The Leakage Problem

I’ve spent more than ten years talking about the cost of construction of physical infrastructure, starting with subways and then branching on to other things, most.

And yet there’s a problem of comparable size when discussing infrastructure waste, which, lacking any better term for it, I am going to call leakage. The definition of leakage is any project that is bundled into an infrastructure package that is not useful to the project under discussion and is not costed together with it. A package, in turn, is any program that considers multiple projects together, such as a stimulus bill, a regular transport investment budget, or a referendum. The motivation for the term leakage is that money deeded to megaprojects leaks to unrelated or semi-related priorities. This often occurs for political reasons but apolitical examples exist as well.

Before going over some examples, I want to clarify that the distinction between leakage and high costs is not ironclad. Sometimes, high costs come from bundled projects that are costed together with the project at hand; in the US they’re called betterments, for example the $100 million 3 km bike lane called the Somerville Community Path for the first, aborted iteration of the Green Line Extension in Boston. This blur is endemic to general improvement projects, such as rail electrification, and also to Northeast Corridor high-speed rail plans, but elsewhere, the distinction is clearer.

Finally, while normally I focus on construction costs for public transport, leakage is a big problem in the United States for highway investment, for political reasons. As I will explain below, I believe that nearly all highway investment in the US is waste thanks to leakage, even ignoring the elevated costs of urban road tunnels.

State of good repair

A month ago, I uploaded a video about the state of good repair grift in the United States. The grift is that SOGR is maintenance spending funded out of other people’s money – namely, a multiyear capital budget – and therefore the agency can spend it with little public oversight. The construction of an expansion may be overly expensive, but at the end of the day, the line opens and the public can verify that it works, even for a legendarily delayed project like Second Avenue Subway, the Berlin-Brandenburg Airport, or the soon-to-open Tel Aviv Subway. It’s a crude mechanism, since the public can’t verify safety or efficiency, but it’s impossible to fake: if nothing opens, it embarrasses all involved publicly, as is the case for California High-Speed Rail. No such mechanism exists for maintenance, and therefore, incompetent agencies have free reins to spend money with nothing to show for it. I recently gave an example of unusually high track renewal costs in Connecticut.

The connection with leakage is that capital plans include renewal and long-term repairs and not just expansion. Thus, SOGR is leakage, and when its costs go out of control, they displace funding that could be used for expansion. The NEC Commission proposal for high-speed rail on the Northeast Corridor calls for a budget of $117 billion in 2020 dollars, but there is extensive leakage to SOGR in the New York area, especially the aforementioned Connecticut plan, and thus for such a high budget the target average speed is about 140 km/h, in line with the upgraded legacy trains that high-speed lines in Europe replace.

Regionally, too, the monetary bonfire that is SOGR sucks the oxygen out of the room. The vast majority of the funds for MTA capital plans in New York is either normal replacement or SOGR, a neverending program whose backlog never shrinks despite billions of dollars in annual funding. The MTA wants to spend $50 billion in the next 5 years on capital improvements; visible expansion, such as Second Avenue Subway phase 2, moving block signaling on more lines, and wheelchair accessibility upgrades at a few stations, consists of only a few billion dollars of this package.

This is not purely an American issue. Germany’s federal plan for transport investment calls for 269.6 billion euros in project capital funding from 2016 to 2030, including a small proportion for projects planned now to be completed after 2031; as detailed on page 14, about half of the funds for both road and rail are to go to maintenance and renewal and only 40% to expansion. But 40% for expansion is still substantially less leakage than seen in American plans like that for New York.

Betterments and other irrelevant projects

Betterments straddle the boundary between high costs and leakage. They can be bundled with the cost of a project, as is the case for the Somerville Community Path for original GLX (but not the current version, from which it was dropped). Or they can be costed separately. The ideal project breakdown will have an explicit itemization letting us tell how much money leaked to betterments; for example, for the first Nice tramway line, the answer is about 30%, going to streetscaping and other such improvements.

Betterments fall into several categories. Some are pure NIMBYism – a selfish community demands something as a precondition of not publicly opposing the project, and the state caves instead of fighting back. In Israel, Haifa demanded that the state pay for trenching portions of the railroad through the southern part of the city as part of the national rail electrification project, making specious claims about the at-grade railway separating the city from the beach and even saying that high-voltage electrification causes cancer. In Toronto, the electrification project for the RER ran into a similar problem: while rail electrification reduces noise emissions, some suburbs still demanded noise walls, and the province caved to the tune of $1 billion.

Such extortion is surplus extraction – Israel and Toronto are both late to electrification, and thus those projects have very high benefit ratios over base costs, encouraging squeaky wheel behavior, raising costs to match benefits. Keeping the surplus with the state is crucial for enabling further expansion, and requires a combination of the political courage to say no and mechanisms to defer commitment until design is more advanced, in order to disempower local communities and empower planners.

Other betterments have a logical reason to be there, such as the streetscape and drainage improvements for the Nice tramway, or to some extent the Somerville Community Path. The problem with them is that chaining them to a megaproject funded by other people’s money means that they have no sense of cost control. A municipality that has to build a bike path out of its own money will never spend $100 million on 3 km; and yet that was the projected cost in Somerville, where the budget was treated as acceptable because it was second-order by broader GLX standards.

Bad expansion projects

Sometimes, infrastructure packages include bad with good projects. The bad projects are then leakage. This is usually the politically hardest nut to crack, because usually this happens in an environment of explicit political negotiation between actors each wanting something for their own narrow interest.

For example, this can be a regional negotiation between urban and non-urban interests. The urban interests want a high-value urban rail line; the rest want a low-value investment, which could be some low-ridership regional rail or a road project. Germany’s underinvestment in high-speed rail essentially comes from this kind of leakage: people who have a non-urban identity or who feel that people with such identity are inherently more morally deserving of subsidy than Berlin or Munich oppose an intercity high-speed rail network, feeling that trains averaging 120-150 km/h are good enough on specious polycentricity grounds. Such negotiation can even turn violent – the Gilets Jaunes riots were mostly white supremacist, but they were white supremacists with a strong anti-urban identity who felt like the diesel taxes were too urban-focused.

In some cases, like that of a riot, there is an easy solution, but when it goes to referendum, it is harder. Southern California in particular has an extreme problem of leakage in referendums, with no short- or medium-term solution but to fund some bad with the good. California’s New Right passed Prop 13, which among other things requires a 2/3 supermajority for tax hikes. To get around it, the state has to promise somthing explicit to every interest group. This is especially acute in Southern California, where “we’re liberal Democrats, we’re doing this” messaging can get 50-60% but not 67% as in the more left-wing San Francisco area and therefore regional ballot measures for increasing sales taxes for transit have to make explicit promises.

The explicit promises for weak projects, which can be low-ridership suburban light rail extensions, bond money for bus operations, road expansion, or road maintenance, damage the system twice. First, they’re weak on a pure benefit-cost ratio. And second, they commit the county too early to specific projects. Early commitment leads to cost overruns, as the ability of nefarious actors (not just communities but also contractors, political power brokers, planners, etc.) to demand extra scope is high, and the prior political commitment makes it too embarrassing to walk away from an overly bloated project. For an example of early commitment (though not of leakage), witness California High-Speed Rail: even now the state pretends it is not canceling the project, and is trying to pitch it as Bakersfield-Merced high-speed rail instead, to avoid the embarrassment.

The issue of roads

I focus on what I am interested in, which is public transport, but the leakage problem is also extensive for roads. In the United States, road money is disbursed to the tune of several tens of billions of dollars per year in the regular process, even without any stimulus funding. It’s such an important part of the mythos of public works that it has to be spread evenly across the states, so that politicians from a bygone era of non-ideological pork money can say they’ve brought in spending to their local districts. I believe there’s even a rule requiring at least 92% of the fuel tax money generated in each state to be spent within the state.

The result is that road money is wasted on low-growth regions. From my perspective, all road money is bad. But let’s put ourselves for a moment in the mindset of a Texan or Bavarian booster: roads are good, climate change is exaggerated, deficits are immoral (German version) or taxes are (Texan version), the measure of a nation’s wealth is how big its SUVs are. In this mindset, road money should be spent prudently in high-growth regions, like the metropolitan areas of the American Sunbelt or the biggest German cities. It definitely should not be spent in declining regions like the Rust Belt, where due to continued road investment and population decline, there is no longer traffic congestion.

And yet, road money is spent in those no-congestion regions. Politicians get to brag about saving a few seconds’ worth of congestion with three-figure million dollar interchanges and bypasses in small Rust Belt towns, complete with political rhetoric about the moral superiority of regions whose best days lay a hundred years ago to regions whose best days lie ahead.

Leakage and consensus

It is easy to get trapped in a consensus in which every region and every interest group gets something. This makes leakage easier: an infrastructure package will then have something for everyone, regardless of any benefit-cost analysis. Once the budget rather than the outcome becomes the main selling point, black holes like SOGR are easy to include.

It’s critical to resist this trend and fight to oppose leakage. Expansion should go to expansion, where investment is needed, and not where it isn’t. Failure to do so leads to hundreds of billions in investment money most of which is wasted independently for the construction cost problem.


  1. Matthew Hutton

    I’m not sure you are right about roads. Rural America having good roads means you could link up things like the national parks with full cama style overnight buses and that would probably be more plausible than improving rail infrastructure to do 100mph

    • Alon Levy

      The main links were completed generations ago – already in the 1950s, spending was a mix of speed and capacity, and since then it’s been almost entirely capacity. Today, even mixed projects like I-69 are pretty rare, and most spending is adding lanes, upgrading interchanges, and other projects that are sold as ways to reduce traffic congestion where in most of the areas where the money is spent there isn’t any.

      • Matthew Hutton

        I’m more thinking the secondary routes which have passing lanes and are well built in America to allow a 65mph average speed vs 40mph on secondary routes in Britain. And if there are medium sized leisure destinations I’m sure those roads could support a good bus service – especially a sleeper service that is difficult to self-drive.

        I agree massive junctions is a waste for towns and cities with 10 minute traffic jams from 8-9am only.

  2. Eric2

    “In this mindset, road money should be spent prudently in high-growth regions, like the metropolitan areas of the American Sunbelt or the biggest German cities. It definitely should not be spent in declining regions like the Rust Belt, where due to continued road investment and population decline, there is no longer traffic congestion.”

    It actually makes a lot of sense to spend money where growth is happening. The issue is further upstream – the policies that determine where growth is happening. Specifically, in the US, land use policy prohibits large coastal cities from growing in population despite the massive demand (as measured in housing prices) for such growth. Remove the land use restrictions and the massive growth will be on the coasts rather than in the Sun Belt. And to the extent there is an honest demand for “prudent spending in high-growth regions”, that demand will focus on coastal cities and to a large extent urban transit.

    (As usual, about 90% of transit policy is just derivative of land use policy)

    • Alon Levy

      Sure, but Texas, Florida, and Georgia will continue growing no matter what – these are not poor areas. Zoning-constrained coastal cities are barely expanding freeways – LA occasionally adds lanes at high costs, but new freeways are prohibitively expensive.

      • Eric2

        I don’t think that’s true. First of all, I have seen statistics that Texas at least is a poor area, the workers it is attracting are low-income on average. Second of all, US population increase is dropping like a stone, and if the coastal cities grow substantially, there will be no remaining increase to be distributed among the Sun Belt cities.

        • Henry Miller

          Not everyone wants to live on the coast. The sun belt can grow at the same time as the coasts. Even if the coasts fixed their zoning so enough things actually got built, that wouldn’t destroy the value of the sun belt for some people.

          • Eric2

            I am highly confident that the potential demand in the major coastal metropolises is enough to absorb the entire country’s population growth. That might mean those metropolises together increase from 30% to 40% of the US population. “Not everyone wants to live on the coast”, sure, but I’m pretty sure 40% do, as long as a large share of jobs and cultural attractions are on the coasts, which they are and will be.

          • Car(e)-Free LA

            They won’t be. Look at the actual topographic lines.

        • Alon Levy

          Houston, Dallas, and Austin all have higher metropolitan per capita incomes than the US average. They’re not much richer than the US average, especially (for Houston) this side of the oil price collapse of 2014-5 – beforehand Houston was richer than Chicago – but they are richer. They’re especially richer if you only look at labor income, which exclusion mostly screws dedicated retirement regions like Florida, but even on all-source income, they’re richer. They’re not Vegas or Phoenix, whose growth is genuinely just a push factor out of California.

          • Reedman Bassoon

            Florida is now the third most populous US state (surpassing New York). Is Florida (no state income tax) getting a “push factor” from people leaving (high tax) New York, New Jersey, Connecticut? P.S. Florida is the “most coastal” of the 48 contiguous states. It just splits it between the Atlantic Ocean and the Gulf Of Mexico. Alaska is in a class by itself.

      • Nathanael

        Climate change predictions say that Texas, Florida, and to a lesser extent Georgia are gonna have to lose population — no real choice, Texas lacks the water and Florida lacks the landmass once it floods — and it’s going to move to the Great Lakes region. That is, the Rust Belt. Expect it. I mean, Phoenix will probably depopulate first, but if you’re long-term planning, you know what’s coming.

        • Eric2

          Water is not an issue – solar/wind will power desalination if necessary (this is much cheaper than moving the people). I’m not sure why you think Georgia will suffer, little of the population lives near the coast. Similarly Phoenix, they will just keep their air conditioning on. Florida will have real issues though.

          • michaelrjames

            @Eric2: “Water is not an issue – solar/wind will power desalination ”

            Quite apart from the energy implications of large-scale desalination (which would ratchet up carbon pollution even more), desalination has to happen on the coasts. Perhaps this is not so much an issue for Florida but it would be a big thing for Texas. It would also ramp up the amount of energy used, for all that pumping uphill and over long distances. That would then require an entire new long-distance water infrastructure to be built. The surprising thing about the US is how reliant upon groundwater* so much of it is. That means that there is no existing mains water infrastructure for large parts of the country. The water issue in the US is absolutely huge and it boggles the mind, yet very little is being done about it. Denial and NIMTOO is the current response. Unless you’re talking about desalination (ie. reverse osmosis) as a means of recycling waste water in cities. But Texas and Florida won’t consider any limits on consumption, say like CA, so what chance for asking them to use recycled water from their sewers? In fact LA puts its purified waste water back into the aquifer to help recharge it, or at least slow its depletion. It is also another (big) step in purification in that it will be decades or centuries before that water re-emerges. (BTW, this is another aspect in which CA is progressive.)

            *The Ogallala and Floridan aquifers in Texas and Florida are being depleted at approximately ten times their natural rates of replenishment. It’s the very definition of unsustainability. By resorting to massive and ever-deeper pumping projects the crunch may be forestalled for another few decades but that is part of the problem, in that it leads to denial and failure to act today.
            Incidentally, while climate change gets blamed for the Florida problem of saltwater intrusion into its aquifers (and sometimes bubbling up into the urban areas), actually the main cause is overdrawing on the aquifer. In a low land like Florida it means that at some point it causes saltwater (from ocean or marshes) to flow into the fresh, ruining the entire aquifer.

          • michaelrjames


            I forgot to mention that the collapse of that Miami apartment block is probably related to this same issue. The entire substrata of Miami (and most of Florida) is limestone which has pros and cons. It is the reason Florida has a gigantic freshwater aquifer to draw upon. Without it Florida would not be developed. But then overdrawing on it creates the saltwater intrusion problem and the sinkhole problem. Oh, and why Miami is sinking (much more than climate change is causing the ocean to “rise”; watch out for when climate change really kicks in).
            The news reports about the building collapse are all about deaths and then about building standards and another huge American problem, lack of maintenance. However I still believe they’ll discover that geology was the ultimate cause of collapse. Unless there is a big coverup because the implications are both huge and unthinkable, with no quick fixes.

          • michaelrjames

            Here’s another story about water shortages.

            The California tourist town that’s running out of water: ‘It’s a shock’ Lush Mendocino draws nearly 2m visitors a year. But drought is threatening to sink its key industry
            Maanvi Singh, 08 Aug 2021.

            As they explain, the groundwater is not deep because of the geology and so there is no solution from digging deeper, and this too is why it is running dry (it simply is not a big store like many aquifers; it runs off into the ocean).

            But it is curious. I thought that part of Northern California coast was quite wet–and when I check it out, it is: Fort Bragg (closest big town to Mendocino, and where they currently truck water from) gets 1100mm and 111 days of rain. For comparison, most of Southern CA gets 300-400mm and ≈35d rain; Seattle gets 948mm and Vancouver BC gets 1460mm and 160d rain. But it’s not even mentioned in the article! So a solution is right there (unless Mendocino is in some kind of weird rain shadow?)
            The thing is that >40% of household water use is toilet + laundry and this can be served by untreated rainwater. Perhaps it is not as big a solution for a restaurant but still. Here (the driest continent) it has been determined that to cope with recurring droughts household water tanks should be >30,000L, ideally >50,000L, but these things are relatively cheap (these days the replumbing of wc and laundry will probably cost more!). Is this another case of the American mind being incapable of thinking beyond existing experience?

    • Aaron Moser

      Say what you will about O’Toole but when it comes to the urban growth boundaries he is comically right. UGB are even stopping transit oriented development in places like Portland ironically(look on google satellite how much farm land is close to downtown Portland) but the urbanists will never get rid of them because its part of their religion.

  3. fjod

    You mention about money being ‘wasted on low-growth regions’. What do you make of the argument that by maximising ROI you end up repeatedly boosting the same place and thereby creating or entrenching low-growth regions?

    Say (vastly oversimplified) a new metro line boosts productivity by 5%. If you build it in city A (avg salary $50000) you boost the economy by $2500/head. If you build it in city B (avg salary $100000) you boost the economy by $5000/head. If you are looking to maximise ROI you will always build it in city B. Now city B is even richer, and hence even more deserving of ROI-maximising investment in the future.

    This is not a theoretical question – it has been identified in the UK where salaries are higher in London and so this effect has emerged over the last few decades of transport funding decisions assessed primarily by benefit-cost ratio. The effect has even caused changes to investment appraisal methods.

    • keaswaran

      This argument seems to take cities as the moral unit of evaluation. If you take human individuals as the moral unit of evaluation, and if there are human individuals whose choice to locate in City A or City B is determined by the investment in housing and infrastructure in the cities, then I think the question disappears.

      If some cities can absorb investments in ways that expand more opportunities to more people, and others can’t, then it seems in many ways akin to the issue where some cities need expensive mitigation of flooding or drought conditions in order to accommodate the same number of new people as other cities in more favorable locations, so that we should be focusing development in the places where people can better be accommodated rather than splitting it equally between a good place for people and a less good (or actively bad) place for people.

      • fjod

        Most people don’t think like that though. Most, if not all, people have something (support networks, families, non-work commitments, language) tying them to a particular place, so they do not simply choose to locate in city A or city B based on financial reward. Humans do not work how you suggest (it’d be much simpler if we did!).

        And repeatedly boosting one area of a polity at the expense of others causes negative externalities. You get poor opportunities in peripheral areas, often to the point where they end up shrinking with all the crap that causes – Detroit is the obvious example of a city crippled by shrinking pop. You also get huge housing costs in the boosted core (which of course makes moving from core to periphery much harder). There are other ways this sort of bad outcome can be wrought apart from with infrastructure; the euro has done the same in making Germany overly competitive at the expense of e.g. Greece.

        • Nilo

          Detroit was harmed by its suburbs, not due to “ROI” investment, and using the fact the British has one of the world’s most destructive housing policies doesn’t seem like a good excuse either. Japan has basically allowed anyone to move to Tokyo and Osaka, such that they remain stagnant or weak growing despite a shrinking country.

          • fjod

            I didn’t say Detroit was harmed by ROI investment, I said it’s a good example of some of the problems a shrinking city can cause. And the high-rents-in-the-boosted-core phenomenon is hardly unique to the UK (I don’t think I even implied it was).

            Japan also – quite famously! – does low-ROI investment in declining rural areas.

        • adirondacker12800

          Detroit isn’t on the periphery. It’s at the center of the country’s 14th largest metro area.

          • Herbert

            Can you name the fourteenth largest metro area in any other country? Would you recognize the name of the town at its center in most places?

          • adirondacker12800

            Easily, the same way I checked that Detroit has apparently moved down to 14th in the 2020 estimates, with Wikipedia. 14th in the U,K. is Edinburgh, 14th in Canada is Oshawa, Matsuyama in Japan, Montpelier in France. 14th in Austrtalia is a wide place in the road in Queensland. It’s Metropolitan Statistical Area is the same size as Hamburg’s EMR and it’s Combined Statistical Area is the same size as Stuttgart or Munich. Give Canada a whirl

          • Matthew Hutton

            And to be fair Edinburgh and Montpellier are places I think lots of people will have heard of.

    • Alon Levy

      The US doesn’t do benefit-cost analyses and spends road money disproportionately in low-growth regions – Texas is a massive net donor of fuel taxes (I’m told on Twitter it’s the only state where the 92% limit is binding). And it’s not even low-growth as in zoning-constrained coastal cities, which don’t build new roads, but depopulating Rust Belt regions. And yet the same effect is obtained there.

      And even with this tax subsidy, Americans are certain that the Rust Belt didn’t fail but was failed by outside forces, like globalism, high finance, and other stand-ins that a few generations ago would just have been called the Jews. Today it’s not PC to say that so the populist grifters say the same with the word Jews stripped away. In Upstate New York most people believe they’re tax-subsidizing the city when the opposite is the case.

      • Herbert

        The Ruhr had and has many of the same problems the old industrial regions of Silesia, northern England or the US have and had. But the Ruhr is doing better and its reputation now even includes stuff like “culture” and “greenery”. What’s different in the Ruhr? Local policies? Federal policies?

        • borners

          @Herbert I’d say both local and federal policies worked quite well together. In transport they do a much better job than the North of England does intergrating the centres together with trains and usings light-rail within the cities. Cologne Stadtbahn kills the Manchester Metrolink in ridership/km. And well Midwest only has one regional rail line worthy of the name. That can’t solve all your problems but it helps.

          If you look at the Macro-economic story of say UK vs Germany, the story is that postwar Germany had an open economy with new co-ordination institutions like industry-wide unions, whereas in the UK focused on preserving the 19th century industrial structure of outer Britain until Thatcher when she put them out of the misery (90% of the jobs in the staples are gone by 1970). negotiated a shift to niche-industrial products with high-value complimentary services, for instance many historic German textile firms from the Ruhr survive often as service sector firms in textiles. First world economies preserve industrial sectors with services which outside Finance Germany actually outperforms the UK. Good passenger rail integrates regional labour markets. Investing in downtowns attracts service sector professionals. That said the Ruhr is still in relative decline, the one thing they got wrong was universities.

          Midwest has toxic racial politics, crappy local government design etc, only Minnesota is doing well. But it was New England levels of white until the 2000’s and had its big-city its university town and its state capitol in the same urban area. If Chicago were not governed so badly it would be fine and could maybe drag Wisconsin and Indiana with it. Ohio is screwed.

          • Herbert

            Just because Bochum university is a crime against architecture doesn’t mean it’s a bad university.

            Also iirc the Stadtbahn tunnels were planned in part to give former miners new jobs digging them… It didn’t work and now their maintenance is a drain on municipal budgets

          • borners

            Having an underpowered or overpowered university system is a product of two things, quality and per capita scale, not just in the absolute but in relative terms within a country. Ruhr is at a disadvantage per capita versus Hesse/Baden-Wurttemberg for sure and probably Berlin too.

            As for tunnel problems. Writing from an English perspective, that’s the kind of problem we’d love to have. I’ve just been been comparing lightrail/tram systems and while UK lighrail doesn’t suck at American levels, its still bad by Central European standards.

            @Alon Levy, do you have better phrase for Denmark/Austria/Switzerland etc than “German periphery”? I know you haven’t used that pejoratively since you’ve been clear that these countries have made the best use of German innovations in public transport. I used Central Europe, but I don’t think that’s very good either.

          • Herbert

            Nobody can agree where central europe is, but most agree it somehow contains Germany

          • Nathanael

            Ohio’s problem is that it’s still dominated by rural interests, and the demographics aren’t improving. If the cities of Ohio (Cleveland, Cincy, Columbus, Dayton, Toledo, Akron/Canton, and Youngstown) had sufficient demographic size and political clout, it would have a chance of making sane decisions, but they’re just slightly less in population than the delusional rural-identity people (mostly racists).

      • Nilo

        You should see what people in Illinois south of Chicago think about the city.

        They’re all convinced they’d be better off as another state. Said state of course would be in heavy competition with Mississippi for the poorest in America.

        • Frederick

          On the other hand, USA will be better off if Illinois splits into two states, as Chicago will no longer get dragged down by the burden. Ditto for NYC and upstate New York.

          State boundaries in 1800 already didn’t quite make sense (because they were colonial vestiges) but coming into the 21st century, the old state boundaries have become so stupid and they are hampering America’s growth.

          • michaelrjames

            Or do like the Chinese: any megacity becomes its own province (state). This could solve permanently the Dem’s senate difficulties. Or does it?

          • Eric2

            No, that would hurt Democrats. Illinois now elects two D senators, if split then it would elect two Ds and two Rs. Similarly with NY, CA, etc. I think in general most state splits that aren’t crazy gerrymanders would help Republicans, because Republicans are the majority in most US territory.

            Independently this would also have the bad effect of making Senate races noncompetitive nearly everywhere.

          • michaelrjames


            Hold on. While it may net to zero in some states, of which Illinois is probably one example, that is probably not the case in CA where the LA and Bay Area would result in at least a 2-seat, maybe 3-seat gain. Likewise, Texas is approaching the cusp of going purple and it is mostly happening in their big cities, so any gain in a city is a net gain. Likewise FA and probably GA where the current Dem senators are in a very precarious situation. Not to mention completely new Dem seats in Washington-DC. Would NYC and Boston result in net zero for those states? The US continues to urbanise and so even some of the current situations of big red cities will shift purple then blue over time.
            Someone has probably already modelled this ….

          • Alon Levy

            As it happens, I crayoned state boundary redrawing last year!

            I drew this without looking at who voted what; this applet says Biden wins 15/30 states on this map, the same share he actually won. You can do small tweaks in either direction, for example if you move West Texas back to Texas then the resulting Arizona-NM state is blue, and if you donate enough of Minnesota to Colorado and Illinois it flips back to blue, whereas if you do some Appalachian tweaks you can make the Piedmont state red and with enough expansion you can flip Michigan and Upstate NY to red.

          • michaelrjames

            @Alon re electoral boundaries

            Right. Interesting but merging states or even subregions is a non-starter. Years ago I read something similar but specific to the state most obviously crying out to be divided: California. While NoCal and SoCal is the simplest it has obvious political difficulties but the alternative–I think of between 3 and 5 so as to give something to conservatives, but then that isn’t realistic either. In its own way it is a form of gerrymandering though equi-population electorates in CA would always favour Dems.

            My idea is a bit different in having it city-based. It is often said that city government is the only one that is at all functional in the US. Hence the Bloomberg association of city mayors and a world version too. It seems to have overcome the usual petty partisan divisions and the participants reported satisfaction. It is suggestive that giving them more power over national affairs might be more productive than current arrangements. And, inevitably the world is urbanising such that we”ll all be city-slickers before long. Maybe we’ll eventually revert closer to the feudal arrangement where all those outside the city walls will have to be associated with their nearest city to have any political input ie. protection from their overlord!

            Here it is:

            Plan to split California into six states could end up on 2016 ballot
            Billionaire Tim Draper collects 800,000 signatures, which could put plan experts call a ‘colossal waste of time’ on 2016 ballot
            Amanda Holpuch, Wednesday 16 July 2014

            “The chances of this passing are zero, zip, nada,” said UC Berkeley law professor John Yoo in an email. In an article he co-wrote on the subject, Yoo said that Draper’s argument that California is “nearly ungovernable” doesn’t stand up to criticism because that claim can’t be proven.
            If Draper’s measure does make it onto the ballot and is supported by a majority of California’s more than 17.7m registered voters, one large thing stands in the way of his plan: the federal government. The US constitution requires the consent of a state’s legislature and Congress to create new states, meld two states together or make other complicated modifications.

          • Alon Levy

            I don’t think city governance in the US is especially functional. The only place I can concretely point out to you where city > state is New York City, but only because the governor happens to be Cuomo; de Blasio is bad and Adams is a corrupt machine politician who is so traumatized by almost losing the election he wants to change the electoral system. In the US about the only places where government even approximates something that works are ones where a) there’s interpartisan competition and b) the Democrats are currently winning (Congress, Virginia, Minnesota, etc.) – without such interpartisan competition politics becomes just a contest of who can dole out jobs to supporters.

            In France, of note, the lack of local empowerment has been a big benefit to intercity transport and housing, and some of the worst abuses are local, not national. Mandatory pork at schools is not state policy – the food halls are at the commune level, so from time to time a racist mayor makes Jewish and Muslim children eat pork and gets told to stop by state courts.

            So even if you have to have subnational government, it should be at high enough a level that there’s meaningful ideological politics, more regional than local.

            (P.S. speaking of the article you link to, the reference to John Yoo as “UC Berkeley law professor” rather than the more significant “Bush torture lawyer” is just bad.)

          • michaelrjames


            We’re just talking about different quanta of dysfunctionality. The point of that article about dividing CA into 6 states (!) was that it was in response to its perceived total dysfunction and ungovernability; and here it is going into another Recall election. This, the richest and most progressive state, not just in the USA but world. We’ve seen federal politics in a total mess, at least since GWB if not Reagan, maybe Nixon. And still Biden will do nothing about the absurd and anti-democratic filibuster, and predictions are that the usual gridlock will resume after next midterms (!?). Obviously NYC is hardly a posterchild for great government but it looked pretty good in US terms during Bloomberg’s reign.

            One can also look at very successful city-states like Singapore and formerly HK. The argument has been that the unit of government that is closest to the business end of governing, ie. service delivery, has to work the best because feedback on dysfunction is pretty much instantaneous. Then scale delivers a pool of competent people to run the joint.

            I wouldn’t discuss France in this context because, as you know, I think it is one of the better run nations, and at all levels. It has actually managed cohesive governance across nation-province-town-city quite well. (I’ve wondered whether France’s peculiar dual mandates may be partly responsible. National pollies who are also mayors etc.) The Anglosphere can’t run any halfway competent rail system (the UK has been living off legacy for >50y) but even without takt or EMU-TGVs (les pauvres!) France still runs a quite good city and inter-city and international rail system:-)

          • Eric2

            Nice crayon, but keep in mind these 30 states have drastically different populations. For example the state centered on St Louis appears to have population ~4 million while the state centered on LA has population ~24 million. So if this is an attempt at making the Senate fairer, it’s barely an improvement over the current situation. Also, the population distribution is changing all the time, which means that tweaks that work now will stop working in a decade or two (similar to congressional gerrymanders, which are redone with every census).

            Now if the point is to unite economic areas so that they can manage their affairs more efficiently, this crayon would do a good job. But I doubt the gains are enough to justify the disruption involved.

          • Herbert

            City states always have the freeloader issue. Well, sub national city states anyway. The main reason Brandenburg has a tax base at all are people living just across the state line from Berlin and paying taxes in Brandenburg. It’s even worse for Bremen, but then Bremerhaven is a bit of a rust belt

          • Herbert

            Where on earth did you get the idea that California is the most progressive state in the world? Is it more progressive than the Netherlands? Catalonia? Suriname?

          • Alon Levy

            It’s about as progressive on climate as Bavaria. On some other issues, like access to education and transportation in general, California is to the right of Bavaria.

          • michaelrjames

            Re Progessiveness of CA.

            I knew some would object to my throwaway line. But I think many lose sight of the historic influence of California. For example where they lead on vehicle emissions the rest of the US inevitably follows (even with the name “CAFE standards”), and so the world. It’s true on many enviro concerns. On education, Alon is talking about the post-Reagan era when their inability to raise taxes has put horrible financial pressure on what was once the best university system, not only in the US, but arguably the world. IMO, the three-tier UC system is an excellent model and for other jurisdictions more willing to fund such a system.
            Not to mention it being the epicentre of many cultural (’60s youthquake, protest movements, music, arts) and scientific (biotech, biomed etc) and hi-tech (Silicon Valley, HP, Apple, Google, FB etc). Yes, it may be a mixed bag what it does with all this creation but there’s not another place that wouldn’t wish to have a fraction of it.

          • Alon Levy

            CAFE standards are the progressivity of 40 years ago. In the 21st century, California’s environmental policy is horrific – so NIMBY people on net move from temperate coastal areas to inland areas with heavy AC and car use, unable to effect change toward public transit, not good on energy for how much solar incidence it has.

            The UC system is still good, the problem is that it charges thousands of dollars per semester. And the primary and secondary education system is, because of the New Right (not exactly Reagan, but Prop 13, which is ideologically the same), bad.

          • michaelrjames


            As I have come to realise with some of our differences of opinion, they are based on generational experiences and perceptions. On this one I probably concede, in that undeniably CA has gone backwards on many things. Most notably and sadly education, because the money thing increasingly negates the original spirit of it: excellent education opportunities for all, and on a merit basis.
            I seem to recall that CAFE standards were still being contested during the Trump era, ie. the latest update on them by CA; and CA was the lead state in contesting that rollback in the courts (presumably since reversed by Biden?).
            I haven’t checked recently but CA was always way ahead of all other states on electricity consumption per cap. Remarkably, and unlike almost anywhere else has not increased its per capita consumption over decades.

            In a recent post you wrote that Texas is the biggest net contributor to federal fuel taxes. Of course that reflects that they drive more, a lot more and probably with less efficient and/or bigger vehicles, than anyone else (probably on the planet). A lot of those will be commuters in its ridiculously sprawled and car dependent cities. I also read that Texas produces more electricity than all of the UK (which is about 3x its size) as if that is a good thing instead of just massively inefficient and wasteful. IIRC Dallas uses more than 3x electricity per person than NYC (here it is: Dallas 16,116 kWh, NYC 4,696 kWh per household). Of course its grid is standalone so it can’t share with neighbouring states! The paradox is that it also produces lots of wind power but then just wastes it.
            Texas uses 483TWh annually compared to CA’s 202TWh despite the much smaller population. Per capita TX uses 3.3 times more than CA. CA’s per capita use is the same as the EU’s median use. Pretty good for the largest state in the U.S.A. Dare one also mention it is where the US’s hopes for EV became serious …

      • df1982

        I’m not sure why you’re bringing anti-semitism into this: I think the people who live in the rust belt are all too aware that it was the heads of the auto companies and other manufacturers who consciously de-industrialised the region by outsourcing production to lower-wage countries (not to mention it’s perfectly possible to be against finance capital and neoliberal globalisation without being latently anti-semitic). If there’s any racism around this it comes out as anti-Mexican sentiment, not anti-Jewish.

        But the bigger issue is, what do you do with a region that has fallen victim to economic decline? Do you just starve it of all funding and create a vicious circle because you can supposedly get a higher rate of return on investing in growth regions, with the result that tens of millions of people either have to leave their homes or suffer unending economic depression? Or do you try to invest in such a way that the depressed areas perform an economic turnaround?

        By the way, you’re living in a city where the latter alternative is precisely what happened. Berlin in 1990 was far poorer than the likes of Frankfurt and Munich, its economy was stagnant and its population was shrinking. But the concerted public investment program in transport links and other infrastructure has helped to resuscitate its fortunes, to the point that it’s now a magnet for much of Europe’s youth. Sure, this was mainly motivated by an imperial vanity project of restoring the old capital to its former glory (and contrasts with the asset stripping carried out on the rest of the former GDR), but there’s no reason the same thing can’t be done for Detroit, Cleveland, et al. Or should the German government have consigned Berlin to a death spiral while pumping money into the western boom cities?

        • Alon Levy

          So, the situation in Berlin is unique, for a couple of different reasons:

          1. The city inherited the infrastructure of a dominant capital that, before the war, was among the world’s top 10 cities. As soon as the Wall fell, Wessis swarmed Mitte, Prenzlauer Berg, and Friedrichshain, for their proximity. This isn’t true of any post-industrial American city, unless you count Philadelphia. Detroit notably lacks this infrastructure, because it grew too late (1910s-40s, not 1850s-1900s) and was one of the origins of job sprawl, so other than abundant substandard housing it doesn’t have Berlin’s production amenities.

          2. West Germany kept its capital in abeyance, distributing institutions between different cities expecting reunification, rather than just consolidating everything in Frankfurt, so Berlin was a natural place for the capital and associated infrastructure (new development where the Wall had been, Hauptbahnhof, BER, etc.).

          3. West Berlin was not poor. The economic statistics are really a story of reintegration of the East, which happened everywhere in East Germany – but this also includes depopulating parts, like Saxony-Anhalt, that do not feel like they’re growing even if incomes are up by a lot. Wages were raised to compete with Western employers, which is a good thing and is also not what the local elites who are paying these wages want when they demand state investment in infrastructure to be paid to contractors they own.

          4. Berlin has very high levels of Anglophony, way better than Munich and Hamburg, maybe better than Frankfurt. This means that pan-European industries have an easier time locating here, hence the tech startups that pretend they’re in San Francisco. For the same reason, finance and corporate branches relocating from London after Brexit are converging on Amsterdam as their new location, and not on Paris or Frankfurt.

          • michaelrjames

            Alon: “Berlin has very high levels of Anglophony … For the same reason, finance and corporate branches relocating from London after Brexit are converging on Amsterdam as their new location, and not on Paris or Frankfurt.”

            The logic is good but the details, not quite:

            Finance firms find new capitals amid mini-exodus from London
            Hans van Leeuwen, 30 Jan 2021.

            … of 413 London-based financial services firms which moved operations to the EU last year, 128 went to Dublin, 88 to Paris, 84 to Luxembourg, 56 to Frankfurt, 47 to Amsterdam and 10 to Brussels.
            “This first wave of relocations is very heavily sector-oriented. I can imagine multiple financial centres in the EU thriving as sector specialists. I can’t imagine them rivalling generalist London,” he said. The partial exception was Paris, which has something of a specialisation in insurance and asset management, but was hosting a wider range of financial services players. Firms were not moving their EU headquarters there, but were opening secondary offices because it was a top choice for staff leaving London.
            “It will be interesting to see if Paris continues to thrive and grow as a more generalist financial centre,” Mr Wright said.

            Dublin kind of proves the Anglophony point though doubtless other issues played a role, tax for one. But I’d say Paris is reaping the benefit of an easy 2h12m Paris to London Eurostar, as well as language not being a realworld problem for these kinds of workers, most of whom would be quite familiar and comfortable with Paris. And actually, despite the francophilia trope, more comfortable than Frankfurt or Berlin. In fact, I recall I predicted exactly this on your blog back in pre-Brexit days.

            It also happens that this effect is germane to your article, in that Brexit is driving these fintech functions and jobs to many centres across the EU, which is good compared to being concentrated in a single place. Of course with my biases it wouldn’t trouble me if The City went all Detroit but nah that’s not going to happen either.

            [And yeah, your points are all good and I’m just picking at one minor nit.]

          • Herbert

            In some ways reunification was as much a hit to Berlin as it was a boon. Unlike Paris or London or even Madrid or Rome, Berlin doesn’t automatically get federal money shoveled in for no particular reason other than prestige. But during partition, the East would concentrate its resources in east Berlin (the only subway built by east Germany was an U5 extension) and despite West Berlin not ever being technically 100% part of west Germany (Berlin only sent non voting delegates to the Bundestag) there were a lot of west Berlin specific subsidies, from air travel to “Berlin made” industry. With reunification that suddenly dried up and instead of waxing poetical about the “city on the frontlines” Bavarian or Rhineland conservatives would now decry the moral decrepitude of the “atheist” or “lazy” or “drug addicted” cosmopolitan capital. And then the CDU caused a banking crash that emptied the municipal budget… You haven’t lived thru the Wowereit/Sarrazin years, but austerity in Berlin during that era was as severe as it was dumb. Berlin’s civil service is still licking its wounds from that hit

          • df1982

            Re: Alon’s comment, there’s so much wrong to this it’s hard to know where to begin. Even West Berlin was markedly poorer than the main cities of West Germany. It had no industry and virtually no commerce, and its economy was being propped up by the FRG essentially for propaganda purposes (as Herbert outlines).

            And Berlin is nowhere near as anglophone as Frankfurt (and probably not Munich or Hamburg either). There are some anglophone ghettos in Neukölln and Prenzlauer Berg, but there are also huge swathes of the outer east where you’d be hard pressed to find anyone who speaks more than a basic level of English. You can’t say the same about any of the major West German cities.

            Also none of this is any reason why the policy of regenerating Berlin through infrastructure investment couldn’t be replicated in rust belt areas. Detroit, for instance, actually has pretty good bones: it has an expansive freight rail system that could be re-purposed as commuter rail and wide boulevards that can have transport lanes (BRT or streetcar) inserted into them. And the depopulated inner areas have a lot less NIMBY resistance to being transformed into medium-density neighbourhoods than comparable low-density suburbs in more prosperous metro areas do. Indeed, there’s already an artistic/cultural gentrification vanguard moving to Detroit much as they did to Berlin and NY in the 70s and 80s.

          • Alon Levy

            When you say things like “Berlin is probably not as Anglophone as Munich,” are you saying this out of knowledge? Because you don’t live in Germany, and your takes on Germany look extremely filtered through a few layers of badness (e.g. you said a few months ago that Kurdistan protests are illegal here; they’re not and I see them at Alexanderplatz all the time, with Kurdistan flags). The Neukölln-as-a-bubble bit is something that looks taken directly out of CityLab; Germans stereotype Neukölln as a poor neighborhood full of Middle Easterners and criminals (the average Aryan thinks these are the same), as in the depiction in 4 Blocks, or Bild’s demagoguing about the corona outbreak there last summer. There’s also a parallel stereotype of Neukölln as a gentrifying neighborhood, but in the German discourse it’s a secondary stereotype, it’s only in English that people act like Neukölln is the new Charlottenburg rather than the new Seine-Saint-Denis.

            And no, Marzahn is not somehow real Berlin. Those supposed vast swaths are a significant but small minority of the city’s population, maybe a quarter once you strip the inner parts of Lichtenberg and Pankow and the neighborhoods inside the Ring. We’re not Stockholm or anything like that but we’re not Paris either and I say this having lived in all three cities and both noticing how Anglophone people are and what other people say about how Anglophone people are (and “Berlin > Munich, Hamburg” is an unrecognizable take here, and yes, I’ve asked repeatedly).

          • Reedman Bassoon

            Let’s say you are a person choosing where to live. If you live in Detroit and work in the suburbs, you will pay a 1.2% Detroit City Income Tax. If you move to Livonia, you won’t pay any city income tax. If you are choosing where to put your business, Detroit has a 2% corporate tax, Livonia has none. If you are going to take a job at a supermarket, taking the job at a Detroit supermarket will cost you 1.2% versus taking the same job at the Livonia supermarket (P.S. living and working in Detroit will cost you 2.4%)(P.P.S Detroit isn’t the only city in Michigan with a city income tax, but it is the highest of them.).

          • Nilo

            Detroit’s freight rail system bar like 1 maybe 2 lines is terrible for commuter rail. It’s both significantly sparser than older cities (cf. Chicago and Cleveland), but also has no one living around it because the auto industry smartly located all their factories around it. Detroit’s development in Alon parlance follows roads, not rails. It is not Chicago. There is certainly a way to fix Detroit with some mass transit investments, tax reform, and highway removal, but it’s going to be slow, and those investments will only pay dividends when some other industry starts generating growth in Detroit again.

          • Herbert

            The northernmost part of Neukölln (nicknamed “Kreuzkölln“ due to the proximity of Xberg) is indeed gentryfying and Berliners by and large know that. But the reputation of Neukölln as a whole is still that put in the media by the former district mayors Buschkowsky and Giffey who despite being SPD like to present their tenure as “law& order” and “tough on crime” and stuff.

            Now as for how Anglophone Munich is… That depends to some extent on what you define as “Munich”. Munich city boundaries are relatively narrow which means both the tram and the subway cross the city boundaries (but so do Hamburg subway and Nuremberg subway [to Fürth which formes a contiguous urban era and was almost annexed in the 1920s]). Munich has reinvented itself after the war as an automotive (BMW) and research hub, but especially the research stuff is often outside Munich proper in places like Garching, Oberschleißheim or (in aviation) Oberpfaffenhofen, the airport or even the military airfield at Manching. So if you go to e.g. Hasenbergl, you’ll have different rates of English speakers than in Garching or the “Schickeria”

          • df1982

            Re: Alon, thanks for the assumption (and the attendant straw man arguments) but I actually do live in Berlin, can speak German, and have direct knowledge of the city. I also have lived in Frankfurt and commute there frequently for work (or at least did so pre-Corona), and I can definitively say it is more anglophone than Berlin. I said “probably” for Munich and Hamburg because I am less familiar with them and don’t have stats at hand. I didn’t claim that the outer eastern districts are “real Berlin” (whatever that’s supposed to mean), I said that there are large parts of the former East Berlin where the level of English-language ability is very low, and this is not the case in the major western cities. This is pretty indisputable.

            I also said there are anglophone ghettos IN Neukölln (to be more specific, certain areas bordering Kreuzberg), not that the whole Bezirk is like that. Again this is pretty indisputable, but the Bezirk as a whole is diverse, since it stretches all the way to the border with Brandenburg. Nobody thinks it is the new Charlottenburg, God knows where you got that idea from (Charlottenburg is a distinctly musty, conservative middle class area). For Germans it’s understood a case of Kreuzberg gentrification expanding outwards, for Americans it’s probably seen as a twin of Bushwick.

            Finally, I didn’t say that pro-Kurdistan protests are illegal in Germany, but that publicly displaying symbols from the PKK is, since it is a proscribed group in Germany. Again, this is indisputable and amply documented (see this TAZ article if you’re interested:!5629632/). It’s obviously not universally enforced, rather it is deliberately designed to be something the cops at their discretion can use if they want to target specific people. But the law is on the books and there are multiple recent cases of activists being charged under the statute.

          • df1982

            @Nilo, the freight lines would still be useful as regional lines to places like Pontiac and Ann Arbor, and further afield Flint, Lansing and even Toledo. There are some stillborn plans about doing this. Also plenty of the brownfields areas are ripe for TOD conversion, since a lot of its is now defunct. But I agree that the main roads have more potential for mass transit within Detroit proper.

          • Alon Levy

            [I deleted my own comment and am reposting here to fix the threading, nothing nefarious is happening.]


            Kreuzberg is undergoing white flight as well (as is Prenzlauer Berg), but yes, the Hermannplatz area is falsely believed locally to be gentrifying, whereas Neukölln writ large is correctly believed to be a racially diverse neighborhood and incorrectly believed to be dangerous (how I wish 4 Blocks were about Amara…).

            Something I recently read is that Munich, too, has net emigration of German citizens (link) but is gaining more than enough foreigners to offset. Some people just don’t like living next to foreigners. I suspect this is where the “people left behind in rural Saxony become fascist” issue comes from – they were always racist, and that’s why they’re not going to move to urban job opportunities, whereas more open people from those same areas move away and the ones who come to Berlin tell me how happy they are they got out.


            [For some reason I thought you lived in Australia, like Michael James? Sorry…]

            Yeah, so a specific symbol is proscribed, which is about as connected with the “Germany suppresses free speech” line as the prohibition on specific Nazi symbols is with general censorship (which, were it real, would close FAZ, Die Welt, Bild, and other racist papers).

            The white flight in Neukölln isn’t just the Bezirk level. A lot of the links are rotting due to ongoing site migration, but here are 2019 numbers and here are 2018 ones. Neukölln the Ortsteil is losing Germans without migration background and gaining people with migration background; ditto other supposedly-gentrifying areas like Kreuzberg and P-berg. The highest-migration-background-% Ortsteilen are Gesundbrunnen, Tiergarten, Neukölln, Wedding, and Kreuzberg, I think? My impression judging by names on buildings is that immigrants of different social classes live in the same buildings – I don’t remember ever seeing a building with lots of Turkish and Arab names and no other names or no other non-German names (and yes, I look), and all-gentrifier buildings, like the one I currently live in, look rare. (And I don’t remember obvious all-German buildings in Neukölln either.)

            The Berlin vs. Munich Anglophony is something I’ve heard from multiple people who’ve lived in both cities.


            Re Anglophony and work, I saw this FT article about Amsterdam while going on a Twitter rabbit hole looking for that US state crayon map that I linked.

          • michaelrjames

            @Aloon: “Re Anglophony and work, I saw this FT article about Amsterdam… ”

            Yeah, that was about one particular class of stock trading, one that has very low returns because it is largely automated and relies on scale (see below). It fits in with the point made by my AFR article on specialisation of different cities in the EU. You extrapolated from it too much. (The AFR–Australian Financial Review–is the Oz version of FT; in fact that article may have been reprinted from the FT, which they do quite a bit on European matters.)

            London’s status as a financial centre isn’t as secure as some might think
            The centre of Europe’s share trading has shifted to Amsterdam – this is a problem for London that is only going to grow
            Dan Davies, 13 Feb 2021.

            In itself, as William Wright from the New Financial thinktank says, the fact that Amsterdam has overtaken London in terms of equity trading volume is unlikely to make much difference to anyone’s profit and loss account (or a country’s tax-take). Measuring trading volume is in any case something of an abstract concept, which these days depends more on the way in which computer algorithms split up orders to get the best price, than the actual amount of business. The commission on equity trading is incredibly low and the exchange fees even lower.
            But there is a reason why stock exchanges compete so aggressively to be number one in market share. Financial market trading is the ultimate case of increasing returns to scale – the more business you do, the easier it is for someone to find a buyer or seller, and so the more attractive a place you are to do business. And the same dynamic works in reverse. When you get a negative shock to your market share, it also tends to be self-reinforcing. London also lost huge market share in January in euro-denominated swaps trading.
            The immediate cause of all this is that the UK failed to negotiate an “equivalence” deal with Europe for its trading venues after Brexit. For the purposes of EU regulation, a number of British trading platforms ceased to exist at the end of 2020.

          • df1982

            That’s OK, I’ve led a peripatetic life, but in general it’s best to avoid dismissing someone’s views on the basis of where you assume they live. I would be careful with the “people with migration background” statistic, since it can cover two very different groups: people from high-income countries moving to Berlin for professional/lifestyle reasons, and those from lower-income regions (Africa, Turkey, the Middle East, predominantly) migrating out of existential necessity. Prenzlauer Berg is definitely the former (there are a lot of non-Germans, but it’s still overwhelmingly white), while Neukölln was traditionally the latter, but the “Kreuzkölln” areas are becoming more and more the former.

          • michaelrjames

            Hah, I inadvertently turned you into a Dutchman with Aloon! Or is it A Loon?

          • Herbert

            If someone who studied at LMU or TU Munich moves to Garching to have a job in research after graduating, that counts as “having moved out of Munich”. The combination of world class r&d in Munich suburbs (including a living Nobel laureate working at Garching) and the municipal boundary not including them leads to some of those oddities

        • Herbert

          If your differentiating between “good” and “productive” industrial capital and “bad” or “exploitative” financial capital, you’re already halfway on the path to antisemitism.

      • Nathanael

        From what I can tell the money isn’t even going to proper Rust Belt cities like Cleveland or Toledo or Rochester, it’s going to the rural Midwest or Appalachia. (Appalachia was never Rust Belt.)

  4. Sarapen

    But telling politicians to “Just Say No to Leakage” is a technical solution to a political problem. There is little political incentive to say no and every reason to say yes. I can’t speak to the other examples but for Ontario, it’s not as if Doug Ford wants excellent public transit but is held back by suburban NIMBYs: he is in fact one of the suburban NIMBYs and therefore usually rules in their favour. The leakage is a feature, not a bug. You’re essentially saying that the problem is how scarce resources are allocated, which is fundamentally an issue of politics and not something to be solved by a city planner or such.

    • Alon Levy

      It’s not exactly resource allocation, since leakage is a negative-sum game. In the case of the Toronto RER, the electrification project will reduce noise, because electric trains are less noisy than diesel trains even when they run faster. But the benefit-cost ratio for the project was so high that instead of fighting it, Metrolinx caved early and agreed to the noise walls. So it’s perhaps better to say that surplus wasn’t extracted but destroyed.

      This is also what we’re seeing with Second Avenue Subway. The various schemes used to limit neighborhood impact at high cost didn’t actually reduce negative impact. The decision to mine the stations instead of building them cut-and-cover meant less neighborhood impact but over a longer period of time, about 5 years instead of 1.5-2. Value was destroyed, but the mere perception of reduced impact led to the more expensive choice.

      So in that sense, I think it’s valuable to let agencies know that this dynamic exists and they should be careful to avoid it. In the case of very high ROI projects like late electrification or Northeast Corridor HSR, the best choice is probably to intentionally overexpand to a lower ROI, and yet this has a lot of potential for leakage, i.e. for including extra projects that aren’t actually good (like the 1.5-orders-of-magnitude-too-expensive SOGR program in the US) rather than things that are justifiable on their own.

      • Herbert

        I think the French combination of tram construction and city beautification is a good idea overall. You’re changing your city already, so why not use the opportunity to do several things at once?

        • Alon Levy

          Yes, and when it’s something relatively small, it’s fine. The problem is when you have a city – let’s call it “San Francisco” – building bike paths for the cost of a subway, excusing everything on “we need to rebuild the street” grounds.

          • michaelrjames

            I guess you are referring to the ≈$300m bikeway+pedestrian path proposed for the (West) Bay Bridge but it’s really the cost of one car lane on that bridge. It would create a kind of symmetry–the world’s most expensive bikepath to accompany the most expensive bridge per mile (east span). In fact they are overengineering the bikepath so as to be able to take a maintenance vehicle on the premise that it will prevent closing a traffic lane when maintenance access is required. Again, it’s all about prioritising cars. If one was paranoid one might imagine a near-future conversion of that ‘bikepath’ to permanent vehicle use–or maybe one bus lane instead of taking two car lanes from the bridge. It has even been suggested for the, admittedly underused, San Rafael bridge bikepath. One wonders what a lightweight bike/ped path (not suitable for any vehicle) would cost by comparison?

            And yes this comparison (below) hasn’t been updated to include that cost (which in any case hasn’t been built) but I reckon the cost comparisons are worthwhile posting anyway. I’ve converted all the figures to billions which IMO makes the comparison easier for the eye.


            1 mile of Bay Bridge eastern span: $2,068,343,000……………….$2.07
            1 year of the SFMTA budget: ~$1,371,100,000……………………..$1.37
            1 mile of Central Subway: $1,000,000,000……………………………$1.00
            1 mile of Doyle Drive: $571,066,667…………………………………….$0.571
            1 mile of basic protected bikeway: $445,000………………..$0.00045
            1 new traffic signal: $280,000……………………………………………..$0.00028
            1 curb extension: $78,000…………………………………………………..$0.000078
            Percentage of MTA budget spent on biking: Less than 1%.

          • michaelrjames

            @Eric2: “He’s probably referring to this $300M/mile project for Market Street”

            Maybe. Though the $300m West Bay Bridge bike + pedestrian path is about 3 miles, and is a new bit of bridge hanging off the existing bridge (cannot be slung under due to shipping clearance requirements), and as I said the bridge supervisors want it to take full size vehicles (and probably very heavy maintenance vehicles bloated in the usual American tradition) so it is not quite as outrageous as first seems. That is, it’s not just laying down some green paint though that is probably what they should do (on an existing car lane):

            Petition for a Quick-Build Bike Lane on the Bay Bridge
            The Brooklyn Bridge already has a bike and ped path–and NYC intends to add more space for both. Advocates want the Bay Area to do the same on the Oakland Bay Bridge
            By Roger Rudick, Feb 1, 2021

            Editorial: Motorists, Be Thankful about Cost of Bay Bridge Bike Path
            The high cost isn’t really about bikes, it’s about giving almost everything to cars
            By Roger Rudick, Nov 20, 2018

          • Matthew Hutton

            Alon I have gone and pointed out your style of analysis to my local representatives in the UK on bike lanes and they realise the value and are impressed. So I think change can happen.

            I mean moving from a cycle lane costing £250k/km to £50k/km obviously makes it much easier to justify construction.

  5. adirondacker12800

    maintenance spending funded out of other people’s money
    Martians don’t pay Federal or state taxes. It’s not other people’s money. It’s just coming out of a different budget. Why shouldn’t people who pay a lot of taxes expect that some of it get spent on them?

    declining regions like the Rust Belt, where due to continued road investment and population decline, there is no longer traffic congestion.
    The Northeast and Midwest have toll roads which get little Federal money. Little state money either. If New Jersey Turnpike toll payers want to spend toll money adding lanes they can. Extend that logic, the Second Avenue Subway shouldn’t be built and a second set of Hudson River tunnels is silly.

    • Alon Levy

      New York hasn’t really been called Rust Belt in 20+ years. When I talk about the Rust Belt, I talk about new interchanges in small towns in Ohio and Michigan, and not about very busy Northeastern or Chicago roads.

      And other people’s money is really about outside infusion. Massachusetts is of course a net tax donor, and within Massachusetts Somerville is a net tax donor, and yet the Community Path was not funded locally and therefore Somerville’s civic elites didn’t mind that it was horrendously expensive.

      • Herbert

        In Franconia you can sometimes sell expensive projects with lots of state funding at least in part on “so that the money doesn’t end up in Munich”

        If there’s one thing most Franconians agree on, it’s that they dislike Munich and see the Bavarian government as sucking money out of Franconia and putting it into Munich and adjacent areas.

        Just look at the number of top research facilities that Munich and surroundings have acquired in the last seventy years. .

      • adirondacker12800

        The Rust Belt is the Northeast and Midwest. The mill now being The Lofts at the Mill luxury condos doesn’t make it less deindustrialized.

        The DOT or the toll roads either, don’t do things for the fun of it. There has to be some sort of reason for a project. You may not like the reason but there is a reason. A lot of it is because there is this pesky thing called time. When it passes things wear out. Then need to be replaced. Then there are those pesky pesky users who do things like move to suburbs. And then need more capacity. And very very rarely that additional capacity is things like new railroad tunnels under the Hudson River.

        • Henry Miller

          The reason the DOT does things because there is money. Often there are several different buckets of money, so one bucket really is starved for money while another has plenty for useless projects. Thus you get state DOTs overbuilding highways to nowhere, while in the same area county bridges are failing inspection but there is no money to replace them – both budgets are filled by the same gas tax, but the state DOT is getting a larger share of the money and the counties are not able to access any of it for work they need to do. Thus states will raise gas taxes because of a real problem in back county roads, and the DOT will announce another highway to nowhere funding by those taxes, with the counties will just barely have enough to repair a few of the worst bridges.

          • adirondacker12800

            If they didn’t have any money they wouldn’t be able to do anything. The employees would wander off quite quickly too. Which is it? “several different buckets of money” or “budgets are filled by the same gas tax”. It’s a pity the gas tax doesn’t fill the budget of of the Federal or state DOTs. Little if any of it filters down to county or municipal budgets. It’s why there are local roads maintained by the municipality, county roads maintained by the county, state highways maintained by the state and Federal highways maintained by the state with Federal contributions.

          • Henry Miller

            Several different buckets of money all filled by the same tax. thus one bucket has more than it objectively needs and instead of giving that money to a different bucket they find more projects for that bucket. In the mean time the other bucket doesn’t have enough money for the projects they need to do. In the case of my example, the State DOT only funds state roads (an interesting quirk, national roads in the US are state roads with federal funding), and not county roads: thus the county isn’t getting enough money to repair their small roads while the state is building large projects that are not needed.

            Each state does things differently, of course. Still the above seems to happen a lot in some variations. Keep in mind though that I’m not making any claim that the counties are doing a good job of controlling waste – I have no insight if they are getting enough money to repair their roads but spending it on expensive projects.

          • adirondacker12800

            gas tax doesn’t fill the budget of of the Federal or state DOTs. Toll roads don’t get any tax money.
            None of them do anything because they think it will be fun. somebody somewhere took measurements, somebody else evaluated them and wrote down all the reasons the project should be done. Then there was a variety of reviews. It’s too bad you don’t like their reasons but there are reasons.

  6. seangillis78

    I disagree that State of Good Repair is leakage, as you’ve defined leakage. Some of it might be incredibely ineffective and wasteful, but I hope you are not arguing that state of good repair and general maintenance is not important. Scrimping on maintenance is going to lead to less reliability and generally shabiness, and in worst cases compromised safety.

    What would be more useful is looking at best practice for ‘standard’ life-cycle spending on various components, especially the nuts and bolts of the working systems: tracks, road beds, switches, wiring and power systems, garages, etc.

    I would agree that some renovations are less useful then others. Prettying up the subway stops might be easy to sell, but not a good use of cash. Hard to say though.

    • ericson2314

      Based on prior posts, I a pretty sure the argument is on what the specific maintenance is, not maintainence in general.

      I see a lot of the same stuff in the tech industry, where people write shit code, and then dump money on “site reliability engineers” and not fixing the underlying problems once and for all. So yes there is are specific patronage / regulatory capture problems in the the public sector, but I also think there is a tendency for productivity to sag in loose labor markets.

    • Nathanael

      NYC Subway in particular was in a practically derelict state by the 1970s; “state of good repair” means replacing everything but the tunnels. Now, is NYC managing to do do this in a particularly bad way — replacing parts with archaic out-of-date parts rather than modernizing while they replace — spending too much to get too little — well, yes. But there was no getting around the rebuilding of most of it. Just the damaged walls and water infiltration alone were huge, and expensive-to-fix, problems. And they’re not “sexy” problems with good-looking ribbon-cuttings, unfortunately — water problems never are. (Speaking of water problems, Daylight Tibbetts Creek!)

      • Alon Levy

        New York did that in the 1980s-90s. Since then, SOGR has been a black hole of ever-escalating flagging rules and no decrease in the backlog.

  7. Pingback: The Leakage Problem | Later On
  8. Herbert

    The Wiesbaden tram vote is a perfect example of suburban vs urban voters. The urban core voted yes, outlying districts voted no, overall the project was rejected 2:1

  9. Aaron Moser

    “the Gilets Jaunes riots were mostly white supremacist, but they were white supremacists with a strong anti-urban identity who felt like the diesel taxes were too urban-focused.”

    I guess this blog is turning into a Twitter thread now. Hahahaha

      • Oreg

        Most likely he’s referring to the “white supremacist” part – but you knew that.

        On a similar note, in an earlier comment you labeled the Frankfurter Allgemeine Zeitung (FAZ) a “racist paper”. Would you elaborate?

        • Alon Levy

          It published an op-ed by an elder SPD politician who sits on the Berlin Holocaust memorial’s board who ranted about how ethnic minorities have a responsibility to accept majority rule and drew equivalence between AfD and immigration activism. And it’s impossible to read anything about immigration there without being bombarded with a litany of statistics about social problems.

  10. Pingback: How to Build High-Speed Rail with Money the United States Has | Pedestrian Observations
  11. Pingback: Cost- and Project-Centric Plans | Pedestrian Observations
  12. Pingback: Institutional Issues: Coordination | Pedestrian Observations

Leave a Reply to Nilo Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.