Doug MacDonald is a Sightline fellow. He served as Secretary of Transportation for Washington State from 2001 to 2007 and now lives in the Greenwood neighborhood in Seattle. He mostly rides the Metro 358, 5, 48 and 70 and the Sound Transit 550, plus whatever comes along in the downtown transit tunnel.
It’s a fundamentally worthy public enterprise, facing the toughest of challenges: sustaining service to its r iders while it stands awash in a rising tide of red ink. But King County Metro Transit—the largest transit agency in Washington State, providing more rides each year than the rest of the state’s transit agencies combined—needs a strong and focused board of directors to guide the system through this crisis.
Good luck finding that kind of leadership. In recent years, the political oversight structures governing Metro have demonstrated little aptitude for making smart strategic choices, achieving greater efficiencies of service, and setting sound priorities. As the gap between Metro’s needs and funding has widened, it’s been increasingly unclear whether anyone’s truly minding that gap.
But before we dive into Metro’s governance problems, let’s talk about the agency’s biggest recent success: its remarkable tally of ridership growth. For the first half of 2009, Metro buses and electric trolleys fielded almost 380,000 boardings on its buses and electric trolleys on an average weekday. That’s 60,000 more boardings a day than in 2005—an impressive 20 percent gain.
We are a nonprofit. Donate now to support more research like this!
True, ridership has slipped a bit since last year, as the economy has stumbled and gas prices have eased. Metro’s ridership gains are impressive nonetheless. For perspective, 60,000 new boardings is more than twice Sound Transit’s total projected light rail boardings (largely taken from existing bus trips) after the new tracks are finally open next year all the way to the airport. Or consider Portland, where combined bus and light rail system boardings have grown by just 4 percent since 2005.
Yet despite Metro’s ridership strengths—and in some ways because of them—the organization is in big trouble. Surging ridership has added to costs; the economic downturn has constricted sales tax revenues; and Metro, like any organization, is not free from a few leaks in its own roof.
Metro operates an astonishingly complex array of routes and services, a result of history and the enthusiasm with which new missions have been pressed upon it. It has an undeniably high cost structure. And burgeoning ridership comes with higher visibility, and elevated expectations from a public that cut little slack for late or over-crowded buses—of which there now are more than there used to be.
Those challenges are exacerbated by the fragility and Byzantine complexity of the agency’s finances. Metro’s proposed 2-year budget—which is a moving target, given all the crucial deliberations still ahead in November—stands at about $1.2 billion. Those funds come from county transit sales tax revenues. Plus fares. Plus payments from more than a dozen local cities, including Seattle to help support bus runs in the local precincts. Plus money paid Metro by Sound Transit to operate quite a few of the Sound Transit Ride the Wave Regional Express bus routes. Plus important, but limited, flows of federal transit aid.
But the recession has cut deeply into Metro’s crucial sale tax revenues. At the same time, the cities that pay for some of Metro’s services are facing their own budget crises, threatening funding cuts for transit—a situation that will only worsen if Tim Eyman’s I-1033 wraps a bottom-of-the-economic-cycle tourniquet on future services. Sound Transit’s revenue projections have fallen 20% below the long-term estimates from last fall—meaning that it may be forced to look for potential cuts in its payments to Metro.
As those losses have hit Metro from one side, Metro riders have shouldered fare increases on the other. There has been one fare increase already, another is coming, and there will be more.
Putting it all together, the agency faces a serious budget shortfall. The exact size of the depends on whom you ask, and how many of the piecemeal fixes have been taken account of in the moving target. But $200 million over the next two years is a fair round number.
WHO MAKES THE BIG DECISIONS?
These budget woes have riveted the public’s attention on seemingly inescapable cuts to Metro services. The questions reverberate: When will they come? How deep will they be? Whose routes will feel the axe?
But first the all-important threshold question: Who actually makes the difficult decisions? Who exercises responsibility and accountability for steering Metro through this crisis? Is there a table somewhere with a sign that says “The Bus Stops Here”?
King County Metro’s management team, including General Manager Kevin Desmond and colleagues, reports to King County Executive Kurt Triplett. As the professional managers, they matter a lot. But setting Metro’s budget—the power of the purse strings over what Metro does and how it does it—is the responsibility of the King County Council. It’s to the Council that Triplett and Desmond have submitted Metro’s next $1.2 billion proposed two-year budget.
However, in 1992, when Metro was folded into county government, the County Council provided a layer of broader, but not necessarily sharper, oversight of Metro Transit through a policy and strategy review body called the Regional Transit Committee.
Think of the Regional Transit Committee, therefore, as the closest thing we have to a board of directors specifically for Metro, but not exactly. The Committee is made of three county council members (Constantine, von Reichbauer and Hague), two Seattle city council members (Drago and Clark) and local elected officials sent from suburban cities, one each from Sammamish, Federal Way, Issaquah, Maple Valley, Renton, Bellevue, Burien and Kent. This Committee, in turn, reports up to the County Council.
Unfortunately, the lingua franca of this hydra-headed governance structure is local politics, not transit management. So it’s unsurprising that for months, the discussion of the budget crisis has revolved almost wholly around the high-visibility and politically charged debate over service cuts. Much of this debate has focused on the so called “40/40/20 rule” that was adopted in 2002 by a county council vote that was divided on parochial lines. The 40/40/20 canon dictates that transit service hours added by Metro, with a few exceptions, should be bestowed in the ratio of 40 percent to the eastside, 40 percent to the south part of the county, and only 20 percent to the Seattle/North region.
Whatever defense could have been made for 40/40/20 in past, the rule is now completely out of date. And in an era of deficit budgets, using 40/40/20 to guide bus route contraction serves as well as cutting cheese with a hammer and spreading it with a screwdriver. State Senator Fred Jarrett recently declared in Crosscut that it should be laid to rest. The Municipal League of King County had offered the same view last November and recommended a new course to the
Regional Transit Committee five months ago.
Service cuts will ultimately depend on how much money needs to be saved, and no proposal currently on the table offers cuts at a scale that yield more than fraction of the dollar savings and revenue gains that must be found to close the budget gap.
Which leads to a crucial question—one that demands at least as much attention from Metro’s political directors as whose route will get whacked: how much money can metro save without cutting service? How many ways are there to make Metro’s existing service more efficient?
TRANSIT EFFICIENCIES: FROM TALKING THE TALK TO WALKING THE WALK
Squeezing out efficiencies is hard work, but in the end it’s a better way to spend time than arguing endlessly about where to put the service cuts. The efficiencies can be hard to find—but they’re there. They won’t solve the revenue problem or allow the system to escape some cuts in service if current trends hold. But they can help. And a great place to start looking for them is this 370 page performance audit, requested, to its credit, by the County Council and released last month by the King County auditor.
Produced with the help of a hired squad of true experts, the audit so far has only seen one recommendation gain much public notice. Metro, the audit preliminarily suggested last summer, had built up an over-sized reserve fund for future bus purchases. The fund’s $105 million could be applied from the reserve fund to the crisis of the coming budget’s operating deficit. Metro basically agreed, with a few small reservations about precisely how much could prudently be taken from the reserve. It now looks as though the booty will be applied in four shares over four years of the deficit crisis.
The rest of the audit now sits on the shelf. There are signs in the County Executive race that the candidates at least know about it, though one must wonder if they have any idea what’s in it. Over at Metro itself, some of its recommendations certainly have gained traction. The Regional Transit Committee mentions it as a source of possible savings that would shrink the service cut need, but hasn’t taken up its specifics or shown evidence of the hard policy choices it will have to make if the audit is to yield real and long-term value.
Overall, it’s a good audit, balancing tough insights with respectful recording of some of Metro’s strengths: its sound preventive bus maintenance; the agency’s generally good practices in collecting and using ridership data; and recent positive trends in communicating with riders through long-overdue improvements to its website.
But the audit gives no free rides when it turns to problems. Like this one: Do Metro’s complicated daily schedules make the most efficient use of its drivers and its more than 1,400 buses and trolleys?
The audit recites the myriad factors involved in serving its riders: putting its buses and trolleys on a huge array of routes; dispatching buses from widely separated bus barns, with each bus matched to a qualified driver; scheduling each driver’s work day for expected traffic conditions, while allowing for the reasonable breaks required for safety and stipulated in the labor agreement. Grappling with the audit’s implications requires the diligent reader to plow through the “round trip cycle time analysis” and the “recovery time to in service time ratio” and more. At least you will never board the bus again without marveling that every morning Metro gets this system to run at all, and even produces three schedule revisions a year to refine the product.
The “recovery time to in service time ratio,” however, is not just techno-babble, but a key to improving Metro’s efficiency. Recovery time is the cushion built into every bus run—the time when the bus and driver wait to start their next trip along the route. In-service time, the flip side of the ratio, is the time the bus and driver are actually driving the route picking up and dropping off riders. Adequate recovery time is essential to keeping the day’s schedule on-time, protecting the reliability of connections and assuring that operators get the breaks they need in a long shift of tough driving in traffic. Metro has to get the right balance for every schedule on every route. Too little recovery time to in-service time and the schedule falls apart. Too much, and bus buses and drivers—expensive assets, to be sure—spend more time than necessary parked at the beginning or end of a run.
The audit reports that Metro schedules show recovery time at troublingly high levels, at least by comparison to other systems. Metro’s buses on the daily routes spend on average 29 percent of their time in the recovery mode of the round trip cycle time and only 71 percent of the time in the in-service mode.
That’s a lot higher ration for recovery time than at counterpart systems in Phoenix (20 percent), Denver (24 percent) Portland (21 percent), San Jose (24 percent), San Diego (21 percent) or Santa Monica (18 percent). Higher, if you have been following this, is not better!
What should the ideal ratio be? There’s no easy answer, because local conditions vary so much—not just between, say, Seattle and Phoenix, but even among different routes just in the Metro system. And even if you do set a tighter target, there’s no guarantee you’ll meet it, given the enormous complexity inherent in route scheduling.
For that task of actually building the schedules, Metro employs a sophisticated software package called HASTUS to optimize its routes and schedules. HASTUS is used by more than 200 transit companies all over the world, including places like New York, Los Angeles, Stockholm, Singapore and Sydney.
But, according to the audit, Metro doesn’t benefit from the full power of HASTUS. It hasn’t maintained or updated necessary customizations to local conditions. Metro’s schedule builders—former rank-and-file bus drivers with deep knowledge of bus routes and traffic patterns—haven’t had all the training they deserve on the software. The agency hasn’t used HASTUS modules that would avoid cumbersome manual processing. It hasn’t properly used data fields on which HASTUS depends to generate optimal results. And it hasn’t developed or maintained widely accepted performance measures of operations efficiency that would better inform itself and its overseers and customers how it is doing.
Significant tens of millions of dollars of savings could lie in better route scheduling by making it possible to run the routes with fewer buses and drivers. The savings are no silver bullet. They’ll have to be eked out route by route and dollar by dollar. Years, not months, of patient improvement, at a cost, will be required to achieve them.
Desmond and his team already have promised that at least one dormant module in HASTUS will be running in time to help the coming annual February schedule adjustment. But money for IT support and software training is tough to come by if it has to be traded against saving even one trip from a service cut. Will the politicians be far-sighted enough to spend scarce money today for better efficiency tomorrow? Early indications here seem good that at least some money for these purposes will be approved even in the face of the deficit but the Regional Transit Committee has yet to fully consider where this course will take it.
How many dollars can be saved, and how quickly, depends on hard choices the policymakers will have to resolve. For example, savings from shrinking the scheduled recovery times can be a lot
greater if the system is planned against the standard that 90 percent of the runs must run on-time rather than 95 percent. In other words, to save a lot of money, schedule performance may slip a little. Slightly more often then today, a rider may miss a counted-on transfer. Is that too high a price for a lower across-the-board recovery time ratio? These are policy choices. The public is entitled to have the policy-makers—the County Council and the Regional Transit Committee—make the choices and be accountable for the trade-offs they select.
Indeed, the audit itself only presents the opening questions that policy-makers have to consider as route efficiency is examined. There are at least two big policy areas that the audit does not tackle directly—and that can only be addressed by competent leadership willing to make tough choices.
The first of these areas takes us back to the complicated mechanics of any program of service cuts. County Executive Triplett’s proposed solution to the service contraction problem presents an alternative to using the 40/40/20 rule to allocate the cuts. That is just to take off nine percent of the service more or less evenly across the system. The plan Triplett has put on the table would play out in four phases from February 2010 through September 2011, altogether cutting 310,000 service hours and yielding across the two year period about $25 million in savings.
This approach sounds, at first blush, like at least a plausible alternative to the 40/40/20 canon. But it too is a dud. For technical reasons, across-the-board cuts are a hopelessly unwieldy parameter to add into the service delivery schedule for vast and interconnected transit system, and probably would result in route schedules and bus and driver utilization that would be even more inefficient than today’s practices. This is cutting and spreading cheese with a fork. There’s no obvious magic formula for allocating cuts. Sooner or later the policy-makers have to give Desmond and his management team the leeway to make cuts by hard-nosed selection of routes with the lowest productivity and benefits to riders.
A second major policy choice that Metro’s leadership must make involves overhauling and fine-tuning its route offerings. Metro is long overdue for a basic route restructuring that would make the system more efficient. This is a step beyond tinkering with the schedules on the existing routes, as contained in the audit recommendation Basic route structures and their alteration: Are the policy-makers committed enough to improving Metro’s efficiency to lift the lid off that neighborhood-by-neighborhood can of worms? The politicians really have no choice if the future is to see a stronger, more efficient Metro. It will be interesting to see whether they are up to the challenge.
Questions from the audit don’t get easier as they directly approach problems in Metro’s labor agreement with its operators. Like this one: Can Metro manage effectively the task of providing relief overage for all its operator shifts?
Just as a hospital must staff every nursing station on every shift, Metro has to staff every scheduled bus run with a driver. That means a relief operator to cover every planned absence (vacations or military leave) and every unplanned absence (sick leave or an injury). It’s complicated and expensive to do this right, because it’s costly to keep a big roster of relief operators. And even more expensive not to do it right, since driver shortages mean missed runs, busted schedules and angry customers.
This is no simple problem. There are 1300 regular full-time drivers, along with 900 part-time operators driving regular assignments. Covering relief shifts for that staffing requires Metro to maintain approximately 500 additional full-time operators available for temporary assignments or call-in. They line up on two separate bid lists, one generally covering for short-term and unplanned absences and another for longer term planned absence, and each list separately maintained for each of the seven terminals from which the routes originate. A smooth schedule of planned absences means savings: you need a shorter roster of relief operators. Ditto if unplanned sick leave works only reasonable, not unreasonable, disruption to operator schedules.
Over the last decade, evolving terms of the labor agreement haven’t given much support for Metro efficiency on relief staffing. Vacation schedules, for example, for all practical purposes are set by the operators for themselves. Metro management is required to accommodate the vacation schedule desires of its operators “to the maximum extent possible.” When a lot of operators choose the same day for a vacation, a vacation peak use day occurs for which Metro needs an especially large roster of relief operators. The audit analyzed vacation use for a few months of 2009 and found huge variability. On some weekdays, 60 operators had arranged their vacations. On others, 120 operators took vacation. Weekend to weekend showed a similar range of variability. That’s expensive for managing relief coverage: in the millions of dollars.
Covering for operators’ sick leave, an unplanned absence is even harder. Under its labor agreement, Metro today has few tools to exercise reasonable control on the small number of employees in this organization, like any other, who will, if they can, take sick leave that they shouldn’t.
Prior to 2001, management could require independent medical verification of an operator’s sick leave instances after two days had been used on three occasions in twelve months. The rules were loosened in 2001: drivers could claim sick leave at least seven and sometimes as many as twelve instances before verification could be sought, and no independent verification to be asked for if an employee had more than three months of accrued sick leave on the books. The rules were loosened again in 2004. As long as the operator had any accrued sick leave on the books, only self-certification of sick leave was required, however many times it was used.
According to the performance audit, sick leave use at Metro after the change in 2001 grew at three times the rate of growth of total labor hours. This picture translates into ever higher relief operator costs, especially when, as the audit describes, 65 percent of the time when coverage is needed for an absent part-time operator, a full-time operator is called in and paid a full shift. It would be cheaper to backfill a part-time operator with a part-time relief—but that’s generally not allowed. And it would be even cheaper to use overtime with a full-time operator—but that’s also usually not practicable under the agreement.
Of course, Metro can’t now unilaterally change the rules in the labor agreement. That has to be taken up in bargaining the next agreement. That’s coming soon. But the audit found that even within the framework of the current labor agreement, if Metro’s staff were using all the tools it owns in the HASTUS software, reasonable service reliability could still be maintained with seventy fewer employees on one of the relief bid lists—suggesting a savings possibility of millions of dollars.
It will take leadership from the County Council and the Regional Transit Committee to frame expectations for the next collective bargaining so that progress can be made toward a fair, but more efficient, set of rules. It’s a tough issue. Metro, including its union, needs the support and loyalty of the people who get on the bus. Very few of them going to work in big organizations live under r
ules that they or their co-workers are never required to more than self-certify their sick leave. Where will the politicians come down on this kind of point when the labor agreement is next bargained?
DON’T FORGET PARATRANSIT
If getting the right mix of efficiency and service is hard when it requires spending on sophisticated technology, as with service scheduling, and even harder when labor rules hamper efficiency, it’s harder still when service level choices involve really sensitive issues of policy choice: service to the disabled. Who wants to grab that high-voltage wire?
To this topic the performance audit brings important facts. In short, Metro’s Access paratransit van services, required under the Americans with Disabilities Act (ADA) for service to citizens who can’t use the regular fixed route buses, finds itself in 2009 providing first cabin service at economy cabin fares.
Metro Access service sees about 1.1 million boardings a year—roughly one percent of Metro’s total boardings. The fare is one dollar. The average cost of a ride to Metro is about $40. Total annual operating expenses are a little more than $50 million, about nine percent of Metro’s total operating costs. In recent years costs have risen at about twice the rate of increase of the Consumer Price Index.
The audit compared Metro Accessto its counterparts in Minneapolis, Portland and Denver. It found that Metro has highest costs per mile, the highest costs per hour and fell short only of Denver in having the highest costs per boarding. It is cold comfort to know that when the Access dispatchers, as they do about six percent of the time, send a private taxi rather than the Accessvan, the average cost at $40 per ride is virtually the same as for the van.
Metro’s own agency staff for Access is bigger than at peer programs in Minneapolis, Denver and Portland. Yet Metro also has not been enforcing productivity incentives and penalties recently entered into with its outside private contractors that provide the 440 drivers and mechanics, as well as administrative personnel and almost 100 people involved in dispatch and scheduling.
Policies set for Access at Metro try to give its patrons the best possible service. Those policies, however, go beyond the requirements of the ADA. Service extends for more hours than required. Service extends to local areas beyond the required proximity to the fixed route bus system More than half of Access patrons are beneficiaries of door-to-door service that may be fundamentally required by their needs (and which counterpart agencies in other cities also provide), yet go beyond curb-to-curb service minimally required by the regulations.
Meanwhile, federal regulations limit the fare for paratransit service to double the cost of a regular bus fare. For Metro, that would be a fare for Access of $3.50, or double today’s $1.75 base fare. That’s a ceiling, not a benchmark, and few other systems charge the highest fare level. But few discount the fare as steeply as Metro.
Indeed, there’s a pattern here, because the audit also commented on the 50 cent discount fare for senior and disabled riders even on the regular buses. Federal requirements and the policy adopted for Metro itself in 1999 stipulate that there should be at least a fifty percent discount from the peak fare. (Disclosure: the writer uses the discount card.) That would be a dollar, not today’s 50 cents, against today’s peak fare of $2.00. In February the discount fare will go to 75 cents against the new peak fare of $2.25. “Almost universally,” according to the audit, “Metro’s regional and national peers do not offer fares discounted as much.”
Eight months ago, well in advance of the audit, Metro management asked the King County Council to review the fare policies for senior and disabled fares on the regular buses. The Regional Transit Committee apparently has put the matter on hold. For the Access program fare, tucked into County Executive Triplet’s proposed budget is a fare increase from $1.00 to $1.25 to take effect in 2011. Its fate is not yet known. Nor can one predict whether the Regional Transit Committee or the county council will be prepared to look further at fare, service or cost issues at Access while the rest of Metro’s portfolio should also be under a magnifying glass for service and cost efficiencies. Given the observations made in the audit, however, it seems hard to see how politicians who say they are committed to policy leadership and accountability can avoid taking up the hard questions.
THE BOTTOM LINE
The bottom line is that Metro needs help from just about everyone. Help from experts who can offer the kinds of insights contained in the performance audit, as we have seen here. Help from citizens who must be deeply concerned about the future of public transit. And help especially from transit riders, who have to balance their critical voices for protecting and improving service with their loyalty and support for the system they so badly need. Help from Metro’s management and union, who can bring fresh thinking to the crisis now at hand. Help from business leaders, who recognize that a strong transit system is fundamental to a sensible overall transportation program, and that the workforce needs the bus to get to work, at almost all hours of the day!
But most of all, Metro needs help in the form of leadership choices on policy issues—and that’s not the same as micro-management –from the elected officials who need to make the hard choices that will improve the efficiency with which Metro spends precious taxpayer and farebox dollars.
The Regional Transit Committee held a regular monthly meeting on Wednesday, October 21. The meeting had just two agenda items. First, predictably, another go-round on service cut allocation and the 40/40/20 canon. Discussion was entirely inconclusive except to hear a formal position from the Suburban Cities Association that there should be no change in the formula. From that flowed consensus that there needed to be more discussion and another meeting. And on it goes, as it has for months. This has to be an enormous frustration for those calling for a fresh approach and asking for some actual decision-making by those whose job is to make decisions.
The other item on the agenda was a briefing on how Metro is preparing to handle snow and floods.
Nothing on the performance audit or any other efficiency topic. That will be in next year’s work plan, we are told.