided can be controlled absolutely and adjusted quickly without regard to what others are doing.

Drivers can be hired at lower salaries than those covered by Metro's labor contracts and they would be less inclined to strike because suburban labor laws are tougher on collective bargaining rights and public employe strikes than those in the District of Columbia.

Deficit control is a key issue and it goes to the heart of Tony Gardner's $1 million bill, which Arlington tax-payers will pick up two years from now. (Under Metro procedures, local jurisdictions pay back unbudgeted deficits two years later so the money can be fitted into the normal local budget cycle.)

Three things happened that made a hash of Metro's projections of what Arlington's bus deficit would be. First, a fare increase in July 1977 was budgeted, but never adopted by the Metro board. Citizen opposition beat it.

Second, many transit riders were lost in the early months of operation of Metro's Blue Line.Whereas the riders once had express buses direct to downtown, they then found themselves forced to transfer to the subway at the Pentagon and Rosslyn. That transfer itself added some time and incovenience to many trips and the irritation was increased by the fact that the subway was highly unreliable in the first three months of operations. Nevertheles, before the strike, ridership had been recovering steadily.

Third, the reductions in bus service that accompanied the beginning of the Blue Line were not as great as had been anticipated in the budget, because some express service paralleling the subway was reinstituted. Thus, the cost of running the buses in Arlington exceeded the budget.

Although all of those factors were known, their impact on the Arlington budget was not projected by Metro until the fiscal year was eight months old. "That was one of the things that really amazed us," Gardner said. "We were proceeding merrily along toward the end of the fiscal year with no warning."

To avoid this problem, the Metro staff has been ordered to develop an "early warning system" on big budget problems. Local governments can then adjust the level of service they provide to meet the financial circumstances.

Much of the problem stems from the fact that, five years after Metro assumed control of four privately held bus companies and made them into Metrobus, it still does not have a total understanding of what it is running.

Bus riders and drivers alike have long complained that Metro management's primary concern was getting the subway in the ground and that the bus was second priority. Jackson Graham, Metro's first general manager and the man who was in charge at the time of Metro's bus takeover, conceded as much in the interview in April.

The bus takeover, he said, "diverted a lot of energy" from Metro construction, his first passion. It was clear from his tone of voice that he resented the diversion. And it was equally clear at the time from interviews in bus garages that everybody in Metrobus felt they were considered second string.

Graham never specifically designated a manager for transit operations. His successor, Theodore C. Lutz, finally did so last Nov. 3, after the Metro board had spent months deciding to keep the bus management in-house after considering an offer from an outside management firm.

Thus, lines of authority for Metrobus have been clear only since November. Other problems remain.

Bus scheduling is still an arcane, enormously involved craft done largely by hand. Bus runs have to be scheduled so that drivers get proper relief time and so that their runs start and conclude at garages at times that are compatible with service on the street. Union work rules require a certain number of straight-time shifts and a certain number of split shifts. During rush hour, many buses dead-head to the beginning of the line, thus making only one productive run on a round trip.

Outside experts, looking at the Metrobus system, are convinced that far too much driver time is lost in layovers and dead-head time because of the method Metro uses for scheduling.

No system has been developed so that Metro knows, down to the penny, which of its bus trips attract riders and which are losers in every way. When Metro thus starts reducing service at the insistence of local jurisdictions, the reductions are not made in an educated way.

That summary of problems leaves too negative an impression. Budget specialists and politicans from every local government were interviewed in connection with these articles. To a man, they believe that Lutz is turning the elephant around.

For the last three years, Metro has lived within its projected budget.Lutz has imposed hiring freezes and other actions to partially offset the revenue losses that have occurred because of decreasing bus ridership and the lack of a fare increase in July 1977. Just last Thursday Lutz proudly announced that Metro's accounting of fiscal 1978 was completed and that Metro had managed to operate at $200,000 less than its projected deficit.

[That is a systemwide total. Because of shifts in accounting within the area, Tony Gardner's surprise is still about $1 million.]

A consultant is preparing a report on how Metro can computerize its bus scheduling and at the same time improve driver efficiency.

Eckhard Bennewitz, a tough, bottom-line man who learned in Robert S. McNamara's Pentagon, is busy delaying approvals on things that cost money and devising a system that will give Metro cost and revenue facts on every bus trip made.

That leaves the second major incentive for local governments to look to their own resources for bus operation: the union contract Metro has with its bus drivers.

One statistical comparison is all that is needed to understand why this is a huge issue. The average Metrobus driver earns about $18,800 a year. The average full-time driver for Ride-On, the Montgomery County minibus service, earns $10,000 a year.

The contract Metro has with Local 689 of the Amalgamated Transit Union does not permit part-time drivers. Metro would like to cut costs by using part-timers on some rush-hour bus routes. Simply stated, two rush-hour bus runs per day probably equal four hours of work. There is no real need for another four hours of driver time. But Metro guarantees its drivers eight-hour days and 40-hour weeks under its contract.

If part-time drivers could be hired, more service at cheaper cost to the taxpayer could be provided.

Before last week's transit strike, Metro and the union had been negotiating and then preparing for arbitration on part-time employment, among other issues. It was clearly revealed in federal court testimony growing from the strike that Metro was using the cost of living pay demand of the striking workers as a bargaining chip to be played against Metro's desire to hire part-timers.

The judge ordered expedited arbitration on the cost of living pay, the key striker demand. But that arbitration, which was heard last Thursday, will settle only the issue of whether the strikers should have received a quarterly cost of living increase automatically while their new contract was in arbitration.

The question of part-time drivers, and the question of whether Metro can win some reduction or change in the cost of living clause in a new contract, remain under binding arbitration. It is technically possible that the strikers could win a cost of living increase on their old contract in the court-ordered expedited arbitration, but lose it or have it modified in the subsequent arbitration on the new contract.

Small, county-based systems would be less susceptible to wildcat strikes or even to unions, especially in Virginia where it is illegal for local governments to enter into collective bargaining contracts with public employe groups.

John Shacochis, a Fairfax County supervisor who is an alternate on the Metro board, referred to that the day the Metro strike began when he said, "I hope the people involved in this realies they have just given the people on the Virginia side pushing for their own bus system their strongest argument. Under Virginia law, (commuters) will be free from this kind of harassment."

There is no guarantee, however, that a number of small, locally run bus operations would be any more efficient in the long run than a well-managed Metrobus system.

Jeffrey Bruggeman, a transportation planner for Peat. Marwick Mitchell & Co. who is familiar with the Metro problem, said that, in the long run, such operations would probably cost local governments as much as Metro does today.

"It becomes a question of whether the labor contracts are similar to Metro's and of whether there are economies of scale in operating a metropolitan-wide bus operation, such as Metro's," Bruggeman said.

"I think the main thing local operations would give (local governments) would be increased ability to tailor service to demand," Bruggeman said. "Ultimately, the kind of service Montgomery County is providing has got to cost more (than Metro) just because the amount of service they are putting out there is greater" than what Metro would provide.

Regardless of who runs the buses, however, it is essential to the financial health of the subway system that there be a closely coordinated bus system.

"The rail system does not work without the bus system," said Littleton MacDorman, the chief Peat, Marwick consultant on the Metro project. "You cannot look at them independently."

Long-range transit plans have always included buses to provide neighborhood service, with convenient connections to subway stations in areas close to train lines.In other areas, buses were to continue providing long-haul service.

Peat, Marwick Mitchell has done several projections on what transit will cost taxpayers in 1990. Bus service - regardless of who provides it - ranges from 40 to 50 percent of the cost, depending on the projection.

NEXT: Whose Dollar Is It?