While I don't know the particulars of the Autobahn privatization controversy, I do know about toll booth and parking meter privatization in the United States, and yes, they tend to be bad deals for the tax payers, because the government wants a quick hit of money, in exchange for multiple times that money over the long term. Take for instance, Chicago's parking meter privatization, where the city got $1.15 billion in exchange for all the revenue from parking meters over the next 75 years In 2015, the meters brought in $156 million, so we're talking at least 7x.[0]
[0] http://chicago.suntimes.com/news/parking-meters-garages-took...
But a government has the task to always use the least costly bid. And that can also be "build it from scratch". In this example, that’d be cheaper.
But now you're saying they should buy vehicles, hire drivers, hire an iOS dev, an Android dev, a back-end coder, an ops team to watch the servers 24/7, a coder to handle payment back-end, a general manager to run the whole thing, pay AWS or GCP or a bare metal hoster for the servers, hire customer support to handle issues, pay Google (or Bing) maps commercial license fee, cover gas, maintenance, tires for all of the aforementioned vehicles and still have a lot of money left over from that fat $610,000 check to justify this project?
I think people suggesting "build it from scratch" mean build a local Uber-like service from scratch, which like Uber would not own vehicles itself or pay for gas or maintenance. That gets rid of the buying vehicles, and covering gas, maintenance, and tires.
Dispatching drivers could be done with pairs of text messages to the driver, of the form "pick up <NAME> at <ADDRESS>" and "drop off <NAME> at <ADDRESS>". The driver can tap on the addresses in those text messages to open them in the driver's smartphone's map application. That gets rid of the need for a commercial map license.
I don't see any need for a payment back end. The town can require drivers to accept payments directly using one or more specified methods (such as Square).
It probably doesn't need native iOS and Android apps. A web-based application should be sufficient.
They already have a website. They should be able to add the back end stuff to that, so no new costs for ops or hosting. (That is assuming that they have access to the code for their site and can make changes. Their site is run by Pathfive.ca, which seems to be a company that specializes in building and hosting websites for educational institutions and local governments. It may be that the town only supplies content to fill in templates).
Why hasn't anybody executed on that strategy (in practice)?
Forget the government affiliation, someone could be minting money either by competing with Uber on cost or just strategically choosing smaller markets that are not yet covered by Uber.
Why aren't they?
For comparison, in an average German town, a bus costs annually $80'000, including maintenance and driver. From scratch or rented from a company offering bus services.
a typical California city must contribute quite a significant sum so that the bus driver can retire at, say, 70% of their highest salary level. (e.g. over the past year, about 20% of Los Angeles's budget went to paying for the retirement of past employees.)
how does Germany do it?
The rest is handled by the employee.
Pensions are paid by the employees pension insurance, and the federal pension insurance, not by the employer.
There was a distant time where some bus drivers were also appointed Beamte but that practice was done away decades ago, so there are only very few such drivers left.
So what are the most successful examples of non flashy websites running on cheap hosting you'd recommend?
Things working the way they do, the government could not build a system that even approaches a service like Uber.
Although, I really don't know how to make the rules better.
Anyway, from personal, anecdotal experience, I have hired building contractors, first time around I took the cheapest offer, then two years later I needed the same work done for another room and then didn't take the cheapest offer (the one I took was about 25% more). The result could not have been more different. The cheap room has cost me more than 2x the original price, because I've had to redo most of it during the last 7 years and the second room shows no signs of needing any more work anytime soon.
So, taking the longer term view, sometimes it's better to pay more upfront to get more quality. Except that quality in this sense is really hard to quantify and put into rules.
I'm no particular fan of Uber. But the idea that some small city government could create a more economical mobile taxi app seems unlikely to me.
Now maybe this should be some shared local government infrastructure app but those sort of things aren't that common. I know I use checks to make various payments to my town because the online service actually costs me an incremental fee.
lolno
A government has the task to provide the best service for its citizens with its budget. We've seen what least costly bid leads to: degradation in quality of all public services impacted.
That is similar to saying that you're wasting money by paying a plumber to fix your pipes, instead of buying all the tools and fixing it yourself.
If the government doesn't have the capital, expertise, or budget for investment in infrastructure, it makes perfect sense to outsource it.
Why would a county pay a bus driver, mechanics, and invest in a fleet of buses when it's less expensive to outsource it?
That’s a good argument. Especially once economics of scale come into play.
But that’s not the case in this example. The government is outsourcing something that would be massively cheaper if purpose-built for their town.
Uber has a lot of expenses and complexity to handle corner cases that will never occur in this town – outsourcing is wasteful. That’s like building a heating system into a house in Phoenix.
I don't understand. Software has ultimate economies of scale. Also, they're not paying Uber a fixed cost, they're strictly subsidizing the ride. They can opt out at any time, if I'm reading the article correctly.
> Uber has a lot of expenses and complexity to handle corner cases that will never occur in this town – outsourcing is wasteful. That’s like building a heating system into a house in Phoenix.
One of us is misunderstanding the point of software. Just because you're not using certain capabilities doesn't mean you're charged extra for them. To your point, did you waste money by buying your IDE if you don't use each and every feature, because outsourcing your dev environment is "wasteful"?
But you are.
Such a system for a small town doesn’t need native apps, it doesn’t need to look beautiful.
You can use a single PhoneGap app for all OSes, also the same on the web.
You can avoid the entire payment infrastructure, and have riders pay the drivers directly. Alternatively, use Square, and you won’t have to make your own payment system either way.
A lot of uber’s costs only appear at huge scale.
This town doesn’t need uber, it needs a simple ride sharing service like every town in the second or third world has – and those even work without any app.
Explain that last bit. Economies of scale work very well for software. Whenever there's a new security patch or a feature that Uber implements, you get it for free if you use Uber as a platform, in that you don't pay any additional cost. Also, you don't need to pay money for maintenance/servers.
At this point I'm not quite sure if you're trolling or not. That'd be like advocating that a town's government use a home-coded text editor instead of Microsoft Word, because you don't need any of that fancy stuff that you get for no incremental cost.
If anyone of you is a city and wants that: Email me under the address in my profile.
That's not a valid point. You're subsidizing a good. That's like me saying that since I'm willing to buy my friend a beer, then beer is free.
Development time is most definitely not free, and neither is software update, ops, or support. Things that Uber gives for free.
Just a lot smaller in scope. See Mercedes' ridesharing offering in the EU, it's exactly that.
Only a few dozen devs, yet cheaper than Uber, with similar features.
Uber has already built an app. The city is subsidizing people for transportation. There is no more software development that needs to be done.
You comparing cheaper services to Uber is like comparing a large industrial farm to a local farm. Sure, the industrial farm costs more, but you're not paying for it. You're just paying for food stamps.
Why would you duplicate effort that already exists?
I don't see any way this discussion can continue in a useful way like that.
That said, you're not the only one - although it might also be due to the SV mindset.
> Why would you duplicate effort that already exists?
Because it's cheaper.
I have single-time costs to pay for developing such a system, and afterwards minor maintenance costs.
Or I can pay every year for Uber, and create high profits for their investors for them to build golden castles. If an investor makes profit from taxes, something's already gone wrong.
You're still suggesting an economy based on leasing, but that's not affordable in the long term.
Simply: Tax money is wasted. Already rich investors end up enriching themselves further, tax money ends up in the taxes of the rich instead of improving the life of the poor.
Sometimes it makes sense (see economies of scale), but often it does not, because the private contractor has to design one system that works everywhere, and has to deal with a lot more complexity due to that than a local system. Often expenses also occur elsewhere that wouldn’t in a local system.
And paying out to investors is also a common example of additional expenses. Especially if the state has to provide a guarantee the company will be a specific amount of money in profit from the riders of the project.
That’s also why in Germany right now there’s lots of protest against privatizing the Autobahn.