Hacker News Reader

Comment by jordigg | Hacker News Reader

Really happy seeing these initiatives and the way the White House is thinking about bringing technology to the government.

I just can't understand why other government bodies and countries still give all their work to big corporations asking for ridiculous amounts of money for delivering questionable work quality. Their only thinking is how we can deliver the worst software ever that require us to maintain it for as many years as possible.

Give the work to smart folks who are willing to make it happen because they believe in that country and how they can make a true impact and you'll get wonderful software at a reasonable price that will just work.

View on HN

You bring up a great point and a big challenge is the way government purchases things. There have been a lot of laws passed to ensure that the gov't buys things in a fair way (as not to just give the contract to a friend) but over time those laws and policies have resulted in an extremely complex and time consuming process that only huge companies can devote the resources to "apply" for a job and win it. Once a law is passed and a policy is in place, it can be very hard to unwind (power of inertia) which is why it can take a crisis like Healthcare.gov to make sweeping changes. Teams like 18f are designed to help the gov't build the right tech in house, skipping the procurement process, and using agile methods and modern tech to build faster and cheaper.

To be clear, 18F is not be "skipping the procurement process" and has released their own Agile BPA which is great. Skipping the procurement process would be the sign of a bad workaround.

A critical part to fixing "the problem" is both creating vehicles for new kinds of companies to access government and reduce the overall risk of government officials purchasing from those companies.

"Nobody ever got fired for buying IBM / Accenture / Whatever" applies in government as well, but the appetite for risk can be even lower than in the private sector, with higher costs of entry.

Top-level butt-coverage from the White House / OSTP and another agency demonstrating that it is possible and creating the contract vehicles to get more agile / lean / innovative businesses in the door goes a long way to solving this problem.

It isn't just taking on risk by trying something new. Even a known better solution can result in assuming risk.

Take a long existing systematic problem which produces 100 bad effects a day (people getting double billed, database errors, what have you). Someone thinks of an improvement to reduce this to 10 bad effects a day instead. A major improvement. But they don't do it.

Why? Because anyone who approves of the new system is now the person responsible for the 10 errors a day. It goes from 100 errors per day because that is just how the system works to 10 errors a day because of Bob two cubes over. Bob ends up being blamed for the 10 errors far more than being congratulated for the 90 that no longer occur.

Additionally, there are people that believe that the private sector spurred by competition can provide a better, cheaper product. Personally, I think that's more a feature of new and small companies without entrenched interests and inertia. As you noted the requirements for government contracts have made their entry into bids harder than larger companies that can afford the large upfront bidding cost in time, effort, lobbying, etc.

In my eyes, it's entirely possible the USDS is and ca be successful because it's new. I think it's (unfortunately) entirely possible that in 10-20 years they will suffer from the same bloat and inertia problems large companies and existing government agencies do. I'm not sure how to combat that, but having a way to revitalize large companies or government agencies without losing too much of the existing internal knowledge and expertise would go a long way towards helping combat this, but I'm not sure what that is. There are probably numerous successful companies that have done this that I'm entirely unaware of, so maybe there is a good amount of information on how to combat this...

GDS are doing a great job on this in the UK by building a platform for small companies to pre-register as government IT suppliers, covering everything from phone systems to cloud hosting and bespoke software. One of the things I really like about the work they're doing is that one of their performance metrics is increasing the number of small businesses getting government contracts.

There's almost no pipeline for "smart software folks with industry experience" into DC right now. "Big corporations asking for ridiculous amounts of money" as you put it, is realistically the only option available to the decision-makers.

USDS and 18f are trying to change that reality in part by creating that pipeline. It only works if good people are willing to put in the work.

Often the thought isn't "let's give this to a big corporation." Often, the way the proposals and contracts are written, only big companies could reasonably respond.

18F is helping to change that with our Agile BPA!

Because those giant companies have an army of managers and accountants that will be sure to check every box and cross every t. A group of Valley engineers is thoroughly less thorough.

Reply on news.ycombinator.com