Innovations for Successful Societies Innovations for Successful Societies AN INITIATIVE OF THE WOODROW WILSON SCHOOL OF PUBLIC AND INTERNATIONAL AFFAIRS AND THE BOBST CENTER FOR PEACE AND JUSTICE Innovations for Successful Societies Innovations for Successful Societies Series: Centers of Government Interview no.: C 5 Innovations for Successful Societies Innovations for Successful Societies Interviewee: Matthew Gallagher Interviewer: Michael Scharff Date of Interview: 23 October 2012 Location: Annapolis, MD United States of America Innovations for Successful Societies Innovations for Successful Societies Innovations for Successful Societies, Bobst Center for Peace and Justice Princeton University, 83 Prospect Avenue, Princeton, New Jersey, 08544, USA www.princeton.edu/successfulsocieties SCHARFF: This is Michael Scharff, the date is October 23rd, 2012, I am in Annapolis, Maryland speaking with Matt Gallagher who is Governor Martin O'Malley's chief of staff. Sir, thank you very much for sitting down with me. GALLAGHER: Sure. SCHARFF: I just wanted to ask if you could introduce yourself in a bit of detail and tell us how you came to your current role. GALLAGHER: Sure, my name is Matt Gallagher, I serve as Governor O'Malley's chief of staff. I have been chief of staff a little bit over three years. When then Mayor O'Malley was elected governor I entered the administration as a deputy chief of staff in charge of operations. I graduated to chief of staff a little over three years ago. Prior to that I was with the governor for almost the entire time that he was mayor of Baltimore and I served as director of his CitiStat program during my 6-1/2 years in Baltimore city. Shortly prior to that I had managed a private sector business transition team effort for the then new O'Malley administration when he first was elected mayor. SCHARFF: Thanks. One of the interesting sort of takeaways is this idea that a governor comes in and there is already an existing management program here in the form of Managing for Results. I wonder if you can help the listener to understand a little bit more about what that program was and perhaps what it wasn't addressing that StateStat was able to sort of come in and do GALLAGHER: Managing for Results was what I would call kind of your typical performance measurement program. Each agency of state government had a handful, maybe half a dozen, high-level outcome measures that were generally measured on an annual basis and were reported out as part of the budget process. The intent was to help inform, particularly on the legislative side, funding decisions at the agency level. Our feeling was-well, we had a couple of feelings about it. One was that it was not really operationalized which I can elaborate on in a little bit. We also felt that it wasn't intense enough to actually move those half-dozen outcome measures because there was very little underlying what I would call performance measurement infrastructure that was actually going to help move those larger outcome measures. SCHARFF: Do you mind if I just push a little more on the details of when you say it wasn't operationalized what does that mean? What are you referring to? GALLAGHER: In terms of-when you're managing a very complex organization and the State of Maryland is a 35 billion dollar annual enterprise, we have a lot of lines of business, we're responsible for a great many things, it is a little bit inadequate to say that you're going to have four to six measures for a whole agency that might be a billion or a two-billion dollar enterprise on its own. It is even more inadequate to think that simply having those four measures is going to in itself kind of move the numbers and move the performance of the agency. Our feeling is that a performance measurement is going to be most effective-you have to have a real infrastructure in place. You have to measure and manage considerably more indicators. You have to increase the frequency of data collection and analysis. It is really kind of the week-to-week, month-to-month grind, managing the data, managing the agency's performance that ultimately will give you the biggest bounce in terms of those big, high-level, kind of macro outcome measures. SCHARFF: Right. GALLAGHER: So what you'll observe with StateStat-at its precursor when we were involved in the city, CitiStat, is really a much more aggressive performance measurement program in terms of the number of measures, in terms of the frequency of meetings, all with an eye toward just being more impactful as a way that you actually manage what goes on in the enterprise. SCHARFF: It is interesting because in some ways, often StateStat is often referred to as a performance management tool. GALLAGHER: Right. SCHARFF: But in some ways we also can look at it from the angle of it being a way to sort of reform the functioning of the cabinet itself and holding not only the cabinet secretaries to account but then helping them to actually manage the people in their various departments. GALLAGHER: Absolutely. SCHARFF: And agencies. GALLAGHER: You know often times performance measurement programs, the ones that I studied back when I worked in the city of Philadelphia, when I was in graduate school, when you work in this field, you would have models where frequency sometimes is quarterly. That was considered pretty aggressive. What most people don't appreciate until they come and actually see the sessions that we have, the vast majority of our agencies are coming in every fourteen days. At the most it is once every four weeks or once a month. We feel like that if you're touching these agencies twelve to twenty-four or twenty-five times a year, you've got a structured environment like where you've got the key decision makers present. You've got a built-in follow up mechanism, you're reviewing data kind of constantly. That it does a bunch of things for you. It accelerates your decision making. It accelerates the feedback loop. It gives you an opportunity to observe where you have management capacity and deficiencies at the agency level. It helps you identify future leaders. It improves the level of connectiveness between the governor's office or the chief executive's office and what actually happens at the agency level. SCHARFF: Can I ask you about the cabinet itself? GALLAGHER: Sure. SCHARFF: How often does the full cabinet meet? GALLAGHER: We have a full cabinet, it is probably about 24 or so different agencies and department heads and then we have a series of subcabinets. Our subcabinets meet much more frequently than our overall cabinet does. The overall cabinet meets a couple of times a year. The subcabinets meet at most every couple of months. In the StateStat meetings, whether it is BayStat which basically convenes the whole environmental subcabinet. We have a homeland security stat session where we basically convene all of the public safety-related agencies. There is probably a higher level of cabinet agency interaction than there has been in the past in terms of the frequency. We don't always call them cabinet meetings, we generally call them stat meetings. SCHARFF: So in some ways the stat meetings, they don't entirely replace-. GALLAGHER: Not entirely. SCHARFF: Replace the cabinet meeting but a lot of the substance of what might normally in many states be discussed in a cabinet meeting is being taken care of in your StateStat meeting. GALLAGHER: Yes. SCHARFF: Of the members of the full cabinet how many are appointed by the governor? GALLAGHER: All, except the head of the Maryland State Department of Education that the School Superintendent for the State is appointed by the state School Board. But the governor makes the appointments to the state school board. So there is a certain level of connectiveness that happens. SCHARFF: I'm curious if we could sort of go back to early in the term and then move forward. GALLAGHER: Sure. SCHARFF: You guys come to office. There are obviously a number of challenges to governing any sort of bureaucracy. And I'm curious-- you've had the CitiStat model-you must be already thinking during that transition period of adopting that model for the state, is that true? GALLAGHER: Absolutely. When the governor-when then Mayor O'Malley was running for governor, it was kind of a plank in his platform that he was going to bring CitiStat to the state, that he was going to bring that same management approach and techniques and apply it to state government. Having been in Baltimore city we had very frequent interactions with state government whether it was the Department of Juvenile Services, whether it was the Division of Parole and Probation, whether it was the Maryland State Police, whether it was the Department of Social Services. We were acutely aware of what we thought were very significant performance deficiencies because often times they impacted us as a local jurisdiction. So when we came into state government and when we were doing the transition from city hall to the state house, one of the things that we tasked our transition teams to do was to kind of prepare us to hit the ground running so that we would be in a position to initiate StateStat meetings very quickly. I think we had our first StateStat meeting, I believe it was between two and three months into the first term. Literally, as soon as we hit the StateStat room we gather, kind of identify some staffing resources that we wanted to apply to StateStat, we were off to the races. SCHARFF: What specifically is one of the transition teams that is focusing on bringing StateStat? What are they looking at and what are they doing? GALLAGHER: This is back in 2006, this is after we won the election. Basically you have a period between the first week of November and I guess the third week of January where you are not in office yet but you are preparing to come into office. We formed a transition team for each of the agencies of state government. The transition team was charged with making recommendations from a financial management standpoint, performance measurement standpoint, an operational service standpoint. And all of these reports were delivered to the governor and the lieutenant governor early in the term. As we ramped up agency participation in StateStat process the early transition report became important, kind of fodder for areas of focus for our StateStat team, kind of flagging issues we needed to get into. Basically we would take the transition team report for those agencies, we would take the audit history of those agencies which was pre-existing and then we would ask the career people in the agencies who weren't going to turn over with the transition in executive leadership, give us all the reports that you use to manage right now. We would take those three buckets of information and we'd try to merge them into what would ultimately become the StateStat template for that agency. What data we were going to expect them to report either every two weeks or on a monthly basis. Then that data, those reports become the backdrop of what we talk about in StateStat meetings. SCHARFF: You talk about some of the challenges that StateStat sought to address. I'm curious if we could look at really specifically what are the key things that StateStat was going to help the governor to do as it applies to governing in Maryland specifically? GALLAGHER: If you go to the State's website you'll find very clearly defined a series of goals that we eventually kind of moved into, fifteen target areas that encompassed a very significant part of the governor's governing agenda. StateStat is the primary mechanism through which we monitor and try to accelerate progress in those areas. More specifically there were certain deficiencies that we were very aware of coming in the door. Things like very significant backlog of DNA testing for people in Maryland who were supposed to have given DNA samples given past activity they were involved in, criminal activity, If they were on parole and probation. We had a series of uncollected samples in the corrections system. We had a series of untested samples backlogged. So this was a performance area that we decided to put a lot of very early attention on. When you go over to StateStat they'll be able to show you some pretty amazing data in terms of what we were able to do in terms of increasing our throughput on processing and then what the corresponding impact was in terms of hits and arrest as a result of clearing the backlog and now getting current. We were, coming from the city, Baltimore city is, you know, an urban setting. It has had some challenges in terms of its management of public safety and crime. We knew very well having run a local police department that the overwhelming majority of our most violent offenders were people who were very well known to the Parole and Probation System which is a state function, it is not a local function. So we knew that if we wanted to positively impact what happened with violent crime in places like Baltimore and really around the state, that there was a great opportunity to do a much better job of focusing in on the most violent offenders, those who were most prone to kind of reoffend. What we did is we refocused our parole and probation efforts with very, very intensive monitoring of kind of the top tier of offenders. We call it the VPI, the Violence Prevention Initiative. We worked with Ph.D.s and real strong data people at the University of Pennsylvania to come up with indicators and profiles that we would apply to the clients that we had in the system. We think that that has been a very significant contributor to the very significant reductions we've been able to achieve over the last six years in violent crime. SCHARFF: So you're talking about crime. I'm curious, whenever we see a head of state come into office there is literally 150, 250 things that they could be doing and need to be doing. All those fires have to be put out. I'm wondering if the government had a particular strategy to prioritize? GALLAGHER: I don't know if I would say he had a strategy to prioritize or not. I think that our initial strategy was to go out and hire a very competent and accomplished cabinet to manage these agencies. I think it was to put in place the infrastructure he felt he needed to manage through the StateStat process. I've used the word operationalize a couple of times, we are a very operationally-focused government. It was the same way the governor was when he was mayor. You really have to get into the guts of what the organization does if you expect them to perform and be successful. So while we're curious about indicators like safe, healthy, happy child-we want to get down into the payroll system so we know what our level of absenteeism is. We want to get down into work order and complaint system so we know what the public demand for service is or what our output is in terms of productivity. We want to get down into the overtime system, also in the payroll system. One of the more significant cost centers of any government is your personnel cost. It is a little bit less so at the state than it is at the local level, but if you can get your employees showing up on time, if you can make them productive within the regular work day, if you can really reduce and minimize the level of absenteeism or employee injuries that happen, the more you have an available workforce, the more you can kind of contain those, what I would call regular time costs. It really improves your financial flexibility. It also improves your operational flexibility. SCHARFF: I'm just curious because one of the big challenges sort of managing from the top, getting all of your agency heads whom you have now appointed, marching in lockstep with what you as the governor want to be achieving. In those first few months and maybe continuing today, was there any one sort of phrase or statement that sort of summed up the governor's vision and that helped everyone in their jobs on a day-to-day basis sort of continue to drive towards a certain sort of focal point? GALLAGHER: Yes, I mean we're not big on mission statement on the back of the business card. That's not our thing. I think that there are probably a couple of different things you could say, what gets measured gets managed. It is just a results-oriented environment where people are expected to produce. That is generally the way that we roll. SCHARFF: And that was clear to the people that the governor was appointing early on. We're about results. GALLAGHER: I'll tell you I sat through every cabinet interview during the transition process. I was like a panelist recommending people to the governor for these cabinet positions. The overwhelming majority of the applicants that were in our cabinet pool knew about CitiStat, knew about his approach to management. Many of them came into the interview saying "I'm really excited to be part of your StateStat process, here are my ideas in terms of performance measurement." It was really his calling card in terms of government operations and management. So most people kind of knew what was coming. SCHARFF: I wonder if the political appointees who had been in place before and even the civil servants from the previous administration, under a previous party were sort of a bit nervous about these changes that were coming down the pike. GALLAGHER: The smart ones were. SCHARFF: Because in many ways this is getting-you're now asking people in your agencies, your civil servants to perhaps be doing more with less. GALLAGHER: Yes, in many cases. SCHARFF: Were there changes to the incentive structure for these employees in these sort of line agencies? GALLAGHER: Beyond continued employment? SCHARFF: Beyond continued employment, any sort of pay for performance incentives, bonus structures? GALLAGHER: Not many. SCHARFF: They're now going to be doing their job when they weren't? GALLAGHER: In some cases. I mean the overwhelming majority of state employees I think are dedicated and hardworking. I think that this helps focus them a little bit more in terms of the things that we've prioritized. It helps them prioritize a little bit more. The reality is that less than a year into office the state, as the country was, was entering the great recession and a very significant downturn in state and local government revenues. So we all had to manage under a lot of duress. In terms of performance bonuses, in terms of incentive pay, while there may have been small efforts or targeted efforts, by and large that was not an option that was available to us. SCHARFF: And that never has been an option? GALLAGHER: If you go back to our time in the city we had incentive pay. We had instances where not kind of your typical definition of managed competition but you would pit solid waste collection bureaus against each other in terms of the ways they managed it over time and complaints. We looked for opportunities to try and spur that kind of internal competition. SCHARFF: One of the other challenges too that a governor faces in many settings is just sort of reporting structure and how it is designed in the sense that in some cases you have your agency heads with direct reporting lines to the governor himself. I'm curious if that was an issue early on and one that you sought to address and if you have addressed it. GALLAGHER: The governor is the governor, he can interact with anybody he wants any time he wants. So he remains in very regular direct contact via e-mail and cell phone with the members of his cabinet. The internal reporting structure, the way that it works is all of the state government is divided up between the chief of staff and the deputy chiefs of staff. Each agency has a direct report within the governor's office. The two dozen or so cabinet secretaries report directly to a deputy chief of staff or a chief of staff. And then the deputy chiefs and the chief report to the governor but the governor also has the ability to go directly to cabinet secretaries. It is kind of on the people in the governor's office, myself and the deputies, as well as the cabinet secretaries to kind of stay in very regular contact so that if the governor initiates something on his own, the deputy chiefs of staff are informed by the cabinet secretary or if the government initiates it with myself or the deputy chiefs, we then kind of push the information out to the agency heads. SCHARFF: So then where are the policy analysts, the analysts on StateStat come into play with this reporting structure and where do they sit? GALLAGHER: The StateStat office sits in the governor's office. The StateStat office reports directly to the governor's chief of staff. The StateStat office every week is generating anywhere from fifty to 75 pages of analysis in support of the meetings that will happen that week, the four, five or six meetings that will happen that week. All of that information, the night before the meeting, goes into my briefing book, goes into the governor's briefing book. So whether or not the governor is able to attend the meeting, whether it is me attending the meeting or the relevant deputy chief of staff attending the meeting, we convene it, we run it and then there is a built-in follow up mechanism where all the commitments that are made in the course of the meeting get memorialized in a follow up memo. The follow up memo goes out to the agency head or agency heads within 24 hours of the meeting. Then they have to report back on those things before the next meeting, whether it is in two weeks or a month. So just to give you an example (hands interviewer memo). This is-I'll show you an example of a briefing memo that comes in and is in my book when I leave last night when we have the StateStat meeting today it is what I will have up in front of me as we're going through and questioning the agencies about what they're doing. But the analysts and the StateStat office report up through me. They essentially service the governor, myself, and the deputy chiefs of staff. They are working with their portfolio of agencies every day. One, to kind of get ready for these meetings and these sessions and then two to manage the follow up that goes on between sessions. SCHARFF: That is actually what I was trying to get at. When bottlenecks and other issues arise following the meetings, they have their sort of marked calendars saying what they have to do. The agency head runs into trouble or the folks in the agency run into trouble are they then sort of contacting their assigned analyst for help who then sort of, if necessary, kicks it up to you and gets the governor's office involved-? GALLAGHER: Often times. SCHARFF: To follow up. GALLAGHER: It depends on the significance of the issue. Some of it might be dealt with at the analyst level. Some of it might be dealt with at the contact with the governor's office whether it is me, whether it is the deputy chief of staff. SCHARFF: So early on when StateStat is just getting formed do you sort of come up with a couple of different pilots for where StateStat will be-and then you'll roll it out on a larger scale? GALLAGHER: Initially we were rolling out an agency every two to three weeks. So the analyst would-we would basically say we're going to have the Department of Public Safety and Correctional Services start February 1st. If you back up from that you know three or four weeks, overseeing point of contact with the deputy chief of staff, whoever is Public Safety, sits down with the agency head, the agency head's management team and the StateStat analyst. They develop the template, performance measures that are going to be this evolving document that we use to measure and track what goes on at the agency level. The agency then kind of comments on it, says these are good measures, these are bad measures, here are some alternatives. You kind of agree on what the product is going to be and then they start populating the data. As you go through the process certain measures get added, other measures fall away, it evolves. It is not like a static thing. It gets reformed, new issues emerge, other issues get resolved, you try to refocus the template as much as you can. SCHARFF: And who is making that decision on the refocusing? GALLAGHER: It is generally a collaborative decision. I mean the governor's office ultimately has the final say but we want agencies to recommend different measures that they think should be pursued. We want there to be a certain level of interaction. At the end of the day though it is the governor's office that ultimately decides what is going to be reported and the frequency. SCHARFF: Essentially what you want to see being done on the ground, the results that you want to be seen. GALLAGHER: Yes, the information that we think is important to manage. Last week, October 19th there was a DBED (Department of Business and Economic Development) Stat, there was an SHA (State Highway Administration) Stat. I go into SHA and I can go through, I can pull up the template, I can pull up the briefing, I can pull up any of the content and materials that happened that day. When I look at this (internet problems). Once I get my IPad working I'll show it to you. SCHARFF: No problem. You needed a state law to be passed in order to get the StateStat program up and running, is that for funding? GALLAGHER: No, it was-. SCHARFF: It would seem like something that doesn't necessarily-many readers and listeners would say this is not---. GALLAGHER: The situation was that Managing for Results had been codified in the law and given what we were doing with StateStat and kind of overhauling Managing for Results, we thought the most transparent route was to kind of update the law so people were aware of what we were doing and to what extent State Stat was essentially going to be substituted for Managing for Results. So coming in the door, being new to state government, we thought it was the most transparent way to kind of approach what we were doing and try to get some legislative buy-in. SCHARFF: Where does the funding for the office come from? GALLAGHER: It sits in the governor's office. SCHARFF: So it is appropriated from that? GALLAGHER: Yes, this is just an example, this is what a template looks like for an agency. This is the State Highway Administration. We like to put the photos of the top managers to kind of really reinforce the accountability of what is going on. You go through and it is a series of measures and data periods and averages and max and min and year-to-date. Then this kind of template which the agency populates every two weeks in this case, we take that and we analyze it. There will be a briefing memo that goes along with it that will say-this will be the related PowerPoint. Then we take all those charts. All that data, the raw data, we boil it down to a series of charts, graphs, pictures and maps which we then project up on the screen in the StateStat room and then those visuals become the backdrop of what we discuss in the course of the meeting. The analyst will work to kind of pull all that information together, analyze it, resolve any data issues so that we can have as constructive a conversation in the course of the meeting as possible. SCHARFF: Early on I'd imagine it was particularly difficult to set goals in the absence of concrete data. GALLAGHER: Initially you would set intermediate goals like this is where we would like to be next month; this is where we would like to be two months from now. The goal-setting process has got to be partially formed by how many resources you have, what the past performance of the agency is, there are a lot of things that get cranked into that. So it is not something you can necessarily do right off the bat. Here are some more examples. SCHARFF: How difficult is it to actually obtain that data and dig it out from within the agencies? GALLAGHER: It is tough for the agencies initially. This is what the memo looks like. It goes to the governor, myself, kind of the relevant key people in the governor's office and then it will have, in this case a bunch of questions. Then there will be some charts that the questions will kind of relate back to, the things that we're going to talk about in the course of the meeting. Then these things get cut out of the memo, put into a PowerPoint so that as we're talking about, in this case, facility locations and the jobs associated with each one of them, this will go up on the screen while we're talking about it. People generally respond pretty well to the visuals and the trends. It just makes a little bit better conversation. SCHARFF: Do the agencies typically have one or two or three individuals who are in charge of getting that data to the analysts in the StateStat office? GALLAGHER: The agencies become very skilled at kind of jockeying the data around and doing their reporting. So the first couple of months it is pretty excruciating for the agencies but when they realize this isn't going away they become much more efficient at it. They become much better at kind of pulling the data, refreshing the data, making sure that it is formatted appropriately. What most agencies gravitate towards is kind of one StateStat coordinator for their agency whose job it is to kind of analyze the data, compile the data, manage the interface with the StateStat office and anticipate where we're going so they can try to get ahead of it. The best is when you've got a really good cabinet secretary, you've got a really good management team, you've got people who are kind of oriented this way and they come into the meeting and they say we know what you're going to ask us, we're on top of this, this is our corrective action plan. Or, I hope you guys noticed like this 20% improvement we had in this area. People can be proactive about it. That's the opportunity. If you work at this, if you've got the right people in place. Some people are a little more passive. They sit back. They just kind of take their punishment in the course of the meeting. That's not what we're looking for. SCHARFF: You wouldn't say that's the normal. GALLAGHER: That's not the norm now. SCHARFF: It was? Because there was that need-. GALLAGHER: People were more accustomed to-. SCHARFF: To change expectations right? GALLAGHER: People are more accustomed to it now. SCHARFF: What brings about that change? How does an executive, in the absence of essentially just keeping their appointment, what-? GALLAGHER: You create a series of expectations and then you demand they meet those expectations more often than not. If they don't, without valid reason, without valid explanation then there have to be consequences either from an appointment standpoint, funding standpoint or responsibility standpoint. That's how this stuff works. SCHARFF: Are there consequences? Just to play devil's advocate. GALLAGHER: Sure. SCHARFF: I can imagine many appointees are probably confidents, friends, colleagues, have a professional working relationship with the governor. You don't necessarily want to trim agency funding because it can have a negative impact on your programs. GALLAGHER: Often times if you're trimming the funding it is to give it to somebody else who can get the job done. So you move duties and responsibilities around. In terms of-I don't want to oversell the consequences and because it is the exception, it's not the rule. But if an agency is not performing, if you are meeting with that agency literally every fourteen days so you're meeting with them around this stuff and there are performance issues 24, 25 times a year, if the performance is not turning around, if you've tried new strategies, if you've tried additional resources, if they continually demonstrate inability to get things done, you then have to make changes, either at the agency head or within the management team. When you look at the composition of the cabinet, we have had to make changes in the composition of the cabinet both when we were in city hall and when we're in the state house. Getting to these-when you're managing an enterprise this large, eventually you stop grading on effort, you actually have to produce. SCHARFF: We have a section in the case study called overcoming obstacles. Essentially we're letting the reformer tell their story of how they-why they set up a program and how they set up a program. The program is running its course but then all of a sudden there are hiccups and unexpected hiccups at that. It sounds like perhaps the need to change out certain personnel in order to not only get results in that particular agency but also signal to others in the enterprise that I'm serious about what I've set out to do is important. In as much detail as you can or are willing to share, I'm curious if we could try to build that out a bit. I think it is particularly instructive. GALLAGHER: If you're thinking about it, you've got 24 or 26 cabinet agencies, big major enterprises that you're running. Let's say that each of those 25 kind of secretaries who are responsible for these agency enterprises, let's say that each of them have five key appointees, five key staff members that help them manage everything they do. So now you're talking about six times 25 and that's 150 people. Let's say that you vet 90% in terms of hiring the best, the brightest, work hard, smart, all that type of stuff. That still means that you've got 15 people that aren't going to cut it, that aren't going to be successful. That looked good on paper but couldn't get it done or had performed really well in the past in similar type jobs but for whatever reason were not able to get it done. That creates a potential very significant deficiency in terms of your overall management of what goes on in state government. It's not because they're bad people, it's not because you're mean, but often times you have to make midcourse corrections. What happens with a program like StateStat and CitiStat before that is you're observing them, in a lot of cases much more directly than you normally would and with much greater frequency than you normally would. So you get a sense of whether or not things are getting better. You get more immediate feedback and you get to see how they interact with the governor's office, with their colleagues and peers at the agency level and whether or not they're capable at that point in time of getting something done. SCHARFF: Is there a specific example you could offer of where you did need to go and make that midcourse correction with a particular agency? GALLAGHER: Yes, there have been multiple examples of where we felt like we had to go in a different direction with the leadership of the agency or deputy secretary or assistant secretary where we just didn't feel that they were going to be able to advance the mission in ways that we thought were necessary. SCHARFF: We can always scratch things from the official record. GALLAGHER: I'm not going to get into the specifics of which cabinet secretaries were fired or that sort of thing. SCHARFF: What would be the issues though? GALLAGHER: Performance issues. I mean we had a cabinet secretary who had a very operationally-intensive job, had a large clientele of at-risk people that they were responsible for managing. We just did not see the level of responsiveness that we expected. We did not see some of the outcomes being driven in the direction we wanted them to be. Lost confidence; we had to make a change. SCHARFF: Governor Daniels told us that in that case, a very similar case that he faced, there is a sort of probation period when trouble started and policy analysts in his office played a mediation role in trying to get things up to speed. In this particular instance that you mentioned is there an effort to support, to really dig in, to try to find the source of the issue? GALLAGHER: Very rarely when someone is removed from their position in state government, city government, is it because of one individual thing. It is almost always cumulative. There are usually a number of opportunities to try and revive the situation or to bring it back on track. Just like anything else it is disruptive to have to make changes. It is tough. When you're deciding to kind of go in a different direction there are cost implications, there is kind of continuity of operations implications. They are not decisions that are made lightly. This is all true-I don't have any reservations about any of the things that I said but proportionally you have to understand the frequency of which these types of things happen. SCHARFF: In terms of ensuring that that data that is coming up from your agency is clean, what are the checks that the office has to ensure that that data is-? GALLAGHER: Data accuracy is always very, very important, depending on the data source, depending on the indicator. You sometimes have audit functions associated with that. Things like crime data goes through multiple levels of review. The other thing is that often times the frequency with which we're pulling data you will pick up anomalies sometimes and you'll go back and realize somebody transposed a number or there was an erroneous-we didn't pull all the fields we were supposed to pull for a particular two-week period. What happens is if you're looking at two months of data so you have four data points at every two weeks, you'll realize that hey, we've got to go back and do something with that particular number. We also try to marry things up with the financial indicators as much as possible. The systems that always work in any government, at any level, are the payroll system and the financial systems. You know what you spent, you know which people came to work. You know how much they got paid. If you can kind of relate a lot of the performance measures back to those things that is also a very good way to kind of validate what is going on. SCHARFF: Do audits take place at any particular agencies? GALLAGHER: They do. I mean in terms of-if you're the Department of Health and Mental Hygiene or the Department of Juvenile Services, you've got case management functions for the clientele that you're servicing. You've got built-in audit functions with case managers as opposed to case officers at the front line. SCHARFF: But I can imagine-. GALLAGHER: You can always audit more. There is always an opportunity to do that because you're dealing with a lot of data and you really want to make sure that you're managing the integrity of that data. SCHARFF: You were talking about the visual representation in these meetings. You obviously have a lot of data that you're dealing with coming into a meeting. What is the strategy in terms of what we actually make visually available? It is pretty important. That sort of will drive the conversation. How much gets buried in the templates versus how much actually goes up on the screen and who is deciding that? GALLAGHER: Generally the analyst is deciding it in terms of the preparation for the meeting. The people who are driving the meeting whether it is the director, whether it is the governor, whether it is me, we might say I'd like to see a chart on this, I'd like to see a graph on this. Or can you pull up this map? They have the ability to kind of improvise a little bit in the course of the meetings depending on the meeting, depending on the content. It is much better to say on the chart, graph, picture or map than it is in a paragraph. People get it much more easily. If you go to the StateStat website, the briefing memos that are prepared for each meeting which are kind of a Reader's Digest version that is prepared, that strips out some of the more identifying information, more sensitive information but leaves intact most of the charts and graphs and pictures and maps. So that if you are John Q Public you can log onto the website, you can take a look. You can click on the BayStat meeting or the Public Safety meeting and it pulls up the summary of each sheet that we talked about, what was discussed in the course of that meeting. SCHARFF: From a social science standpoint it is often times a challenge to link, to make that correlation between a particular management tool leads to this particular outcome. Are there any specific examples you can give that would actually highlight-there was a pretty strong link between StateStat, that focus on management and results actually produces this outcome. GALLAGHER: There are a lot of results. I mentioned the DNA backlog, that's one. In the course of the BayStat meeting today. We should talk a little bit about BayStat. It is somewhat of an atypical meeting in that the vast majority of our sessions are focused on an individual agency and they are focused on what that individual agency does, the outputs and the outcomes that that individual agency largely controls on its own. BayStat is a much more collaborative forum because when you're talking about the health of the Chesapeake Bay you've got a number of agencies that are contributing to that outcome. You've got the Department of Agriculture, the way that they regulate farming practices. You've got the Department of Natural Resources who are managing public lands and waterways. You've got Maryland Department of the Environment which is responsible for enforcement activity which is responsible for regulations that relates to water and waste water and processing. You've got the Department of Planning which is responsible for land use and land use management. Each of those agencies contribute to the health of the bay. So you can talk to Phil about this when you sit down with him, but he can describe for you kind of our Bay Improvement Plans, our watershed improvement plans, the literally dozens of strategies that we're employing right now to try to improve the health of the Chesapeake Bay. While that sounds good in theory, every year we have to go back to the EPA (Environmental Protection Agency) and say here are our two-year milestones for nitrogen reduction. This is how you quantify what you're doing for the bay. Each of these agencies have strategies that they are responsible for which kind of contribute to this thermometer of like cleanup. So the Department of the Environment is responsible for this much, the Department of Natural Resources is responsible for this much. The Department of Agriculture is responsible for this much. We literally monitor what their success is with all of their strategies. So one of the important things, like you'll hear today probably about cover crops. The Department of Agriculture goes around, takes all the farms in the State of Maryland, signs them up for cover crops so that they're planning these cover crops which absorb the bad nutrients so that they don't run off the farm into the waterways and find their way into the bay. The better job we do of signing up acres of cover crops each year, the less pollution finds its way into the bay. So a big indicator both in terms of cost-effectiveness and pollution prevention through the Department of Ag(Department of Agriculture) is the success of our cover crops program. For the Maryland Department of Environment, where are we on the upgrade of all the wastewater treatment plants in the state? Where are we in terms of replacing outdated septic systems, particularly in the watershed areas? These are things that directly contribute to the amount of pollution that goes into the bay. So all of those things kind of work together; and in our case EPA said we've got the best WIP, Watershed Improvement Program, of any of the other states that abut the Chesapeake Bay. We have hit our two-year milestones in terms of nutrient reductions that are going into the bay and we're constantly trying to quantify those things and make sure it rolls up into the next goal in the next two-year period. SCHARFF: In terms of the governor's time and his schedule, has StateStat in your opinion enabled him to sort of free up time that otherwise would have been spent tracing his appointees? GALLAGHER: Yes. SCHARFF: Is it a consolidated way of getting it done? GALLAGHER: It is a pretty efficient way. It is not like he is bumping into a cabinet secretary at an event and they're grabbing five minutes between the cocktail hour and the dinner. He is getting a lot of very substantive analysis a couple of times a week about what his agencies are doing. He is having a structured meeting with the agency head and his cabinet secretary, with their management team, about what is going on in the agency. When I joined up with Governor O'Malley when he was mayor, I had previously worked for Mayor (Edward) Rendell in Philadelphia and I was skeptical that they would want to spend this much time in these performance measurement meetings. But what happened was people said this is a good use of my time, I get a lot done. There is built-in follow up. This is how we're going to do things. Then there was kind of a proliferation of more and more Stat processes because people just thought it was a good way to spend their time. It is very intensive for the executive; it is very intensive for the people who work in the governor's office or the mayor's office. But at the agency level you might be talking about-if you're a cabinet secretary, two hours every two weeks in terms of the actual meeting. There is prep time involved as well. That is really not that bad in terms of boy I get to sit down with the governor or the chief of staff. I know I'm going to have that meeting twice a month. I know I have a forum to bring these issues up. It is an opportunity for a forward thinking manager at the agency level. SCHARFF: What percentage of these actual meetings is the governor sitting in on? GALLAGHER: He goes to the vast majority of the BayStat meetings although he is in New Hampshire today for the President so he won't be there today. Then it varies really from agency to agency. There are some which he is very intensely involved in; there are others where he very infrequently goes. It just kind of comes down to the time management issue. But every Stat meeting goes on his schedule. So when you look at his schedule and when he looks at his schedule he can pick and chose when he wants to come over. The StateStat room as you can see is a block away from his office in the state house so he is able to frequently pop over and participate. SCHARFF: And if he is not chairing the meeting it is-. GALLAGHER: If he is not chairing it is myself or the overseeing deputy chief of staff or in some cases it is the director of StateStat who is a member of the governor's senior staff. SCHARFF: What sort of message or what sort of signal do you think that his presence or his lack of presence at some agency meetings sends to those folks in the agencies? GALLAGHER: I think that any time you get the governor or a senior member of his team in these meetings it sends a message that this is really important. In terms of the handful of agencies where he is not a frequent attendee it is more so just that they got their stuff together and things are going away, that it doesn't demand the same level of attention of the chief executive. SCHARFF: Or what that agency is working on is not necessarily one of the governor's topmost priorities? GALLAGHER: Well, I think you can say job creation is one of the governor's foremost priorities right now, no question about it. But does he go to the Department of Business and Economic stat meeting very frequently? No. A lot of it is those stat meetings are very focused on pipeline deals that we're working on the economic development front. They're productive sessions in terms of figuring out incentive packages and outreach plans and those kinds of things but generally his most effective intervention is probably not in those meetings, it is kind of dealing with the follow up from those meetings, like when he has to call the CEO of a company, when he has to go visit a Marriott or a Bechtel or something like that. Separate and apart from StateStat you can't get away from the fact that there is one finite resource and that is time. There are seven days in a week; there are 24 hours in a day. You have to figure out how you are going to cut that pie up and maximize the time that you have in terms of your own effectiveness. We find this to be a very useful exercise. We dedicate and devote a lot of time to it but we've also got to be smart about how we spend the governor's StateStat time to get the most impact. SCHARFF: Absolutely, that makes sense. Is there a strategy then to-I would imagine that if I'm an agency head and I know the governor is always going to be there at my StateStat presentation that I'm going to be damn sure that everything-I'm going to cross the T's and dot the I's right? But if he is not necessarily there and there isn't that sort of-I don't want to call it heavy-handed pressure from the top, but that sort of expectation to deliver may be somewhat lessened. GALLAGHER: Maybe. SCHARFF: That's a natural extension of the fact that you do need to manage his schedule, he can't be everywhere. GALLAGHER: But it is also probably the question of do some cabinet secretaries fear him more than they might fear the chief of staff or some deputies that they have to work with every day? SCHARFF: What do you think? GALLAGHER: It depends. SCHARFF: Is there any way for you-I mean when you're chairing these meetings how do you exert pressure, a healthy pressure, how do you do that? GALLAGHER: By cabinet secretary saying we think this is going to take us three months to figure out or it is going to take us three months to implement this, kicking the tires on that kind of statement, seeing if they're being not aggressive enough because they want to under-promise and over-deliver. You've got to know where to prod, you've got to know where to offer support. You've got to be confident enough in your knowledge of what goes on in the agencies and their capabilities to push them more aggressively, to tell them you don't agree with what they're doing. There is a lot of give and take. You're going to be here today and you're going to be here next week or the week after and you're going to see a lot of individual snapshots of meetings. Some might be lovey dovey, everybody is on the same page. It might be celebrating some recent wins, everybody feels pretty good. Then there could be some really bad meetings where an audit report has come out or there has been a really bad story in the newspaper or some type of deficiency has surfaced. We don't have the ability to manage that in terms of people coming in and watching from the outside. You get what you get and you can't get upset. It just depends. Some people-there is a guy at Harvard, at the Kennedy School, Bob Behn, has been here dozens of times across city hall and the state house. He makes a point of coming back and seeing the same meetings over and over again so he can monitor the progression of the issues; he can monitor whether there have been management changes. I think what sometimes people miss when they kind of drop in for the snapshot is that this keeps going on and on and on. It is almost like the baseball season, it is very long. You have a lot of ups and downs, you have winning streaks, you have losing streaks, you have good runs, you have bad runs. It is just a lot of variability there. SCHARFF: Can I ask you about the actual cost. We mentioned before the funding is from within the governor's office but what is the actual dollar amount spent on running the program? GALLAGHER: I don't have the number off the top of my head. Basically we generally run with anywhere from six to eight staff people. Beyond that they've got computers, they've got Blackberries, they've got Ipads, they've got modest equipment cost in terms of projectors and that type of stuff. So 95% of the cost of the program is the staffing. SCHARFF: Salaries. GALLAGHER: Yes. Beyond that agencies spend time on this but it is not like they've gone out and hired new people to manage this. They've basically designated existing staff people to be responsive to it. So we have always reported very low cost associated with the operation of the program. SCHARFF: When you went to hire initially those individuals to serve as the analysts in the office, what was that recruitment process like? What did it look like? If someone else were in your shoes and setting up a similar office, how should they go about that? GALLAGHER: We have generally looked at our Stat employees as people that we're grooming for other positions in government. The general path for a StateStat employee is two to three years. They become subject matter experts, they become very familiar with the governor and the governor's team and then we like to spin them out into more significant management positions. In some cases they might end up going to grad school depending on their age when they come in and what their background experience is. Generally very, very accomplished academically. Generally have had a couple of years' of experience so it is not their first job. In some cases we'll find people that we really, really like at the agency level and we'll bring them in. But it is not like we go out and recruit and say hey we've got an opening and we post the job posting somewhere. It is much more that we find you. We decide that we think you're kind of a Stat person and then we send you over there. Sometimes even if there is not a vacancy if we find a really good person that we like and we think can help the team we bring them on. SCHARFF: I would imagine that because the staff tends to be pretty young on the StateStat side and they're being asked to work with relatively senior, accomplished figures who are the Secretaries and even the second tier within the agencies, that there might be an inherent tension that may arise just from the fact that there is this age difference, that you have a 25 year old or a 26 year old in the governor's office saying "Hey Mr. Secretary or Madam Secretary, I need this information, the governor wants it." GALLAGHER: Yes. SCHARFF: And the Secretary saying, "Not so fast, I didn't get anything about that." GALLAGHER: Can happen. SCHARFF: It does? GALLAGHER: Yes, absolutely. SCHARFF: What does that look like and how do you-I guess I'll first just ask what it looks like and then I'll ask how do you deal with it. GALLAGHER: You just described it. When I ran CitiStat in city hall I was 28, I ran the program. When the mayor wasn't there I was speaking for the mayor. When the analyst is in the meeting the analyst is speaking for the governor. Really they're speaking for the StateStat director who is speaking for the governor. The test that you have to apply in those situations is, is the person being reasonable? Is the person being respectful? You're not asking a 25 or a 27-year-old analyst with an MPP from the Wilson School and a really smart person to go in and butt heads continuously with the agency head. You're going in and you're asking them to be reasonable, to help advance the governor's mission, get the information that they need. You want them to try to work constructively with the Secretary and with the agency head and the agency management team. If it reaches a really bad conflict either the analyst can elevate it up to their boss in the governor's office, either the StateStat director or me or the deputy chief of staff, whoever sees that agency, or the cabinet secretary is going to pick up the phone and say, "I don't like how this person is dealing with me, I think they're being unreasonable." Then you evaluate it. The way I-what generally happens when people complain is they say, they're being too demanding, they're asking for too much data. Then I say, okay, what's the data they're asking for. Then I'll ask why do you think that data is not relevant? How are you managing the absence of that data, like explain it. If they can, say fine, you don't need to collect that, you don't need to report that, we'll go in a different direction. But more often than not you end up backing the analyst. Look, some people approach this and say, "Hey, I'm the subject matter expert, I'm the cabinet Secretary, I've been here for 30 years. I've forgotten more than this individual analyst knows." SCHARFF: Sure. GALLAGHER: But this stuff is not that complicated. I mean, if you're a good generalist, if you're smart, if you're perceptive, if you've got good judgment, if you've got common sense-and I know common sense is not so common. Are people working hard? Are they acting in a way that is consistent with what the governor's goals are? Are we getting the type of performance we're looking for? The learning curve on the analyst's side is not that steep. SCHARFF: Can you, just for color in the case, can you elaborate at all on a particular-without using names-a particular instance where that tension has arisen? GALLAGHER: It is not uncommon-. SCHARFF: What was the issue? GALLAGHER: I'm trying to think of a good example here. Public Safety Agency head, 30 plus year career in law enforcement, and the secretary's analyst was a 25-year-old Stanford grad with a Master's degree, strong-willed young woman who was not shy about pursuing a very aggressive agenda of modernizing technology, using all the tools that should have been at the agency's disposal. They didn't get on great but it didn't matter. We had some things we were trying to get done in that agency. SCHARFF: It is sometimes a personality clash? GALLAGHER: Sometimes. SCHARFF: Is it the agency head saying I just don't like being talked to this way by this know-it-all 25-year-old? GALLAGHER: I don't want to oversell the conflict because it is not the main thing that happens, but when you're dealing with this volume of information, you're dealing with this many transactions in terms of people dealing with each other, there is invariably going to be tension. It is not generally personality-driven, it is performance driven. We're trying to get some things done. SCHARFF: So what did you do when the issue erupted between these two individuals? GALLAGHER: Told the agency head to deal with it, StateStat is not going away. That this is the way that we manage and that his future success was dependent upon his adapting and evolving as a Secretary to a more data-driven approach to managing his agency. SCHARFF: And he did? GALLAGHER: No. SCHARFF: And he's no longer there? GALLAGHER: No. SCHARFF: Are there other costs, not necessarily monetary? GALLAGHER: Time. SCHARFF: To implementing this when it comes to the governor's schedule, so time is one. GALLAGHER: The opportunity cost that you could be doing other things with your time. SCHARFF: Is there a threat that perhaps the focus on a couple of key agencies and a couple of priorities lessens the focus on other-. GALLAGHER: But you can always adjust. SCHARFF: Yes? By elevating some of those issues from other areas? GALLAGHER: I'm going to move this agency to monthly as opposed to every two weeks because somebody else needs more attention. Some agencies actually graduate from coming in every 14 days to coming in once a month. It is indicative of good performance, being responsive, that they don't need the same level of attention. SCHARFF: It must be tough too-when you have your civil servants, your career bureaucrats in these agencies. A governor comes in, these are my priorities for this agency, this is what I want you to work on. That governor leaves. A new governor comes in, these are my priorities for this agency, this is what I want. It happens. GALLAGHER: It happens in every state with every change in leadership. SCHARFF: So it is sort of trying to get everybody oriented around the same set of goals. You mentioned before there really is no incentive structure in terms of pay--. GALLAGHER: Not right now there's not. SCHARFF: Is there a performance-what does the performance management system actually look like? GALLAGHER: From a pay standpoint? SCHARFF: No, just from a management standpoint, evaluations? GALLAGHER: There is a performance evaluation program twice a year. It does factor in to whether or not if you're on the civil service track you get step increases and longevity type pay, so there is a little bit of pay impact, but not like what you would generally think of as pay for performance like where somebody would be eligible for a larger bonus and things like that. It is more whether or not you're going to get that 1 or 1.5% increment, those types of things. SCHARFF: Is there obvious linkage between what you guys are measuring in StateStat and what you're measuring on those individual employee performance evaluations? GALLAGHER: Not that would be memorialized in the performance evaluation program. To the extent an agency is very focused on these things, to the extent the agency head makes it part of their process, it could factor in in a good or bad way. But it is not memorialized in the performance evaluation program within the Department of Human Resources or the HR functions of the individual agencies. SCHARFF: Might it be helpful to have something like that? GALLAGHER: It could be, but honestly the last four years we've generally been managing in a furlough environment, and a pay freeze environment. So it is not a bad long-term ambition but it is also not at or near the top of the list. SCHARFF: Thanks, I appreciate your time. Innovations for Successful Societies Series: Centers of Government Oral History Program Interview number: C 5 ______________________________________________________________________ 21 Use of this transcript is governed by ISS Terms of Use, available at www.princeton.edu/successfulsocieties