Comment by tptacek | Hacker News Reader

Comment by tptacek | Hacker News Reader

Parent Full thread

Why, yes. I do have data. Thank you for asking.

The top 10 cities in the US by GDP are, in order:

* New York

* Los Angeles

* Chicago

* Washington D.C.

* Houston

* Dallas

* Philadelphia

* San Francisco

* Boston

* Atlanta

Here, for your reading pleasure, are the ground travel distances between all of those cities:

    New York
     	Los Angeles		2475
     	Chicago			790
     	Washington D.C.		226 **
     	Houston			1416
     	Dallas			1389
     	Philadelphia		94 **
     	San Francisco		2907
     	Boston			215 **
     	Atlanta			750
    Los Angeles
     	Chicago			1745
     	Washington D.C.		2288
     	Houston			1379
     	Dallas			1235
     	Philadelphia		2401
     	San Francisco		380
     	Boston			2611
     	Atlanta			1947
    Chicago
     	Washington D.C.		696
     	Houston			1083
     	Dallas			924
     	Philadelphia		759
     	San Francisco		2130
     	Boston			983
     	Atlanta			716
    Washington D.C.
     	Houston			1400
     	Dallas			1328
     	Philadelphia		136 **
     	San Francisco		2816
     	Boston			437 **
     	Atlanta			639
    Houston
     	Dallas			239 
     	Philadelphia		1544
     	San Francisco		1927
     	Boston			1848	
     	Atlanta			792
    Dallas
     	Philadelphia		1464
     	San Francisco		1726
     	Boston			1791
     	Atlanta			781
    Philadelphia
     	San Francisco		2876
     	Boston			307 **
     	Atlanta			783 
    San Francisco
     	Boston			3099
     	Atlanta			2472
    Boston
     	Atlanta			1099

I have, I hope helpfully, starred the sub-700-mile routes that are already served by the Acela.

The average ground distance between these cities is (wait for it) 1,356 miles.

Later

You edited your comment upthread to include specific routes.

My response --- apart from wondering why taxpayers in Arizona should fund a rail linkage to Orlando, the 30th biggest economy in the country, to Tampa, the 27th --- is to suggest that adding Florida to this graph would not improve the average cost of the edges.

what does the average distance have to do with anything? Build between the cities where the distance makes sense. looks like there are several options. What about other population centers? Why is the gdp your criteria? Tampa and Orlando and Miami wouldn't make your list.

I said:

most of the economically important trips in the US are more than 600 miles apart, except for the part of the country that is already well-served by rail

You disputed this and asked for data.

I provided data.

If you would like to change the argument to "most of the important theme parks in the US are poorly connected to second-tier Florida cities by rail", I will happily concede that point. :)

In the graph of the 45 trips between top-10 US cities, only two --- LA to SF, and Houston to Dallas --- fall under your "less than 700 mile" criteria. Just 4% of the most economically important trips not already served by the Acela are amenable to rail transit.

Once again: most of the economically important trips in the US are more than 600 miles apart, except for the part of the country that is already well-served by rail.

Incidentally: if you think "top 10" is unfair as a threshold, consider that the 11th city would be Seattle, and the 12th Miami. The average would go up if we included them.

The average distance is irrelevant. Do the distance and population warrant the high-speed train service? Are there lots of flights between any two of these cities?

The population of Spain is 50 million, for example. They've made high-speed rail work effectively.

That's why my data is based on the top US cities by GDP. You can also find the busiest domestic airport pairs to back the analysis up. Oddly, the busiest air route in the US is NYC-Miami --- which is not amenable to rail travel.

I'm simply trying to identify the markets where HSR would work. For some reason, you identify as the critical factor. Then you identify NYC to Miami as the busiest air route, and Miami isn't on your gdp list. So, perhaps we only need one city from the list to another highly populated city?

At any rate, we've both identified a handful of US cities that meet both our criteria. Shouldn't we agree that's where to start building?

So, look: you've lost the context of the thread a little. Skim from the top:

* Someone commented that freight right of ways were impeding passenger rail in the US.

* I said geography had more to do with rail's status in the US than freight.

* You asked, "why are you talking about trips of 1000 miles or more"?

* I said "because those are the most economically important trips in the US".

* You said "no".

* I said, voluminously and conclusively, "yes".

Now you seem to think I'm opposed to SF-LA high speed rail. I am not, nor have I said that I am.

Where rail makes sense, it makes sense. If there's a cost effective way to get passenger rail rights of way from SF to LA, we should do that. We should get 200mph service from Chicago to MSP and to STL. We should get 200mph service from Houston to Dallas.

But even after we do that, rail is going to be a second-tier mode in the US, far surpassed by air, which will through economies of scale also be cost-competitive with HSR even in places where HSR is viable. The tactical routes we're talking about, the under-700-mile routes, will cost hundreds of billions of dollars to deploy, and will probably never recoup their costs or meaningfully change patterns of transportation in the US. HSR between SF and LA isn't going to make SF-NYC any less economically important, and SWA alone is going to remain more important to the economy than Amtrak.

Shit, I think I talked myself out of Chicago-MSP a little there.

Miami is popular as an air route because it's the hub to South and Central America.

I think it's safe to say it wouldn't be a rail hub to those regions. :-)

It's also irrelevant. NYC-Miami is a popular air route. Tampa-Miami is not. The inclusion of Miami in the list I gave upthread would, as I said, make the numbers worse for his argument.

Reply on news.ycombinator.com