Morley Winograd on Measuring the Performance of Local Government

Morley, this interview takes place at a time when the City of Los Angeles and other jurisdictions like Chicago are attempting to apply metrics to their budgeting decisions. Provide the purpose and context for Performance Budgeting.

The increased use of technology by the younger generations and people's experience of the efficiency that technology can bring have caused a disconnect between the kinds of services you get from a government and the kinds of services you get in the private sector. The theory is, if we can bring a greater understanding and a greater use of private sector techniques and technologies to government, we could make it more productive, lower its cost, and keep taxes from going up. I think that's the central motivation behind the efforts that the Mayor of Los Angeles is trying to put in place. 

David Osborne’s “Reinventing Government “ work helped begin a restructuring government revolution. Baltimore and New York, and now Chicago, have adopted results–oriented budgeting. And LA Mayor Garcetti would like to as well. Share the underpinnings of this outcomes approach.

There are two prongs of this approach. One, as you said, comes from the writings of David Osborne on how to do budgets based upon priorities, not upon prior expenditures. I know that the people working for Mayor Garcetti—both Deputy Mayor Rick Cole and Bob Stone, who actually used to work for me in the world of federal reinventing—are deep into that process.

It wouldn’t be the first time that process was used in Los Angeles. Mayor Hahn did at least one if not several budgets using that process. Interestingly enough, back then the deputy mayor worried about that was Doane Liu, who is now a deputy mayor for Mayor Garcetti, although not in charge of this process.

The second prong is generally captured by the word "stats." If anybody has watched "The Wire," there is a dramatic portrayal of the use of CompStat in Baltimore. Obviously they started in New York with Police commissioner Bratton who, luckily for us, was also Police Commissioner in Los Angeles. That's where the first notion comes from, that if you had enough data analytics and enough computer-generated data you could pinpoint the areas of trouble that needed the application of municipal services to fix and allocate your resources more efficiently.

Walk us through how what you're referencing gets applied. 

In a city services context, as opposed to a crime setting, the notion of city stats is a regular (as in every three weeks or so) review of critical metrics. These are the things that Mayor Garcetti, for instance, intends to hold city department heads accountable for. You have a regular review of performance against a set of metrics, and you examine in some detail where it's working and where it's not working.

Because the meetings are frequent and because everybody who can fix the problem is in the room, from a deputy mayor on down to the department head and his people, you have an opportunity to try things, learn from them, and try something new in three weeks instead of six months. When they do work, you can adopt your processes to incorporate what you have learned as quickly as possible. So that learning loop feedback as well as the degree of accountability the system brings to that process of governing and delivering services has been the reason it has been successful in many cities in the United States.

What are the obstacles to the adoption and execution of these stat processes? 

First of all you need good data. Unfortunately in Los Angeles, even though we have a 311 system for calling in non-emergencies, it doesn’t properly capture the detail down to the individual calling and the tracking, and the data systems require that. LA has to use substitute data and/or department-only data that isn't as effective as, say, some of the systems in Chicago, Baltimore and New York.

How do those work?

In those situations where you have a full-blown stats process, it begins with a constituent calling in a problem. "There's a pothole on my street. I would like it fixed." The call triggers a work order to do that. It also triggers a tracking process in the mayor's office. Communication continues with that citizen until the problem is resolved. Now obviously if you think about LA—and that it doesn't even have a budget capable of fixing the sidewalks for this century—you could get a lot of calls with no response. But in other cities where this is in place, the calls do trigger a response, and they do determine the budget based on the volume of work that the departments have to fix the problems as opposed to a political decision that is the way those decisions are made in Los Angeles.

You referenced CompStat on "The Wire." It's been used in police work in many cities. Talk about how it works perfectly when it works in that context. 

In a city context, it means that everybody becomes the eyes and ears of the city's government, particularly its mayor's office, in making sure that services are delivered as promised. First, the departments have to set expectations, and then they have to live up to them based upon whether or not they fixed the pothole on time. Instead of having a large audit department—inspectors and reviewers; checkers checking the checkers—every citizen becomes empowered with the ability to be those eyes and ears. And when that works, city services improve as dramatically as the reduction in crime that Los Angeles has experienced, for instance. 

You say that this was attempted in Mayor Hahn's first term as mayor. What led to its adoption, and how did it unwind after he left?

Well, of course, Mayor Hahn grew up in a family that thought of holding elected office as a public service responsibility. So he was very focused from the beginning on how to improve the satisfaction with the government's delivery of services in LA. This became the methodology that he and his team found that held out the most promise. The promise was fulfilled in many cases by looking at places like Baltimore and New York where it had been put in place.

It was somewhat hamstrung in Los Angeles because the council never approved the technological resources required to make it a truly data-driven process. It had to rely upon substitutes and other tools, and a lot of that had to do with the fact that today, in the City of Los Angeles, if you want a responsive government, you call your city councilman's office and they call half a dozen people in City Hall until the problem gets fixed, and then they report back to you what a great city councilperson they are and ask for your vote. This eliminates all of that from the process, which is why it was met with such hostility by the council.

In all fairness, to understand Mayor Garcetti's focus on this objective, he was in the council and he was the council president for most of those years where this program was unwound. 

It was unwound during the two terms of Mayor Villaraigosa, who abandoned the program as soon as he took office. He also came out of a council background, and he substituted for it reviews of departmental performance every six months and formal meetings, which don’t lead to much in the way of accountability or quick fixes. We've all been to presentations where the presentation is skewed towards a particular perspective rather than the reality on the ground.

In the city hall right now, you have a number of people who have budgetary, oversight responsibility. The CAO, the legislative analysis, the controller, the council—why aren't they sufficient in directing their energies? 

On the controller's side, we all saw the difficulty with that. They're allowed to make audits but not to do anything about the audits, and, unfortunately, it hasn't led any controller to the mayor's office because it's a lot of noise but not a lot of resolution.

Interesting you mentioned the CAO, because under Mayor Hahn, when he moved the stats process into the world of budgets—which, to his credit, is what Garcetti is trying to do in trying to evaluate department heads against these objectives—that's when you really ran into the buzz saw. The CAO eventually quit over Hahn's insistence on the budget side of the process, and, of course, the CLA, who I don't think lived anywhere in Los Angeles County, is no longer with us. The new one might have some different attitudes. But back then he didn't. He actually thought he ran the city. From a city councilperson perspective, this alternative budget process is a big threat.

This is the issue that will potentially define the mayor's relationship with the council because, as has been quoted in the paper, Deputy Mayor Cole sees this as the most logical way to run an organization—that every organizational head, be it a CEO or whatever, has the ability to hire and fire its department heads, in this case its upper management in the private sector. And of course the Charter gave the mayor this newfound power, with the exception of the CAO, to hire and fire. But it did not give the mayor the budgetary authority that the council absolutely holds. They do hold the power of the purse.

And so when these performances issues come up against budgetary decisions, councilmembers tend not to say, “I'm not going to give this program and this guy any more money because it's not working based on the data we have.” Instead they say, “ That program and that guy is a friend of the city council president whose support I need so lets give them what they are asking for. “Politics enters the arena and, at least in Mayor Hahn's case, politics won. 

Chicago and Mayer Rahm Emanuel are cited often as attempting to pursue this stats agenda. Tell us about what precedence and value their experience has for California and Los Angeles.

When Rahm was in the White House, he knew and learned about CompStat through our Reinventing Government Initiative that was run by Vice President Gore, and he was in charge of crime policy issues, cops, and the number of cops, and so forth, for the president. He figured out that this was in fact the way big cities can reduce crime.

When he got to Chicago, he not only found a functioning 311 system with full data analytics, but he also brought an understanding of the power of the tool. It'd be interesting to see how it works in Chicago because, as we all known, Rahm Emanuel also brings one of the finest political minds in the country, if not the world, to the process. How he deals with the attention and power of the mayor's office with an equally powerful governing board in Chicago—it's not a council, but it's equivalent except much more partisan’—will be an interesting story to watch. I don’t know how it's going to unfold.

Some critics would say that politics isn't about efficiency. It's not a matter of building widgets faster and cheaper—it's the process of allocating scarce resources democratically. Stats focused on efficiency is not a useful metric for how we ought to govern ourselves. Your response? 

There is a lot of truth in that comment. I think that ultimately a politician who can hold the line on taxes, which is the source of revenue, and deliver better services has a powerful political argument, particularly when they can display the data in a public and transparent way. I think it has political potential.

You can argue that it hasn't always demonstrated that political potential. I know when I was working for Vice President Gore and we published metrics, people said, “oh you're giving your opposition the ammunition to defeat you with!” The people making that argument were his advisors, so there was certainly some power in that argument. His response always was, “well they'll attack me no matter what. At least in this way we'll argue facts instead of opinions.” And I think the fact that the Mayor of Baltimore, Martin O'Malley, is now the Governor of Maryland and a putative presidential candidate indicates it does have some political power if you can really prove you saved taxpayer money without harming your services. 

You've been involved with these kinds of services for decades. Share with our readers a little bit about your bona fides on this matter. 

I was the director of the reinventing government effort under Vice President Gore in the second term of the Clinton-Gore administration, and within the first six months of my arrival there we had a chance to visit CompStat in its early days. We became convinced it was the single best reinventing government idea we'd ever seen. And so we spent a good deal of time teaching police chiefs around the country how to do it.

We had a chance to audit the places where it was effective. I’ve studied and written about the idea ever since because I think it brings a whole new level of accountability to the government and the people who work for government, and it also gives them an opportunity to learn how to do their job better.

What has to come together in the early months of the Garcetti administration to have this priority of theirs, this stats-based budgeting process, gain momentum and win the day? 

The mayor's office and the mayor himself will need to sit down with his former council colleagues, as well as with the new city councilmembers, to talk about the mutual benefits that might flow from a revision of the budgeting process.

So there have to be ways this is mutually beneficial. For instance, the council is interested in getting people to approve more revenue for fixing the sidewalks and streets of LA, which need it. I think a skeptical public might be convinced to support it if you can demonstrate that the current system is operating as efficiently as possible, or at least close to that goal. Without that kind of data it's just an argument about whether you're doing it right or not.

People can always point to an error or mistakes. I think there are some ways you can come to those mutually beneficial outcomes, and only if you have an understanding like that amongst a majority will the ultimate budgeting process work. After all, the council is the one that holds the purse strings.