“Government.”
Sometimes a swear word, always a reality. In any stable society, there will be government. Government is our regulator, our rules committee. Government is our provider of basic infrastructure and crucial programs that advance the national interest. Government is sometimes a provider of subsidies. And government is always a collector of taxes. We don’t always like it, and we don’t always agree with it, but it is always with us. A society that lacks government falls into violent anarchy.
Given the inevitability of government, at some level of service, it is a truism that we will want from government much the same thing we want from the private sector. We will want courteous and effective service. And we will want affordable prices. Bureaucracy, we know, can become its own excuse for being, but we are also sophisticated enough to realize that large organizations need not be badly run. Like organizations of all kinds, government organizations have the potential to be run well. The better a government agency’s replication templates, the better its performance. Just as in the private sector.
We are Americans. We are optimistic pragmatists. When there’s a better way to do something, we are generally willing to listen. Is there a better technology? Let’s give it a try. Is there a better way to organize a project? Let’s consider it; why not find ways to work smarter and better? Society advances when imaginative people find smarter and better ways to do things.
And what is an organization, after all, but a team of people brought together around a mission and a way of accomplishing the mission? An organization is essentially a choreographed way of accomplishing a task. Is yesterday’s choreography sacred? Generally not, and certainly not in the case of government. The purpose of the law must be fulfilled, but the methods for achieving that purpose are usually flexible. Choreographers have hundreds of ways to design their dances; people who choreograph organizations have equally broad palates to work with.
So let’s strip away the mystique. People in organizations have similar strengths and weaknesses everywhere. Government habits are much like corporate habits, worth nurturing when they serve us well, worth changing when they do not. The smarter we are about this, the more value we will get for our dollar from government.
And being smart isn’t nearly as difficult as some people think it is.
Workflow Matters
Good workflow design is a pleasure for everyone. I have already mentioned the step function improvement in the Driver’s License renewal process at the Motor Vehicle Administration offices nearby. License renewals used to be a huge waste of time. Go there for a driver’s license renewal and you’d be a very long time getting out. Stand in line to get a number. Wait in a large waiting area till your number is called. See the clerk. Take her eye test. Verify your current address. Pay for the new license. Get a piece of paper. Wait in another line to have your picture taken. Stand before the camera. Wait in still another line to receive your license, once your photo had been taken. Took forever, and nobody was trying to be hard on you. That’s just the way it worked.
Now, with the help of technology, the work flow has been streamlined and the process takes only half the time it used to. The clerk does everything. She gives the eye test, takes your money, snaps your picture, and prints out your license. Boom. And you’re done. What a pleasure!
Multiply that savings by tens of thousands of people a year. A fast, well-designed courteous workflow is a huge favor to a busy public.
Bad workflow design, though, remains a curse. One of the Rotarians in our local club is German by birth, but married to an American woman. About a year and a half ago, he applied for American citizenship. The process started and stopped for him, over and over. It took him a year and a half to obtain the citizenship papers he wanted.
Someone in the State Department has a workflow problem. A very bad workflow problem!
And that is hardly the worst of the horror stories.
An old friend of ours, Debb Greenwood of Dickenson, North Dakota, owns a pharmacy. Not all that long ago, she decided it was time to offer diabetes strips, the chemically treated paper that helps a diabetic monitor blood sugar levels. It’s an inexpensive item, a low margin product. All she needed was a government document saying her pharmacy was authorized to carry them. Should have been a snap, right? Wrong.
Welcome to a bureaucratic nightmare that even Franz Kafka might not have imagined possible. It took Debb roughly two thousand hours of paperwork to get her pharmacy certified. Two thousand hours of government-mandated busy work. Someone in government has a workflow redesign problem.
You will have guessed my point already. People dealing with the American government suffer badly when the government is careless about its workflow design. It is time to become much smarter about the way we choreograph the work that government does.
Let’s Hear from the Users
I have just Googled two phrases, “How to register complaints about government inefficiency,” and “How to file a complaint about government inefficiency.” In both instances, Google’s returns the same response: “No results found.” I find that quite interesting. It clearly implies that nowhere in the English-speaking world[1] has any level of government seen fit to solicit citizen complaints about inefficient government workflow. Were the quest for feedback about inefficiency a standard practice, there would have been dozens or even hundreds of direct Google hits for my two searches.
Yet how is government to identify its inefficiencies properly without a well-publicized system for inviting feedback? Bad workflow may sometimes arise from an act of Congress, but it is never an act of God; it can always be corrected.
So the first thing to do is simple. Someone in Washington needs to be the chief workflow fixit person. Same for our state capitals and our big city governments. Reach out to the citizens. Send everyone the same message. “If you have run into ineffective or inefficient government practices, we want to hear from you.” Those who have been put through the wringer on citizenship should be offered a process for documenting their experiences and their complaints. Same for those who have been tied down responding to irrational information requests, especially people in business. It is the government’s job to create effective and efficient workflow procedures so that citizens and businesses can interact with government easily and effectively.
Document the Existing Workflows
There was once a woman who made wonderful pies. Everyone loved her pies, and she enjoyed making them. As her reputation spread, even strangers asked her about her pied. One day she decided to open a pie shop. It was an immediate hit. She made all her favorite pies, and customers flocked to her store.
She became busier and busier. The more pies she made, the more customers she had. After a time, she realized she wasn’t having as much fun any more. She asked one of her friends what to do. The friend gave her a simple piece of advice. “You spend all your time working in the pie shop. But you don’t spend any time working on your pie shop. If you want to get your life back, you also have to spend time working on the pie shop. You have to redesign the way the pie shop makes pies, so that you can have more time to yourself.”
And so it is in government. It isn’t enough to have tens of thousands of Americans working in government. At some point in time, everyone who works in government also needs to spend a few hours working on government.
Working on government means getting the workflow right, just as it would in a pie shop. So, for starters, let’s focus on the workflow issue. How can we know that every government workflow is well designed? The only way to know that is to have a government that regularly documents and evaluates its key workflows with some regularity.
But this isn’t just a job for experts. It’s really everyone’s job. For starters, let’s declare that one day a year will be Workflow Documentation Day. For every government employee – national, state, local, you name it. Each agency gets to pick the day it wants. If I am an agency employee, I list all the workflows I participate in. I pick the one that takes up the largest amount of my time. I document each step in that workflow, and do so in a way that a casual observer would be able to study my documentation and understand what it is I do, in what order, and what outcomes the workflow is expected to produce.
Many workflows have more than one participant. The documentation work isn’t done, of course, until every phase of a major workflow has been diagramed and described. Suppliers of information and receivers of information participate in the same workflow. One might have to stitch together workflow descriptions from half a dozen people, or many more, to have an end to end description of a major workflow.
A diagramed and documented workflow belongs on the wall, somewhere, so that colleagues can see it and study it. Within the limits of security, anyone who interacts with a workflow should be invited to review its steps and offer comments. Informational comments. Comments that raise concerns. Comments that offer ideas for improvement.
There are many ways to document a workflow. When I was at Cummins, our fledgling Just In Time Team videotaped operations on the assembly line. At each work station, we’d videotape the assembly operations for three or four engines in a row. Once we’d finished, Bob Reed, our manufacturing engineer, would list each of the work steps on a flip chart. Typically he’d identify ten or twelve separate steps. He timed each step, took the average from the three or four iterations we’d videotaped, and recorded the time on the flip chart.
This list of activities became a jumping off point for brainstorming, with the team members, about how the work station might be improved, but I am getting ahead of my story.
Later, at Bell Atlantic, as part of a Gemini Consulting team, I participated in the documentation of the provisioning workflow, the process that begins with a service order from a new customer and ends with dial tone being provided to the customer premises. There were only five major steps to this process, as I recall, but we must have identified thirty distinct steps in the overall workflow.
Gemini had a taste for the dramatic. A Gemini workflow diagram was known as a Brown Paper. It was created on a strip of brown butcher paper, four or five feet wide and as much as thirty feet long. Step by step, we’d build box-and-wire diagrams to represent the entire workflow. By the time we were done, the Brown Paper was filled end to end with diagrams and notes. Employees were invited to Brown Paper Fairs, to observe the diagrams, and annotate them with their own comments about strengths, weaknesses, and concerns.
Gemini also did DILO’s, “Day In the Life Of.” In a DILO, a consultant shadows a client employee for a full day, documenting every piece of activity that made up the person’s day. This, too, captures a perspective on company workflows, and teases out the percentage of time individuals are likely to spend on activities that add genuine value.
About the time our joint Bell Atlantic-Gemini team completed this work, a book was published on the secret life of dogs. The author had followed dogs and recorded minute by minute what they do, and then psychologically reconstructed all the emotions the dogs must have felt. We joked about DILO’s for Dogs.
In one important sense, our Brown Papers were never complete. Whenever a piece of information had to be received from the Information Technology folks, or provided to one of their databases, we drew a little cloud. We’d connect it to the main workflow with a dotted line, but we took a pass on tracking down what happened in the guts of the computer software. The mission of our project was to identify waste that could be removed from the physical workflow. We were explicitly instructed to avoid improvement opportunities likely to require significant redesign of the IT support system. IT redesign was said to cost tens of millions of dollars and eat up two or three years as though it were nothing. We would have to live with whatever shortcomings the IT system imposed on the physical system.
As one might imagine, there are many ways to carry out a series of workflow documentation mini-projects. Workflow Documentation Day might produce its own videos, or its own Workflow Fairs, or anything else that gets the point across and properly documents the real life workflow.
What matters is capturing both the main route and the likely detours. A good workflow diagram has to explain why an applicant for citizenship might get sent back to the starting line over and over again, or why a pharmacist might have to invest 2000 hours in documentation work.
A workflow diagram also has to identify the role of legislation in shaping agency workflows, for better and for worse. “I do X because Law 123 tells me I must.” Every step in a workflow that is thought to be mandated by law should be diagrammed and flagged. In a properly documented workflow, the impact of law and regulation should be fully visible. If better law can support a more rational workflow, notations to that effect belong on the workflow.
Will a single Workflow Documentation Day be sufficient to capture all of the major government workflows, especially those that affect customers and suppliers and those whom the government regulates? The answer has to be No. But it gets the process started. It will pay to make a big deal out of the project – a very big deal. if the old and slow choreographies of yesterday are to be melted away, and replaced with faster and friendlier choreographies, the documentation process needs to be an All Hands project.
Redesign Flawed Workflows
Modest amounts of workflow redesign are under way within the Obama Administration. Obama’s point person on this effort is Chief Performance Officer Jeffrey Zients, from the Office of Management and Budget. At one agency, Zients discovered a 210 day hiring process. “You need to shorten that,” he told them. On his next visit, agency personnel proudly bragged about how they had shortened the process, from 210 days to only one hundred eighty! Clearly they didn’t get the picture. Zients took them in hand. By the time they were really done, they’d scrubbed the hiring process down to eighteen or nineteen days.
Such a triumph is a good start, emblematic of what is needed everywhere, but clearly just a beginning.
What we need, once we have pulled off our first Workflow Documentation Day, is a companion day, a Workflow Redesign Day. As one recent saying holds, “With enough eyes, all problems are shallow.” Pick out a Workflow that’s in desperate need of redesign. Put a lot of eyes on it. And ask the right questions:
Goal? What is the final goal that the workflow was created to achieve? Does the goal make sense as defined, or it there a better and smarter way to define the goal?
What actually adds value? Which parts of the workflow add genuine value? And which points in the workflow simply mark time? Or waste time? Or generate unnecessary work?
Adding value is always a vital standard. In school, when the children in the classroom learn the weekly lesson, the teacher has added value. In an urban transportation network, cars add value for commuters when they’re moving. They don’t add value when they are gridlocked and trapped. In an energy infrastructure, a generating plant adds value if it can create electricity without undermining the global climate. A generating plant that damages the climate takes away more value than it adds.
Sitting and waiting my turn at the Motor Vehicle Administration doesn’t add value. Taking the eye test adds value. Confirming my address adds value. Having my photo taken adds value. Printing the drivers license adds value.
Value can take different forms in the world of government. A cop on the beat adds value by deterring crime. Repairing the Statue of Liberty adds value by preserving a precious asset. A mine inspector adds value by denying a mining permit to an irresponsible and unsafe mine operator. Whatever form value takes, that’s what we want government to focus on. Everything else is waste.
If we are to eliminate all the waste, how do we have to modify the workflow? That, of course, is the next question. Once everyone understands what it means to add value, then the next issue is figuring out how to avoid waste. An immigrant married to an American citizen wants to obtain his citizenship papers. The law no doubt requires a series of tests to confirm eligibility. How long should those tests take? It is difficult to imagine more than a few hours of genuine value-added work taking place inside government in connection with a citizenship application. It is difficult to imagine a legitimate reason for waiting periods of more than a month, especially for an applicant who has lived in the United States for years already.
Workflow redesign begins with the distinction between value-added activity and activity that involves pure waste. It eliminates the waste, preserves the activities that add genuine value, and rewrites the workflow rules accordingly.
Perhaps a wasteful workflow exists in its current form as a result of legislation. It is an agency’s responsibility to let Congress know about the waste that legislators have mandated, and to give them options for streamlining both the law and the agency workflow.
Two simple elements are always worth adding to the mix. Bring as many stakeholders into the process as possible. Ask them all for their ideas: “What’s the simplest way to get the real job done?” And, second, make sure that Congress (or the State Legislature, or the City Council) understands where its laws and regulations are overloaded with detail. If there is an efficient way to accomplish an aim, but an outdated law stands in the way, Workflow Redesign will require help from Capitol Hill or its local counterparts.
Public servants can spend their lives working “in” government without ever being invited to work “on” government as well. Workflow Documentation Day both symbolizes and launches a new approach to public service. Workflow Redesign Day begins the rewrite process. Public servants get to take part in making each workflow as smart and efficient as possible. It should be something we always do – consciously set aside time to work “on” government.
Create Fast IT Turnaround
Hardly anything is more satisfying than having a good idea and then seeing your idea put into practice. Not long after we had begun videotaping workstations on the assembly line at the Jamestown Engine Plant, Bob Reed hired two ironworkers, Tony White and Tom Swanson. When they arrived, they were on a one-month contract. They were so valuable they stayed four years. Team members on the assembly line knew them as “Whitey and Tom, the welders.”
I have already mentioned our videotaping and brainstorming work. Team members always had good ideas for how their work stations could be improved. Sometimes there were new flow racks to be built, or special fixtures to be created. One of the upstream-downstream signaling innovations involved colored ping pong balls. The color coding signaled the turbo model required. When the engine arrived at the turbo station, the team member was ready with the right option.
Whitey and Tom almost always knew how to turn a good idea into a workable gadget. A steady flow of good ideas began to flow across the aisle from the assembly line to the welders’ work area. A week after being given a clever idea, or two or three, the welders had just the right gadget ready to go. They created new flow racks to deliver heavy wire crates of parts to the assembly line more smoothly, paired with flow racks that returned the empty wire crates to the aisle for the fork lift driver to remove. Whitey and Tom produced one valuable gimmick after another, always in response to ideas the team members themselves had originated.
Think of it. Hourly workers on a diesel engine assembly line, capable and hard-working people, creating their own ideas for smarter and better tools, and getting amazing turnaround. How many workplaces in America can say the same?
Contrast this experience with IT turnaround in an office environment. It is a tortuous process. At a spring 2010 banquet-style event in Washington, DC, I happened to be seated next to Jim Duffey, Virginia’s new Secretary of Technology. He shared with me the highlights of his new job. Several years earlier, then Governor Mark Warner had hired a major contractor to modernize the state’s information technology infrasructure. It was a ten year, hundred million dollar contract. Secretary Duffey’s assignment? Resolve the open issues, get the project back on schedule, see that it finishes on time. “How many state agencies?” I asked. “Ninety-four.” Secretary Duffey has his hands full.
Note the contrast. Three week turnaround from the welders at the Jamestown Engine Plant. Ten year turnaround for a public sector IT overhaul. Not that it’s much better in the private sector – IT rework is a multi-year nightmare for large corporations just as it is for government agencies.
Is this really necessary? It might not be.
Not all that long ago, I got a glimpse of an entirely new way to think about IT redesign. My employer at the time was Education Resource Strategies (ERS), a nationally prominent consultancy that serves leaders of public school systems. As its specialty, ERS it analyzes a school district’s key databases – budget databases, teacher seniority databases, class enrollment databases – and educates district leaders about the instructional implications of their traditional spending patterns. This always leads to a series of “What If” discussions. Is it better to allocate scarce resources to class size reduction? To the hiring of pull-out teachers? To the hiring of mentors for new teachers? Or any of a number of other options.
Over the years, ERS founder Dr. Karen Hawley Miles had come to appreciate how difficult it is for school system leaders to develop an intuitive feel for the likely budget impact of the resource decisions that mattered to them. She wanted to create an internet-based tool by which they could test different decisions and learn almost instantly what the likely cost implications would be. She lined up a funding package that involved the Bill and Melinda Gates Foundation, and the University of Washington, and I was put to work building the initial model in Excel.
Fast forward to the point where the model is ready to be translated into internet-friendly software code. Karen Baroody, ERS’ Managing Director, engaged a pair of software wizards to port my Excel model to the internet. Think Whitey and Tom II. Guided by the Excel model, the software team built the appropriate interface. User Inputs here, Budget Consequences there.
So far, so good. Now for the wizardry. One of their tools was an Excel-to-Java translator. This was extraordinary. They fed it my Excel file, and almost instantly they had the Java code they needed for running the model on the ERS Tools website.
As one might expect, it took any number of tweaks to refine my initial model. Each new tweak generated a new Excel file, and a corresponding need for new Java code. No problem. The tweaks were converted into Java code just as fast as the original file had been. I was ordered to observe one rule. Keep the user input cells exactly where they were before. Don’t fiddle with their row-and-column locations. Anything else I was free to change. New formulas. New assumptions. No problem.
Baroody’s two software wizards had given ERS the same swift turnaround that Bob Reed’s welders had given the engine assemblers at Jamestown.
Just as a single nugget of gold may signal the presence of a nearby vein, this nugget of experience suggests a rich vein of opportunity. Maybe IT modernization can be done differently, and much more swiftly.
So why not swift IT turnaround in government, in as many areas as possible? Here’s the procedure I propose. Evaluate each IT module that supports the work of a government agency. Ask a simple question. Could its logic structure be reproduced in Excel?
Some modules are no doubt beyond the reach of Excel. I doubt that Excel has the ability to reproduce the government’s weather forecasting software.
But who knows? Excel is a powerful, high capacity tool. I suspect that the logic framework of most IT packages could be reproduced in Excel or something comparable. Let’s take as a working hypothesis the idea that Excel is capable of handling the business logic for most of the IT packages government uses.
Now, let’s imagine a redesigned relationship between an operating government agency and the government IT department that supports it. The agency has a new role, the IT department has a new role. The agency is responsible for modeling its own business practices in Excel, or equivalent. The IT department translates the business model from Excel into software that’s fast and secure.
Now let’s postulate a successful Workflow Redesign exercise. Agency X has found a faster and better way to do business, but it cannot implement its new system without appropriate software changes. Step One: It revises its Excel model to reflect its new way of doing business. It runs appropriate quality tests to make sure its new model performs as it should. Step Two: The IT department takes the new Excel file and converts it, almost instantly, into new code. Same capacity for high transaction volumes. Same database connections. Same security measures.
It probably won’t be an overnight process. On the other hand, it shouldn’t be a two-year process. Let’s press for full turnaround within a month. Preferably faster.
Note what this rearrangement does for us. In the old days (i.e. now), one imagines a workflow change, writes a word description of the change, and then asks IT to translate the word description into operating code. The IT department has to figure out precisely what the operating unit wants, and then create custom-designed code that expresses the business model to the best of its ability. But it might be wrong. So the two groups go back and forth, back and forth, till the business model has been defined sufficiently well that further coding can move forward. It is a long process.
The redesigned approach changes the dynamics. Within the operating agency, one imagines a workflow change. One codifies the workflow change. One revises the Excel model to bring it into line with the new procedures. One submits the new Excel model to rigorous quality testing. And then one hands the new Excel version to the IT department to be translated into the proper code and put into service.
The turnaround queue shrinks from years or months down to weeks, days, or even hours.
This won’t be a complete solution to the problem of IT delay. Software that cannot be represented effectively in Excel will remain out of reach.
That being said, the times favor the change I propose. The world of IT isn’t so hermetic any more. Open source programmers delight in creating standard modules capable of solving routine functions. Invite the open source world to help government take the wait out of IT turnaround. Throw them the Excel-to-Code scenario I have painted and ask their help.
Excel literacy continues to grow. There might have been a time when it wouldn’t be easy to find someone to maintain a formal Excel version of the software service a government agency requires. That won’t be nearly so difficult today.
I wish Secretary Duffey well in Virginia. But I also wish for the day that government agencies and private sector corporations will no longer be tied to multi-year IT projects anytime they want to reinvent the way they do business. Fast turnaround shouldn’t be the exclusive prerogative of engine assemblers with talented welders at their side. Fast turnaround should be a way of life for people who work in offices too, and for the executives who support them.
Customers, Suppliers, and Stakeholders Deserve No Less
Every President takes office with the promise that Government will work better. Every Governor does the same.
These promises often wither once the elected executive is overtaken by the pressure of events. But they don’t have to. Workflows can be documented. More to the point, every important workflow should be documented. Workflows can and should be redesigned, and, where necessary, laws should be revised to support greater effectiveness and greater efficiency.
And IT support can be redone. Today’s IT workflow might have made sense in an earlier time. It no longer fits. Much faster IT turnaround ought to be a high priority.
Citizens and businesses who interact with government should never feel that governmental procedures are endless and hopeless, as ancient as the manual typewriter if not more so. Governmental organizations can adapt to the times, just as other organizations do. If we wish them to.
Steven Howard Johnson, Version 2010-07-27
[1] The United Kingdom, the United States, Canada, Australia, New Zealand, and many others.