In recent weeks, I’ve been busily making what you might call a
negative argument for high-speed rail — pointing out the many ways in
which arguments against HSR are deficient. That’s all well and good,
but positive cases for HSR need to be made, as well.
Now, others have already begun to do this. California has released a lot of documentation and research related to its decision to construct a high-speed rail network. At a more casual level, Chris Bradford and Yonah Freemark
have shown that simply by making a few reasonable assumptions, the
stripped down models used by Ed Glaeser produce results that are far
more favorable to rail investment.
But I think it’s worth
stepping back and considering the decision to build HSR in context. The
biggest problem I have seen with critical arguments is that they ask
the wrong question, namely, given current conditions could a single HSR
line be profitable.
That’s not the way to make these
decisions, and indeed, if we had been making decisions like that all
along, much of the country’s existing infrastructure would never have
been built. So let’s see if we can’t arrive at a better understanding
of the question to which some of us believe HSR is the answer.
this country, we do not build transportation infrastructure for profit.
Perhaps this is upsetting to the libertarians among us, but that’s how
it is and how it should be.
Rather, we build transportation
infrastructure because we recognize that mobility is vital for the
economic, social, and political life of the country. The question then
becomes, what criteria does government use to build new infrastructure?
What is government trying to accomplish?
significant extent, these decisions are not wisely made. Local
governments often plan new roads based on what developers want and what
federal funding is available to them.
Planners often make
plans that attempt to figure out where growth will be or should be, and
their advice is occasionally heeded. Limited cost-benefit analyses and
environmental impact statements figure in decisions but are rarely
The point is this — the decision to build a new
highway or a new airport is rarely the result of a rigorous analysis of
the totality of costs and benefits, to individuals, taxpayers, and
That doesn’t mean that we should feel good about
building rail whether or not it makes economic sense. It does mean that
the existence of highways and non-existence of HSR should not be taken
as evidence of any kind of intrinsic economic goodness on the part of
So, having said all of that, let’s set out a few
key facts for consideration. First, current infrastructure appears to
be inadequate. Roadway congestion in America is quite costly. The Texas
Transportation Institute has estimated annual congestion costs around $80 billion.
That works out to annual
costs of approximately $10 billion for Los Angeles, $8 billion in New
York, $3 billion in Atlanta and Dallas, $2 billion in Philadelphia and
Boston, and so on down the line.
These numbers are open
to dispute; congestion costs are difficult to estimate. But it is clear
that significant amounts of valuable time and fuel are wasted everyday
due to crowded conditions on highways and city streets.
is worsening in large metropolitan areas, and serious congestion costs
can be found in progressively smaller metropolitan areas. Airline delays have also gotten progressively worse in recent years, particularly in our largest metropolitan areas.
congestion at airports and on roads can be reduced by increasing the
cost of using those modes. Congestion declined with increases in the
price of oil in 2007 and 2008. Unfortunately, the economy’s dependence
on oil meant that high oil prices contributed to the onset of a serious recession.
also worth noting that increases in the price of oil and decreases in
driving and flying took place alongside increases in usage of transit
and rail. This reflects that rail is a substitute for highway and air
travel, and that increases in the cost of the latter will boost demand
for the former.
A third piece of information: the American
population is forecast to grow by over 100 million people over the next
Currently, the Census Bureau projects that
the United States will be home to approximately 440 million people in
2050, about 133 million more people than currently reside in the
country. Based on recent trends, it seems likely that most of those
people will live in the country’s metropolitan areas.
on these pieces of information, it seems clear that if nothing is done,
congestion will continue to increase. Ultimately, this will lead to a
steady reduction in mobility within and between the nation’s
metropolitan areas, with negative consequences for economic activity
and social welfare.
The question, then, is this: what should be done about this state of affairs?
potential answer is what you might call the no-build solution — no new
capacity of any sort will be built. Governments might choose several
options within the no build solution; they could truly do nothing, or
they could choose to ration demand by pricing roads and airport slots.
either case, however, the cost of traveling by road or air would rise.
This could be reflected as an increase in time cost or monetary cost or
However the increase occurs, recent experience
suggests that the rising cost of travel by car or plane will lead to an
increase in demand for substitutes. In the absence of action, then,
demand for transit and rail will be enhanced (and cost-benefit analyses
should reflect this).
Note: I’m not mentioning oil prices at
all here. It’s not easy to predict where prices will go tomorrow to say
nothing of four decades down the road.
It does seem likely,
however, that oil price volatility may increase in coming years (and
has already increased since the placid days of the 1990s). In that case
it’s worth considering the value of an acceptable substitute to driving
and flying. The better consumers can adjust to swings in oil prices,
the less economic damage such swings will do.
Now, what if we
choose to build new capacity? This, it seems obvious to me, is the most
likely policy course. Governments will react to complaints about
growing congestion as they always have — by building more capacity.
question then becomes what kind of new capacity should be built.
Glaeser and others have attempted to show that high-speed rail does not
pass a simple cost-benefit test. I’m inclined to disagree with their
assessment, but to a certain extent, that question is beside the point.
Government is going to build more capacity. Given that, what is likely to be the best investment, all things considered?
alternatives, as it turns out, are not all that attractive. Roads do
not appear to pay for themselves any more than railways do. Receipts
from the federal gas tax come close to covering federal highway
expenditures, but gas is used on highways and non-highways alike,
indicating that at the federal level, highways are subsidized.
more detailed analysis by the Texas Department of Transportation led to
the development of the Asset Value Index, which was used to gauge the
life-cycle cost of Texas roads, highway and non-highway.
According to the Texas research,
none of the state’s roads paid for themselves, taking into account all
relevant taxes and fees. In some cases, gas tax rates would have to be
$2 per gallon or more before a piece of infrastructure would break
even. Obviously, a gas tax at that level would correspond to increased
usage of substitutes, like rail and transit.
Environmentally speaking, rail infrastructure is a much better option than new road or air infrastructure. Ed Glaeser concluded that there are positive environmental gains from building rail, and he was comparing rail to the no-build alternative.
If one assumes that some kind of new infrastructure will be built, then the case for rail becomes more compelling still. Life-cycle emissions for rail are considerably lower than those for driving options.
lower than flying options as well, and rail compares most favorably to
life-cycle emissions for small planes, which are most common on the
short routes that would be in direct competition with intercity trains.
Glaeser did cite research
out of Britain as a reason to be wary of the life-cycle emission
benefits of high-speed rail, but even there rail comes out fairly well.
research, a Booz Allen Hamilton study commissioned by the UK Department
for Transportation, shows that on routes where rail currently has a low
market share (which describes most of the routes being considered for
HSR service in America) rail can produce significant emission savings,
even taking construction of HSR systems into account.
particularly notable about this is that it compares HSR construction to
a no-build option, rather than a world in which air services are
expanded to alleviate airport congestion.
Very little of this
will satisfy those who want to see how rail can earn a profit before
moving forward. That’s fine, and as the title indicates, this is just
the beginning of the argument for high-speed rail.
important to understand the context in which the rail decision is being
made. America will be building new intercity transportation
infrastructure in the coming decades, of that we can be sure. In many
cases, the most attractive of available investment options will be