Sorry for the lack of posts and updates to my econ blogs ranking in the past week, I was away at the EARIE conference for this year.

At the moment I live in Japan and my options for Internet access are many. At home I subscribe to a VDSL service that is basically provided by NTT but resold by another company. For around 4,000 yen per month I can buy a nominal line speed of 100Mbps, with no restrictions on usage, except that I can’t run a web server or things like that. My connection uses the regular phone line into my house, although I don’t actually subscribe to a phone service. As well as VDSL, there are ADSL services available from NTT and other providers, plus there are two or three cable TV companies that can provide Internet over their networks, and there are three 3G mobile networks that will provide (somewhat slower) access at reasonable cost. In short, there’s plenty of competition based on multiple physical networks. This is because the density of population makes it viable to build multiple access networks that can compete. In addition I think there is regulation that allows ADSL providers to piggy-back on NTT’s phone network into each house.

In contrast, I came from New Zealand where the population is low and not very dense. Cable TV basically doesn’t exist, and almost all houses have their phone line provided by Telecom, the New Zealand equivalent of NTT. It is possible to get an Internet connection using a 3G mobile network in many places, and there are some other providers for businesses in cities. But the market size and geography simply won’t support the level of infrastructure-based competition that exists in Japan.

One of the interesting challenges that New Zealand faces in this situation is promoting investment in new telecommunications technologies. The fastest available residential connections are basically ADSL running at speeds around 8Mbps or less. Next-generation services like VDSL at 100Mbps, as I have in Japan, require significant additional investment. Since it’s only viable to have a single fixed-line access network, the problem is how to get this investment without ending up with high monopoly prices.

In the past, New Zealand’s regulatory approach was one of ‘light-handed regulation’ in telecommunications. The basic idea was to leave the market be, and if it wasn’t working well, step in with regulated controls later. However, such an approach creates a problem for investment in new technologies. The problem is that such investments are risky at the outset, and it’s not exactly known whether they will be unprofitable, moderately profitable or very profitable. Furthermore, after the investment has taken place, it’s hard to tell whether high profits are simply a risky investment that paid off well, or the result of monopoly power.

In this scenario, if the regulator errs on the side of finding market power, he will tend to regulate prices down and reduce profits in a high-profit outcome. From the investor’s point of view, this reduces his returns in the case that his investment pays off nicely, and makes him less willing to undertake the investment. Put another way, this type of regulatory regime itself increases the risk of the investment, making it less likely. In a worst-case scenario the investment won’t be undertaken at all. That’s bad because even a service provided at a monopoly price is better than not having the service at all.

An alternative approach is to recognise the reality of the situation: The market won’t support extensive competition with multiple networks. Therefore, the need for regulation could be acknowledged at the outset, rather than leaving regulation as a backstop. This doesn’t mean that prices should necessarily be fixed by the government in advance. Rather, the terms of regulation could be specified in advance, to reduce the regulatory uncertainty of the investment. Such potential regulation should also acknowledge the risk involved with the investment at the beginning. After an investment has turned out to be successful, it is easy to underestimate the actual risk that was taken, but higher risk deserves a higher return if it pays off.

To me, it seems that this issue needs further investigation. In a dynamic situation of technological change, and in a small market where infrastructure-based competition is not viable, what is the optimal regulatory regime? Let the market work first and regulate later if necessary? Or specify some kind of regulatory position first? It would be interesting to see the exact factors that drive the choice of regime.

Web 2.0 … It’s last years buzzword for websites that involve some sort of user-generated content (blogs, flickr) or web-based software (Google Docs) or social networking (Facebook) etc. I found the “complete” web 2.0 directory at Go2Web20.net which lists many interesting sites that I’d never heard of. These sites are interesting not because of what they do, but because they haven’t been successful. In fact Go2Web20.net’s directory is more like a graveyard of Web 2.0.

For example, PeopleTrusted.com, “a democracy of trust”. This actually seems like sort of a good idea. It’s a site where people can vote for online retailers and services that they trust. As the site says, “PeopleTrusted helps bring credibility and social reputation to products, services, and companies doing business on the web.” Asymmetric information is a big problem in online markets — it’s easy to set up a website and rip people off. So in theory PeopleTrusted could solve that problem and provide a valuable service. I’ve even written this paper and this paper about exactly that idea. But PeopleTrusted seems tobe having trouble getting off the ground. It has very few members, and very few products have been voted on, so it’s not really a very useful site. I also wonder how prone it is to manipulation? I didn’t check the site carefully enough, but how do they stop people registering fake IDs and polluting their data?

Next up, Yample. I have no idea how they came up with the name ‘Yample’, but I know it’s hard to find a good domain name these days (26econ … ???). Anyway, Yample is “listings gone social”. It’s listings, ie classified ads like Craigslist, only with a “social” aspect where people can create networks of friends, tags, and vote on favourite listings etc. Unfortunately, Yample seems more antisocial than social. They have less than 100 users, most of which seem to have used the site only once.

I might be boring my readers, so just one more, Guruza (I couldn’t steal a copy of their logo easily). It’s yet another “ask an expert” type site, where you can pose a question and experts can answer them (for a fee maybe). This idea has been around for a long time and there are many other such sites, so I don’t know what’s so “Web 2.0″ about it. Anyway, Guruza seems to be suffering from a severe shortage of both questions and experts.

So why have I tortured you with these boring details about unsuccessful websites? I wanted to illustrate the fact that many of these “Web 2.0″ ideas depend on some kind of social or network aspect. All of the three sites above are not bad ideas per se, but need lots of users and usage to become useful resources. The trouble in such cases is how to get off the ground in the first place. When the value of your business depends on how many people use it, people will be reluctant to use it unless they expect others to do so. Even if the site is free to use, it still takes time and effort to register and so on.

In these kind of network markets, probably only a few businesses will survive. There’ll be a few big winners like Facebook and Digg that manage to get off the ground, achieve a critical mass of users, and can then make lots of money. How to maximise your chances of becoming such a winner? First, expectations are critical. If people have pessimistic expectations about the success of your site, they’ll stay away, and their bad expectations will be realised. If you can make people optimistic about the usage of your site, through advertising, promotions or whatever, then you’re more likely to get off the ground.

Second, if possible, leverage off an existing small but well-connected community. Facebook was initially restricted to college students. This was a perfect way to get off the ground, since students already have strong social networks in the real world, and Facebook provided the technology for them to take those networks online. If Facebook had just set up a website and made it open to all, I think it would have been harder to get off the ground.

Third, piggy-back on an existing network if you can. Can you make your service ‘compatible’ with an existing network so that it’s easy for people to switch? This is another thing that Facebook have done right — they make it very easy for you to import your existing networks of contacts from email services like hotmail and gmail. This greatly reduces the pain for users to join Facebook’s network, as they can re-use data that they’ve already plugged in to another network.

In all, starting a network-based business, like a Web 2.0 site that’s based on social networking or user generated content isn’t easy, and I’m not optimistic about the chances of success of many sites in this category.

Prediction markets aren’t the only way to predict things. Probably the most popular way (aside from crystal ball gazing and guessing) is to use mathematical relationships that appear to hold true over time, and project them into the future or apply them to new data values. Author and Yale professor Ian Ayres has put together a list of online tools to do this kind of kind of forecasting on his site. Examples include predicting how long you’ll live, predicting your child’s height as an adult, etc.

Ian also has a book out on the topic of how businesses and other institutions are using data analysis to change the way they work. I haven’t read it yet, but it looks interesting so when I get a chance I’ll read it and post a review here.