Hacker News Reader

Comment by princekolt | Hacker News Reader

> Having full-time software experts running the show turned out to be crucial. Previous incarnations of the project didn’t have a technical leader at the MTA—just old-school senior managers who would try to wrangle the contractors by force of will. The new in-house team, by contrast, was qualified to define exactly what they wanted from software providers in terms those providers could understand. They were qualified to evaluate progress. They could sniff out problems early.

I think this is the most important topic in this article. Government projects cost so much and take so long because there is always a disconnect between the institution paying for the work and the institution doing the work. One can’t reliably measure the other’s capabilities and knowledge and so it becomes a murky relationship (one that is also easily exploited by persons on both sides willing to make their paychecks fatter).

View on HN

As someone who works in the power utility industry - an industry that has slowly moved from monolithic, often publicly-owned monopolies to a competitive market where lots of functions are contracted out to service providers - I feel that a lot of this is symptomatic of that process.

In the 'old days' when we built lots of large infrastructure projects like the New York subway, various energy megaprojects, the interstate highway system, etc. nearly all technical functions were done in-house at the government or quasi-government agency responsible.

Now we have become 'leaner' and pared down many government agencies to project management only (sometimes just business management - projects is contracted out too), but it's not that easy to manage construction of a subway or a 500 kV transmission line if you don't actually have the base of relevant knowledge internally. Suddenly you have 25-year-olds fresh of out of engineering school being christened 'Senior Project Manager' and put in charge of a large, very complex technical project when 40 years ago, they would have been doing basic design and learning their trade as an engineer under the direction of a phalanx of senior engineers. Then we pat ourselves on the back for 'saving' loads of money by reducing headcount in the public sector when actual total cost of ownership has escalated.

The people actually doing the nuts-and-bolts engineering and construction work have no ownership in the project because the second that contractual completion is declared it's off to the next one.

Right now I am working for a power company that refuses to hire its own project or commissioning engineers, instead contracting management, oversight and quality control of its construction projects to a litany of 3rd parties. Unsurprisingly, it means very few if any people in our company have the expertise to understand what's going on when a project is off the rails. The result is endless delays while senior management wrings their hands and brings in yet more consultants to manage the other consultants.

how is this affecting the quality of the infrastructure that gets built? the TCO for these projects is going up, but is the quality (read: reliability/time between failures) comparable to what you could expect from the older systems?

It's hard to make a statement about average quality without a lot of data, but the total variance in quality is certainly increasing. Owners/companies that have strong engineering processes and standards are still building high quality projects and with the benefit of improved technology, it's much more reliable than circa 1970s stuff.

But there are also a lot of failures caused entirely by project management SNAFUs, poor communication between contractors on a project, inadequate testing, etc. that are completely avoidable and I personally as being largely caused by the diffusion of responsibility that occurs when a company is no longer capable of turning a wrench without hiring a contractor.

"Now we have become 'leaner' and pared down"

This happens in the Private Sector as well. I work for a manufacturing company with 6+ facilities (two major, third major greenfield in design and 3 minor facilities) We have 3 engineers for the whole company. We mostly just manage the contracted engineering firms. It sucks for me as a "junior" Senior Project Engineer because I rarely get a chance to learn what my contractors are doing, too busy managing them and coordinating. Naturally, in order to be a better PM I need to know their skills at least at a 500' level. Hard to learn to code PLCs when I'm too busy managing 5 PLC programmers on a project for a month or two, then moving onto something else that might not even have a PLC involved.

Early on in my career, a manager of mine told me about his first few years on the job after being hired by Ontario Hydro (at that time the integrated generation and transmission utility for all of Ontario) in the late 1960s. The first two years was pure training and basically an extension of your engineering degree except specifically applied to the typical problems and practices of the power utility.

Of course, in those days a utility job came with a generous pension, a seniority-based career track and an expectation that you would likely retire there, so anything invested in a young employee was just investing in the company itself.

Large engineering employers no longer feel the same social compact with the profession to train young engineers and make experts out of them. Not only do they not set aside time for employees to learn, there often aren't any seniors around for them to learn from because the company hasn't bothered to create a meaningful career track for engineers. It's promotion up to management or stagnation at entry-level pay. With little meaningful prospect of advancement in their technical career and no pension 'golden handcuffs', there's no incentive for employees to stick around. Once they become experienced, they split for the contractor they used to manage, creating a vicious cycle that drains the utility company of talent.

> anything invested in a young employee was just investing in the company itself.

We often discuss the modern lack of perks you listed, but we rarely discuss this consequence.

I recently left a tech company and was surprised when engineering management showed no interest in learning why I had left. It makes sense not to worry about -me-, but why not worry about the workers who haven't left?

I concluded that since places assume you will leave within 3 years (I was at 3.5), they dont really concern themselves with trying to counter that and consider it normal, which becomes a self-prophecy.

> "25-year-olds fresh of out of engineering school being christened 'Senior Project Manager' and put in charge of a large, very complex technical project"

This is so true. I have nothing to say except you are just hitting so many nails on the head.

> there is always a disconnect

This is incorrect. There are plenty of efficient, well run government institutions and projects. There are also plenty of inefficient and badly run (for the consumer) market-based companies.

I don't know. I agree with your statement about market-based companies. But if you could please cite the fabled "well run government institution", it would probably make your statement a lot stronger. I at least have seen quite a bit of DC and I don't think such a government agency exists.

Perhaps not in the US, and perhaps not now. But the first thing that came to mind for me was the Hong Kong MTR, which started as an extremely successful statutory corporation that only later was privatized (and that with dubious results). Same with the Housing Authority in the same territory, which also was well-run until it was privatized and begun to sell off assets to private enterprises that turned the screws on renters.

There's also the more fundamental question of whether a public good like transit should (and can!) be run for profit and at the same time benefit the entire population and economy. It's possible that's only feasible in some cases, not others. (Of course, NY should be able to turn a healthy budget given the population density.. the mismanagement is evident)

The IRS?

I think it's pretty amazing the number of government services most people spend almost zero time thinking about, and yet, benefit from nearly constantly in their day-to-day lives.

I think, in general, this is one of the attributes of a well-run government institution. It just works and it doesn't draw attention to itself unless necessary to fulfill the job.

In my mind, these organizations are so ubiquitous, and so common, that it seems impossible to not identify them. However, the narrative of "government is bad" seems to be so powerfully blinding that I suppose they can be hard to identify for certain people.

Reply on news.ycombinator.com