Yes some of the big boys are wasteful and overpriced, but I can ensure you that the federal government itself would be 5x worse, especially since, in DC, at least, the typical GS level employee has absolutely no experience in software engineering, and many times less authority to make good decisions.
Anyway, I've seen a few of these types of enterprises sprout up in the last 5 or 6 years in DOD land. They are sometimes referred to as 'boutiques' in contrast to the big players like BAE, Northrop, Lockheed, etc. They tend to have much younger engineers, SV type office environments, and a lot of hubris.
I'm currently making a good salary via cleaning up their MongoDB, Node, and bloated AngularJS messes that never performed half as well as the old systems running Java and a real RDBMS.
Government uses contractors for lots of reasons -- many of which are more about shifting risk/accountability.
Operating in a unionized environment is challenging because of the inability to fire, but that's more a leadership issue than a cost issue. For most of the infrastructure services we built, during the growth phase, we'd be 25% FTE cost (including benefits), 25% contractor cost, and 50% hardware and software, with the mix depending on the service. Software projects had more contractor cost, and some specific, high skill infrastructure areas had more contractor cost.
The key to a sustainable service is distinguishing build/grow from operate. When the service is in a operate/steady state phase, the government employee staff should be running it 100%, and bringing in SME contractors as needed. That's how you operate effectively.
As you do this stuff, you build a cadre of smart employees who develop the ability to build more and be flexible. You never run out of projects. By owning the services and processes, you save money unless you're totally inept.
All three of them are using Cloud Foundry as their platform. By way of disclaimer, I work for Pivotal, which is the leading donor of engineering to Cloud Foundry.
Do you have some more information on this? To my uneducated eye, it looks like GS14 (with benefits) is substantially less than a good software dev would make in the private sector. I would be very interested in a government job if I am incorrect.
Granted the salary at the GS-14 and GS-15 levels are on the lower end than compared to Silicon Valley, but it's actually around what an engineer with similar experience can get in the DC/NoVA area, in my experience.
Not to mention that it's basically impossible to get fired as a GS unless you royally screw up. Usually this means committing a felony outside of work, or committing serial sexual harassment in the workplace, etc.
The article says that 18F is alleged to be:
> acting as both a procurement policymaker and as a tech competitor
This seems like a severe conflict of interest in which they would write procurement contracts favoring themselves. I would be okay with it if they just made 18F the site builder and didn't have them compete with private companies on tender, but if the allegations are to be believed, that doesn't seem to be the case, it seems like 18F is a federal agency bidding against private contracts.
*I work for 18F.
Contractors hate standards, because they scope what they can sell. If you make a simple statement like "use Git for source control", the vendors want to be able to contest that decision, or have the decision made in some sort of "open" form where they get to lobby for some other standard.
If the "conflict of interest" biases taxpayers - awesome. Government procurement is supposed to be 100% bias towards helping taxpayers.
You can definitely rig the corporate IT game.
When people talk about access to high-paying jobs and powerful positions being biased towards upper-middle class whites and Asians, this is usually what they mean (though not for the reasons you may presume). For most choices, there's a threshold above which the remaining choices are effectively equal. If you have a myriad of options that are all competent enough to do the job and in the same price range, how do you go about choosing one? At a certain point, the act making a choice and moving forward is more import than which one you chose.
There may be 10,000 shops qualified to create whatever it is you need. If you happen to know and trust one of them, and the price range is within budget, why bother investigating the other options? The time investment required would be huge, but the payoff wouldn't be very high.
Of course, overstating your needs and/or your budget to benefit someone you know is wrong - it's corruption. But, there's a very similar outcome that's a natural extension of human behavior. It's not detrimental to the parties involved, but may be to the diversity of society as a whole. All else being equal, of course you're going to choose someone you trust, and your friends are probably some of the people you trust most. Since you typically meet your friends while living your everyday life, they're probably people you have a lot in common with - people who had access to the same schools, jobs, and neighborhoods you did.
At first blush, if you need something done and there's a qualified friend who can help you, there's nothing wrong with that. However, after enough iterations of powerful people hiring their friends, we may end up with something resembling an aristocracy - a bunch of powerful friends always hiring each other, and another bunch of non-powerful people who can't get their feet-in-the-door.
Has there ever bean a great answer to this? Is it a conflict of interest or an area of domain expertise.
The answers are that the SEC and the Fed are best for making banking regulations. The EPA is best for regulating the oil industry. And 18F is best for regulating the tech procurement process.
How can it be possible for an entity to have a conflict of interest in deciding to do something itself?
I think once they clarify this, it would be all good.
Consider this scenario: The 18F team works with the Secretary of the Interior's office to draft requirements for a system needed by the Department of the Interior. They gather the requirements and conclude "This isn't a huge system. We can do this ourselves." Do the individuals working at 18F get raises or massive bonuses for doing the work, or do they continue to receive the same civil service graded salary they've always been receiving? It is certainly in the 18F department's interest to have work and continue to exist, but I don't see any huge payoff to the head of 18F the way there would be if a large company got a contract.
Or, to sort of repeat a scenario I offered earlier in this conversation, my wife wants our bathroom remodeled. I interview her and find out what she wants and write it all down. I go over it with a few companies that do bathroom remodeling. I receive their bids. I conclude that I'm capable of doing the work myself more quickly and more cheaply. I award the "contract" to myself. Where's the conflict of interest?
The point of RFPs are to allow fair competition between vendors. Also to allow the government agencies to compare bids apples to apples and score accordingly. You're example is fine as long as you are providing the other bidders with the same information. It breaks down if you have inside information that other bidders don't have.
For example, most RFPs don't include budget info, but often a large portion of the scoring is based around the budget. If 18F had worked to create the criteria, knowing information that wasn't public to bidders, then won said bid, that's where the conflict comes in.
But this is all for not, as 18F has clarified that they don't do this, which is great. I'm all for better requirements, transparency, and lower government spending.
I bid on a lot of RFPs and the process needs fixing for sure. Anything tobreduce the amount of work I need to do to draft a bid is welcomed. Goverment bids take an insane amount of man hours on our part.
Love the work 18F is doing in this space.
*i work at 18F.
Or in other words, I agree with you.
/P.S my depressingly-joke-like-consultants may or may not have just moved on to KPMG... am I hitting too close to home and do i know you? :P
One example out of many - small(ish) bank using their core banking platform (package) - to keep up with regulators, changing markets etc. they would have to create super-complex system with big internal team, evolve it, secure it, fix it etc. One serious bug can bring you literally down.
Or you can buy a packaged software, customize it yourself and you have big vendor's guarantee that SWIFT messages will be delivered and won't get screwed up.
Usually only few, massive banks do these themselves. If they do, they need to have big IT department. I am talking about companies with 5-50 IT people in total, managing everything.
Having worked in big companies where consulting firms like IBM and Deloitte have really screwed things up one thing has become abundantly clear to me.
Building yourself is always better. There are really no caveats to this. A team hired by you is a team aligned with you. A firm servicing a consulting contract will always milk the contract for whatever they can get, especially when they often do really sly things like settle for a lower initial cost but build in onerous clauses that drastically blow out costs for all sorts of reasons.
If anything they should expand 18F/USDS and eliminate the buy-first policy in favor of build-first. Only go to vendors when it's just not feasible to build it with the resources available.