8 thoughts on “Five Ways to Make Government in Hawai’i More Effective

  1. I probably have a lot of comments on this article, but I’ll just share one for now. Abrecrombe started a modernization project of the state financial systems when he was in office. His vision or the vision at the time he was in office, was to buy an all-encompassing system that could do everything from payroll, to accounts payable, procurement, budgeting, and accounting. The benefit would be that since everything is in one system, there would be no need to interface systems together. Information could be at the touch of a button. But when Abrecrombe was ousted and Ige came in, Ige scaled everything back down to payroll only. If I remember correctly Abrecrombe had like a 6-8 year plan to roll the entire system out. My commentary has less to do with which plan is better, but more to do with the fact that it’s really, really tough to have long term planning in a government setting. Because you never know if you will survive to see any fruits of said planning, especially since a 6-8 year plan, may actually be an 8-10 year plan in government terms.

    Oh I didn’t read the entire article, but that payroll modernization part caught my attention.

  2. My commentary has less to do with which plan is better, but more to do with the fact that it’s really, really tough to have long term planning in a government setting.

    I heard this point made several times, either by politicians, public administrators, or professors. One of the reasons is that when a new administration comes in, they want to leave their own mark and legacy. Therefore, continuing a plan or vision from their predecessor is unappealing and likely won’t occur. (One person made a comment about all the plans that were getting cobwebs in some storage, or something to that effect.

    When I heard this, I found this disconcerting, as some important work require more than 4 years. One person recommended having 6 year terms, as a way to address this. If the person could win two terms that’d be 12 years, which isn’t perfect, but a decent compromise.

    Don,

    Do you have any thoughts about Abercrombie’s plan–i.e., did you like it or not? Also, did you read the section on building this type of software in house as opposed to purchasing commercial products?

    I have mixed feelings about this. Off the top of my head, I’d say if the government IT department is really good, then building programs in house seems like the way to go. One drawback is that the commercial products may have worked out kinks and figured out design issues that are the most user friendly. The in-house IT guys might not only have to start from scratch, but they may lack a lot of knowledge about user experience and needs. This could be solved if government workers who use the program communicate really well, but I’m not very optimistic this would occur.

    1. The downside of the 6 year term is having a crappy guy for 6 instead of 4. I would be more in favor of increase the amount of terms the governor can run. Why do we limit it? I can see some benefits, but I’m not sure the benefits are that substantial since it’s people that are voting.

      Abercrombie’s plan was solid. Having numerous systems can or maybe I should say, is a nightmare. Even within the DOE, the multiple systems can cause issues, then when you take it statewide it can get almost unbearable to get quality outputs like reporting, etc. The downside of the plan is whether the state could pull off such a huge undertaking. The meetings I had to attend for Abrecrombie’s plan had auditorium amounts of participants. How can you gather all that info from all those departments and develop a meaningful system? It very hard to imagine. It might have been harder than the rail, because the rail is starting from nothing. A new system has to try to incorporate all existing rules and policies and all the intricacies of the bargaining unit contracts as well as the manual workarounds that have been going on for 40 years (length of time the article said we were on the old system).

      Hiring better people to do the development or anything in that case has two huge issues. One, the state can never be competitive enough in terms of salary to match what the best companies out there are paying. So how well paid do the salaries have to be to get a “good enough” product. My take on UH sports is that UH should always just hire a young coach and take the chance he will be good and hopefully stay loyal to UH, because we cannot compete in terms of salary. The state should probably take a similar approach. But that leads me to the second issue in the state, which is firing crappy workers or even benefiting good workers. We talked about this ad nauseam. But if we going pay higher wages to attract “good workers”, what happens if they are not good. And now they don’t want to leave because they get paid so well. Or as you love to point out, what if the worker starts off good, but the state’s atmosphere makes them a crappy worker.

      Forgot to add that the DOE developed their HR system in house. Well we hired a contractor to develop it, but they pretty much sat in DOE offices and worked like they were DOE employees. If you ask HR, I think they are overall very happy with the product (This was done in the last 10-15 years.). So my guess is it’s possible. It was probably a super painful process though. Much more painful upfront in terms of development, but less so once the product is rollout because they got a lot of what they wanted. This wouldn’t necessarily be the case in an out-of-the box software.

  3. Don,

    I would be more in favor of increase the amount of terms the governor can run. Why do we limit it?

    I think I prefer this idea—increasing the limit to three terms. But why place a limit, any limit? My guess is that advocates of limits fear that incumbents have an unfair advantage and that can maintain power, whether they actually warrant this or not. I guess one’s position on limits comes down to the at least two factors: 1) How much faith do you have in voters being informed and engaged politically, and 2) the extent strength of one’s idealism. For example, a highly idealistic person may want to eliminate term limits regardless of the practical consequences of this.

    The downside of the (Abercrombie) plan is whether the state could pull of such a huge undertaking

    I think your comment can be said about a lot of ideas to improve government. I think we can conceive of many different ways of doing things better—sound ideas exists, at least on a conceptual level. But can those ideas be executed and implemented in a practical way? My sense is that adding this qualification would eliminate a significant majority of the best ideas. Do you guys agree with me on that?

    Hiring better people to do development or anything in that case has two huge issues.

    I agree that attracting talented workers, with government salaries (and benefits), will be really challenging. Having said that, I think the key question is, how much talent would we need to build a decent system? If you need the cream of the crop, then tackling this problem in house doesn’t seem viable. However, if most competent programmers would be able to build a solid system, then the idea becomes more appealing.
    To your second point about rewarding and punishing workers and ensuring that poor workers either improve or get removed, this problem obviously exists in most government agencies. Is the problem a deal-breaker in terms of trying to create this grand system with in house IT department? Off of the top of my head, I wouldn’t say that it is.
    Before going on I want to suggest something quite different. We’re talking about upgrading the software to improve data collection and sharing, particularly between different agencies under the assumption that better software will make government more effective. I think we should examine this assumption. In your department, do you think the software is the main obstacle to improved effectiveness and efficiency? Or do you think there are other ways that would lead to better customer service.

    Let me give one example in my department, relating to maintenance. Suppose a crack occurs in the basketball court and needs to be repaired. To repair this, a request has to be made and has to pass the desks of several supervisor, the last one giving approval (or not). Then the order is given to repair the crack, and this may have to pass through several people before the frontline guys get the order. My sense one wouldn’t need better software to dramatically improve the time it took to make the repair—although it could play a role. The bigger issue, in my view, is the level of motivation to make the repair in the fastest and most effective way—and all the people in the chain have to be motivated. One worker can hold the whole thing up. Even with terrific software, I suspect you’ll still need several people to approve the request. If, for whatever reason, they’re not motivated to get the job done as efficiently as possible that will likely create a lag. Sometimes there’s a bigger issue than the technology. When that’s the case, then the value of upgrading the technology/software may be far less than one thinks.

    Part of the reason I’m suggesting this approach to improving services stems from many of the problems involved with developing software—some of which we’ve discussed. Another challenge is the rate of change for software and technology in general. Constantly upgrading ones software and hardware, in a government setting, seems untenable. Because of that, other ways of improving efficiency and effectiveness seem more compelling. I also think there’s a danger of overestimating the value and importance of better software. In some departments (like yours?), the software and tech might be critical. In my department, I don’t think it is. There would be many other issues I’d tackle before upgrading software. In your department, I could see how you’d feel differently, though.

    Forgot to add that the DOE developed their HR system in house. Well, we hired a contractor to develop it…

    Wait, by “in house,” you mean not buying out of the box software. You don’t mean the state IT guys build the system, right?

    …but they pretty much sat in DOE offices and worked like they were DOE employees. If you ask HR, I think they are overall happy with the product.

    The first part seems disparaging—as if you’re saying the contractors didn’t work well…but now I’m thinking you mean they worked closely with those around them?

    In any event, one thing you didn’t comment on is the importance of the users of the software to have good communication with the developers of it. This seems essential. The users have to clearly and accurately articulate their needs and challenges to the developers, and hopefully the developers have the talent to create software that meets these needs. But the talent of the developers won’t matter if they’re not getting good information.

    1. Is the problem a deal-breaker in terms of trying to create this grand system with in house IT department? Off of the top of my head, I wouldn’t say that it is.

      It’s not necessarily a problem with deal breaker in terms of creating the system. But it will affect the production of these individuals, is what I’m getting at. So the same person works for the State of Hawaii would probably have no where near the production at a Google for example. So it almost like comparing apples to oranges, which means an increase in salaries won’t automatically equal the same production even if we could get employee’s with similar abilities.

    2. If, for whatever reason, they’re not motivated to get the job done as efficiently as possible that will likely create a lag.

      Yeah in the development, you try to parse out what the technology will improve and what will still be dependent on how well the human being does their work. So the easiest example is leave balances. So the new technology may calculate and even show on the employee’s pay statement what a person’s leave balances are, but if a human either at the secretary level of the employee themselves don’t input the leave into the technology the output will still be garbage. However, if a department is still calculating these balances by hand it will make that process more efficient and correct.

      In your department, do you think the software is the main obstacle to improved effectiveness and efficiency?

      No and it can actually make it worse in some instances. Right now the inputs of what is paid is my staff, so 18 or so people. One technology that was part of the option is timeclocks or some kind of time collection hardware/software. So employee’s would swipe in and out and be paid based on that. So on one hand you get more real data, you actually know what employee’s are working, yet you are now opening up the input of what will be paid to 20,000 people inputting things. So you cannot automatically state that the technology will improve efficiency. It could even mean hiring more than 18 people to sort through the data and clean up the discrepancies. However if handled correctly the payroll make become more accurate.

      now I’m thinking you mean they worked closely with those around them?

      Yes I think it definitely helped that the developers spent their day to day with the functional side and it had to help develop a much better product.

      But the talent of the developers won’t matter if they’re not getting good information.

      All out of the box software, you get a good part of what you stated above. This is because you present what you need the system needs to do, then the company will come back with this is how the system will do it. And this back and forth happens multiple times, yet it super hard to cover every scenario. Then workarounds are created to cover each scenario that the out of the box system couldn’t and so you end up with still a lot of inefficiencies. Not to say this cannot happen with home grown systems, because sometimes you just need a human to make a decision and so it’s hard to program everything.

  4. It’s not necessarily a problem with deal breaker in terms of creating the system. But it will affect the production of these individuals, is what I’m getting at. So the same person works for the State of Hawaii would probably have no where near the production at a Google for example.

    What you’re saying sounds right, but is the difference in production really significant? Without knowing more specifics, it’s hard to say. I mean, if the google guy is innovative and award-winning, and the gov’t guy is merely competent and effective–to the point where they build decent infrastructure–then I’d say the difference doesn’t matter. Build something that works and is usable and that would be a lot. Now, if it takes the gov’t guy way longer to do what the Google guy does–say, two years versus one, yeah, that would be a meaningful difference, although I think if the gov’t guy built something solid that would be decent.

    By the way, this is probably obvious, and I’d guess you’d agree with this, but there should be some evaluation and then a prioritizing of what tasks really need really good infrastructure. For example, in my department, we have permits for parks and registration for programs. Yes, it would be nice with a great infrastructure to do these things and serve the public, but I tend to think the amount gained with better infrastructure wouldn’t be that significant, and I imagine other departments have bigger needs–like payroll.

    So the new technology may calculate and even show on the employee’s pay statement what a person’s leave balances are, but if a human either at the secretary level of the employee themselves don’t input the leave into the technology the output will still be garbage.

    Do you mean that the current leave balance we see, either on pay stubs or even the government’s website, is not necessarily up to date? I just assumed it was.

    In any event, while this isn’t good, this seems like relatively minor. In my example with a maintenance request, one person in the chain might be on vacation for several weeks, and no one knows the request is stuck on his “desk.” Obviously that can really slow down the time repairs occur, and that can be a big deal.

    No and it can actually make it worse in some instances.

    Wow, I was a little surprised to hear this. Surely, though an update computer-based system would be way better than doing everything manually, right? For example, you mentioned improved accuracy (with the calculations, I assume). That’s not a small thing.

    So on one hand you get more real data, you actually know what employee’s are working, yet you are now opening up the input of what will be paid to 20,000 people inputting things. So you cannot automatically state that the technology will improve efficiency. It could even mean hiring more than 18 people to sort through the data and clean up the discrepancies.

    I’m skeptical about the idea of letting all the employees input their start and end times. This doesn’t apply to teachers, right? Or would it? How would you guys know if there were discrepancies?

    All out of the box software, you get a good part of what you stated above.

    Yeah, I basically agree (and pretty much everything you said after this).

  5. (Note: I’ve compiled some notes from the article, and I’ll be posting them gradually.)

    People are frustrated with government in Hawai‘i.
    There are many reasons, from the near daily revelations of graft, corruption and bribery among public officials, to rail’s ever-ballooning costs.
    “I strike up conversations with people all the time about what they think of state government. Usually their first reaction is they think of it as corrupt. And unresponsive. That it’s serving some groups and they’re not exactly sure who those groups are,” says Colin Moore, associate professor at UH Mānoa’s Department of Political Science.
    “Any sense that it is directly responding to their own concerns just isn’t there. People will just shrug their shoulders and assume certain interests, like development, will always get what they want and there really won’t be much action on anything else,” says Moore, who is also the director of the UH Public Policy Center.

    The complaint that government isn’t always effective and responsive to the needs of the people is valid, but I don’t think graft, corruption, and bribery are the primary reasons for this—not in terms of the way government agencies operate on a day-to-day basis. This is not to say that graft, corruption, and bribery aren’t problems or never hinder the effectiveness of day-to-day operations, but from my experience other factors seems like more glaring impediments. For example, the performance of an agency or an individual, whether good or bad, is largely treated the same—namely, with indifference. That is, the effectiveness of agency or individual in that agency does not matter much. Several years ago, Planet Money, an NPR economics podcast did a show on a Chinese Communist farm in the 70s. On the farm, regardless of the productivity of individual farmers, the government would collect what was produced and then distribute it equally among farmers. Essentially, the effort and effectiveness—the quality of their work—of each farmer just didn’t matter. One of the farmers succinctly captured the situation when he said, “Work hard, don’t work hard — everyone gets the same. So people don’t want to work.” To me, this describes my experience in government. Barring a major scandal, covered by the media, a government agency can function in a suboptimal way without any meaningful repercussions. Similarly, individual employees within the agency may perform at a suboptimal level without much repercussions. Additionally, if an employee improved his/her performance the consequences would likely be the same. In some instances, supervisors penalize these higher quality workers by relying on them more, giving them more work.

    In my view, the lack of meaningful and appropriate consequences and feedback from performance explains, in more compelling way, the reasons government isn’t so responsive, effective and efficient. From the point of view of the top administrators, why would they drive their departments to optimal performance, if optimal performance will be treated the same as sub-optimal performance? And if this is the attitude of many top administrators, then why would mid-level supervisors and front-line staff strive for optimal performance? In such an environment, an agency may function, but it won’t be functioning optimally. Improvement, acting proactively, strategic planning—there would be no compelling reason to do these things. Now, imagine what happens to these people over many years in this environment. Gradually, maybe imperceptibly, their morale and motivation will erode. Many of the people I worked with came into the agency with genuine enthusiasm. They wanted to do a good job. Over time, in the type of work environment I describe above, this enthusiasm wanes. Some become indifferent to their performance. This is not surprising. What’s actually surprising is the number of employees that continue to function and do their job. It’s a credit to their conscientiousness and professionalism to maintain this, in spite of the indifference to their performance.

    Do you guys agree that performance is often treated with indifference? If not, do you think corruption is a big reason government is ineffective? If you believe this, how would reducing corruption actually improvement government responsiveness and effectiveness?

Leave a Reply

Your email address will not be published.