Current Events:  DOGE — Government Efficiency — Do It Right, or Don’t Bother.

 

I see a lot of hand wringing on social media about Musk and DOGE (the “Department of Government Efficiency”).  Some of it borders on hysteria. 

 

A lot of that hand wringing is about cutting and/or eliminating programs and services.  That’s about the last thing I would do.  But I strongly support taking a big swing at improving government efficiency.  Why?  There’s inefficiency to be reduced, my taxes and yours are too high, and our national debt is too big.  Oh, and as a really experienced and smart 4 star general once said, our national debt is a serious risk to our national security.

 

Full disclosure.  1.)  I’m a political centrist.  Sometimes I vote for Democrats and sometimes I vote for Republicans, 2.)  I’m a bit of a Pollyanna, almost always assuming the good intentions and best of people.  On a few rare occasions, that’s been foolish.  Maybe this is one of those cases.  But almost all the time, it’s been a good assumption and a pleasant way to live life, and 3.) Finally, I am, in fact, thinking out loud here.  I reserve the right to make this thinking better.

 

Like a lot of things, improving government efficiency can be done well or poorly.  Effectively or ineffectively.  Efficiently or inefficiently.  Further down, I provide ideas on how to do it well, effectively and efficiently.  I hope someone is listening.  I hope it might increase your confidence in DOGE’s potential.  If you support the ideas, share them with your network and your elected representatives … tell them that’s how you want DOGE to work.

 

Let’s start with the word “efficiency.”  Simply put, efficiency is output divided by input.  Example from my boyhood:  I deliver 70 newspapers (output), and it takes me 120 minutes of my labor (input).  When I figure out a way to deliver the same 70 newspapers with the same quality in 100 minutes, I’ve improved my efficiency. 

 

Really important point.  To improve efficiency, you can and must measure output and input.  Skeptics will immediately challenge whether you can measure stuff.  They’re mostly wrong on this.  I bet I can come up with a metric that is common sense, practical, and useful for improving the efficiency for almost everything.  Maybe not everything, maybe not perfect, maybe not precise … but practical and useful and easily understood by the people doing the work, by taxpayers, by voters, and by the “customers” of our government.  (I’m glad to entertain challenge questions.  Leave them in Comments.)

 

I spent a lot of my career reducing costs and improving efficiency and productivity.  Without reducing output and without degrading quality.  In fact, we did more and did it better with less.  I know a thing or 2 about how to do it.  Here are a couple of ideas.  They are not theoretical dreaming.  They’ve worked in the real world, time after time.  They’ve worked in big organizations and at large scale, time after time.

 

  1. Just do it.  A lot of success in improving efficiency depends on believing you can and believing you must.  We can and we must. 
  1. Get the same or more output, with the same or better quality, but at a lower cost.  The emphasis here is on “same or more output”.  Not less.  No “customer” of the work loses a damned thing.  That’s improved efficiency.  How?  Often efficiency improvement is achieved by improving processes … reducing defects, re-work, and waste … and speeding up the process.  In my world, automating away labor with software systems was another powerful tool for improving efficiency and lowering costs.   But big caution here:  the federal government in too many cases is far, far from best-in-class at designing and implementing software projects. 
  1. Improve operations and programs that are performing poorly.  By “performing poorly” I mean:  1.) not meeting their performance objectives for delivering value, quality, and efficiency, and 2.) not improving their performance year-over-year.  Fix them … or replace them with something that does work well.  PS.  If you’re not measuring performance, shame on you.  That’s management malpractice.  Start.
  1. Squeeze suppliers.  I bet there are plenty cases where the government buys stuff at grossly inflated prices and astonishingly different prices.  Look at what you buy, look at the prices you’re paying.  Compare your prices to what those outside the government pay.  Compare prices paid across different government organizations.  Demand the best price, everywhere, now.
  1. In-source contractor labor.  There an enormous number of contractors working for the federal government.  Google tells me there are 3.7 million of them, about 40% of the entire federal workforce.  Contracted workers often cost far more than a government employee, even with the govie’s benefits, vacation time, etc.  That’s an unintended consequence of our push for a small government.  The contractors are laughing all the way to the bank.  The math.  Let’s guess a contractor costs $50,000 more than a govie (and that’s conservative).  Let’s replace 50% of the contractors with govies (and that’s conservative too).  That yields 3.7M x 50% x $50,000 = $92 Billion of annual savings.)
  1. Identify activities and programs that are delivering little value, or no value.  Make sure you really understand 2nd and 3rd order effects and unintended consequences of downsizing or stopping them.  Understand that these are a political 3rd rails; they are dangerous.  They will be a huge distraction and drain on those leading DOGE because these programs all have vocal, well-funded, passionate advocates:  lobbyists, special interest groups, and media.  They are the last thing I’d tackle.

Ideas 1 thru 5 require nitty gritty detail work by the operational types, the do-ers.  Good news:  Not glamorous.  Not newsworthy.  Not rocket science.  No silver bullets, big bangs, or moon shots required.  But effective.  Bad news:  They require widespread application and maybe a profound culture change.  I don’t know if they’ll deliver all the $1,000-2,000 Billion goal, but I bet they yield a pleasingly big number. 

 

Idea 6 is for the big thinkers and dreamers.  It will trigger a full-time career for policy wonks, think tanks, legislators, and the media.  Don’t get sucked into that morass.

 


Discover more from Reed Harrison's

Subscribe to get the latest posts sent to your email.

2 thoughts on “Current Events:  DOGE — Government Efficiency — Do It Right, or Don’t Bother.”

  1. Yes, there are huge opportunities to improve efficiency. You and I both have plenty of war stories to share about inefficiency in governments and large companies. That said, I’m not sure I would start by converting contractors to regular employees. In many enterprises, it is easier to hire and fire contractors than employees. It is a bit like the difference between capital and expense. There are good reasons to treat red $$s differently from green $$s — long term investments are different from short term investments — it may be over simplifying matters a bit to suggest we can cut $2T by accounting tricks.

    I learned a lot of valuable lessons about revenue assurance when we both worked at AT&T, but after moving to Microsoft, I learned a lesson that may be even more valuable. I was in a sales meeting, when a (non-US) military customer came to talk to Microsoft. They made it clear they didn’t want mil spec. They liked what Microsoft had to offer because it gave them better value for $$. I’m sure defense contractors charge more for less. But then the customer was asking Microsoft for a bit more “hardening”. Microsoft tried to say they offer value for $$, but you shouldn’t bet your country on their products. I think both sides left agreeing to disagree.

    The take-away lesson is: consumer markets are different from enterprise markets. Consumer products offer better value for $$; enterprises will do well if they can take advantage of those opportunities. Sometimes that is an option, but you might not want to bet the country on a toy. That said, it may be even riskier to bet the country on mil spec (because the toys work better and they cost less).

    Back in the 1980s, I was talking with a classmate (founder of Thinking Machines). He thought he was competing with super-computers (Cray), but he realized he was losing to PCs. He talked about “diseconomies of scale” — computing was cheaper and better in every way on a small computer (PC) than on a super computer (Cray). I mentioned this to some of the economists at Bell Labs and they told me that you Computer Scientists don’t understand economies of scale. They pointed out that scale refers to the size of the market, not the size of the machine.

    There was a time when the department of defense could tell Cray what features they wanted, and Cray would build those features. These days, the department of defense still thinks they can tell Silicon Valley what to build, but those days have long passed. These days, enterprises have to get in line behind consumer. The big bucks are in consumer. Enterprises need to figure out how to get dual use from stuff that was built for consumers — that’s ironic since dual use used to go the other way around.

    When I got started in computing, enterprise was the only business there was. I used to say, I wouldn’t use a computer I could afford. These days, a phone has better value for $$ than the computers I use at work. The challenge in enterprise is how to use off-the-shelf products, and resist the temptation to do what enterprises do. Even the military doesn’t need mil spec for most things they buy. Contracting with the government is more expensive than clicking on a EULA. Government web sites don’t work as well as websites for online banking and online shopping. Amtrak’s web site doesn’t work as well as similar web sites for plane tickets.

    Large enterprises and monopolies are inefficient. That’s what they do. There may be economies of scale/centralization, but those economies almost never compensate for the inefficiencies of monopolies. If we can figure out how to do more with off-the-shelf consumer parts and transparency/competition, we’ll do better than process-engineering/logistics like you would do in a military/enterprise. Rather than hire the beltway bandits as regular employees, I would try to figure out if there are ways to do much of what they do with alternatives from the consumer market. Bottom line: if the beltway bandits had to compete with toy stores, they would offer better value for $$.

  2. Contractor Labor. My experience at AT&T and Bell Labs in operations (software) systems. 10-15% of workforce was contractors. The idea was to use them to deal with peaks and valleys in the work volume and to temporarily fill in for a skill we needed and didn’t have &/or didn’t want long-term. Contractors cost a lot more than an AT&Ter but were “expendable” … they could be cut quickly if the work program or the budget required it. And that provided some additional job security and stability for AT&Ters. I didn’t invent this approach to using contractors but it always seemed sensible to me. The fed government is far from that model. Congress wants small government so it caps the number govies. But Congress is OK with using much more expensive contractors.

    Squeeze suppliers. So, I’m not talking about weapon systems or exotic technologies. (Like healthcare, that ought to be regulated and cost plus a reasonable profit). I’m talking about mundane stuff. Anecdote from the real world. Remote equipment sites needed to have the lawn mowed and the snow plowed. They were all roughly the same size. For some we paid X$ per year. For others, in the same geographical area and for no good reason, we paid 3X$. I advocate finding and fixing those kind of things.

Leave a Reply

Discover more from Reed Harrison's

Subscribe now to keep reading and get access to the full archive.

Continue reading