Last week's floods in Jakarta illustrate that the private sector can provide valuable public goods, available free of charge to just about anyone. If you wanted to see how hard it was to get around the city, a free web link (lewatmana.com) gave access to real-time cameras at various strategic intersections. In the most testing circumstances, the technology kept working. This was a triumph for private sector provision of beneficial public goods.

If, on the other hand, you actually wanted to physically go somewhere, you were pretty well out of luck unless you had a helicopter. The private sector cannot provide public services requiring enormous infrastructure investment where the externalities (benefits that you can't charge for) are large and broadly spread.

Just about all big-city infrastructure is like this. Certainly, there are many examples of public-private partnerships (PPPs), not least in Australia, which pioneered a successful PPP model. It was successful because state governments had run out of borrowing capacity, thus the alternative was that high-return projects would have remained unbuilt.

That said, it was an expensive way of funding infrastructure, and the private sector was usually smart enough to shift the project risk back to the taxpayers. PPPs got quite a few freeways built and, importantly, got all of us accustomed to the idea of routinely paying for what had usually been provided free of charge. It did not, however, give us an adequate public transport system: it was easier to toll motorists than to charge train and bus users the full fare.

Meanwhile, back in flooded Jakarta, PPPs have not been able to rehabilitate the city's drainage or provide a viable mass transit system. Some form of mass public transport has been on the agenda since well before the 1997-8 Asian crisis, but the only legacy is a few abandoned concrete supports.

For these big public projects, the PPP model is an irrelevant distraction. PPPs don't work unless there is a strong guaranteed cash flow, which a toll road or an electricity generator may be able to supply, but neither a sewer nor public transport can provide because only a small part of the benefits can be collected in the form of revenues. The current floods are hugely expensive in damage and disruption to production and commerce, not to mention the heartbreak and trauma to submerged households. But if this problem were to be solved with better infrastructure, you couldn't charge the customer for the benefit of flood protection in the same way you can charge a motorist for using a toll road.

Thus a purely revenue-driven PPP solution for Jakarta's flood problems is impossible. The same goes for public transport. Few cities of the world manage to run a profitable mass transit system, and none where the users of the service are so poor that they can't afford a full-cost service.

Let's not ignore the 'magic of the market'. Let's accept that there will be 'suitable cases for treatment' using PPPs. But let's also observe that cities like London are viable because, long before PPPs were invented, governments found the resources to build the Underground (celebrating its 150th birthday) and get the sewage out of the Thames.

In Sydney, the cost-benefit of the Harbour Bridge must have looked hopeless in the 1920s, when the decision was taken. When the ten-lane bridge was completed early in the 1930s, photos show it being used by a handful of T-model cars and a couple of horse-drawn carts. It took four decades to reach capacity.

But we can now be thankful that this wasn't a PPP project, which would have occupied the prime crossing site with just enough capacity for the then-current traffic. This sort of infrastructure has to be built by governments with a broad, long term vision of what constitutes a viable city.

Governments are often inept at project selection. Pork barreling distorts decisions, and without clear projections of cash flow, cost-benefit estimates can vary enormously, depending on how the externalities are valued. They run the risk of building grandiose white elephants (is the proposed Sunda Straits bridge in this category?).

So government should try hard to winnow out the poor projects. But when a capital city regularly closes down due to flooding, and lack of public transport brings city traffic to a crawl even in the dry weather, there are clearly so many high-social-value projects that the right course is to say 'just do it'.

Photo by Flickr user nSeika.