My Articles

<-My Homepage

Internet Search Advertising (written in mid 2006)

Searching for Competition

by Chris Murray 

"To become aware of the possibility of the search is to be onto something." Walker Percy

Is Internet search-based advertising the perfect business? It has certainly been a successful one, minting hundreds of new millionaires in barely a decade without charging its consumers a dime. And the service offered to consumers is of ever-improving quality, while the advertisements that generate revenue have become less obtrusive.

In its short existence, internet search-based advertising has become a market of cutthroat competition (who remembers Lycos?). It sucks up billions in capital spending and offers high salaries and an employment haven for many of the world’s uber-geeks. But it remains an immature market, and search leader Google maintains outsized profit margins and profit growth (recall that their recently-reported second quarter profit doubled) along with an ever-increasing market share.

Perhaps the greatest imperfection and long-term challenge for this market lies in its seemingly inexorable drift toward monopoly. Reasonable people may ponder (and perhaps worry) whether the fundamental structure of this market naturally leads in that direction. And these same reasonable minds might also need to formulate policies to render this unwelcome scenario less likely.

At first glance, two distinguishing features of the search-based advertising market should, in theory, make it highly competitive. One is the vanishingly small cost a consumer faces when changing search engines: it’s as easy as resetting a home page. The next is the low marginal cost faced by the search engines themselves of serving additional customers. That is, it costs Yahoo virtually nothing when a customer switches to them from another search engine. And the cost of entry for new players, though high, is certainly no barrier for today’s giants of technology. And yet, the market is dominated by only a few firms, with the top three providing over 80% of the search queries and Google closing in on 50%.

The reason for this seeming contradiction can be summed up in one word: price. Or maybe we should say it can be summed up in three: lack of price! As any Econ 101 student knows, price is the fundamental determinant of economic success, because it matches supply and demand, and because it reins in super-sized profits in healthy markets. In a normal market, price and quality form some kind of tradeoff— the consumer can make a rational choice to accept a bit less quality for a slightly lower price. But search engine users don’t pay a price, so a rival search engine cannot undercut a rival’s price by offering slightly inferior service to customers willing (perhaps eager) to make that tradeoff.

The simple structural feature that prices cannot be less than zero turns out to be a severe anti-competitive constraint in the search-based advertising market. Eliminating this constraint would be a driver of competition and a boon for consumers (economically astute readers note that the former almost inevitably implies the latter). And here’s the thing: This market shift cannot come about by government intervention; only demanding consumers and good old entrepreneurial innovation can do the trick.

If a system could be developed where regular users of a search engine are paid to use it, the zero-floor on price would be eliminated and search engines could compete with each other on the basis of price, which is fundamental in competitive markets. That is, some search engines could employ cheaper technology but share a larger percentage of their revenues with users, while others could invest more in their product and share less with users. This would both shrink the current mammoth profit margins of today’s market leaders and increase competition by allowing market participants to compete based on price.

It may sound ridiculous at first, but the main barriers are psychological. The solution to the technical problem of cheaply making frequent secure payments to customers already exists: Ebay’s Paypal® and Google’s often-rumored Google Wallet ® could work. Easily distributable toolbars, much like Google and Yahoo’s current offerings, could track how often customers use their sites and credit them accordingly. Certainly some scam artists would write software programs to impersonate search-engine users and fraudulently collect revenue, but well written policies and machine-learning based fraud detection techniques could largely prevent this. Indeed, today’s cyber-fraudsters already use all kind of techniques to scam the current pay-per-click advertising schemes that provide the bulk of search engine revenue, and search engines are still prospering. But, again, the psychological barriers could be the hardest to overcome. Scam offerings that claim to pay users large sums of money to view advertisements (like this one) naturally make users skeptical, but if a strong and reliable search engine offered a paid-to-search service promising reasonable payment, most consumers would likely find it credible. Certainly it would take customers a while to get used to the idea, but we’ve all grown accustomed to far stranger things in recent years. In time, the expectation of getting paid to use a search engine could easily become commonplace, just as we are now accustomed to having television content subsidized by advertisements.

But perhaps, as customers, we may be content with the status quo. If seems like we’re getting a pretty good deal now: the titans of tech are spending billions a year to give us free and ever-improving service. But remember one of the basic principles of economics laid out by economics professor Gregory Mankiw: rational people think at the margin. As consumers, our attention is valuable. Currently, all of the advertiser-value created when we view ads in a search engine is taken by the search engine itself. The benefit for us, which is the use of the search engine, is tiny on the margin. That is, we can easily switch to another search engine and get results that are 99.9% as good. Rational people should happily use a second-best search engine that’s 99.9% as good as the first-best one in order to get paid for it. And if that second-best search engine fails to find something, the best one is still free.

How would the search-based advertising market look if customers were paid for their searches? The current market leaders could ride their high horse and refuse to pay customers, allowing smaller almost-as-good upstarts to offer to pay customers and thus steal the leaders’ market share. Or market leaders could agree to payout to customers (though perhaps a smaller percentage) and maintain higher market share but with smaller profit margins. Either way, customers are the winners. "When the trade or practice becomes thoroughly established and well known, the competition reduces them to the level of other trades" –Adam Smith, The Wealth of Nations

Here’s to competition.



Oil Demand Elasticity (also 2006)

Burning cheap oil requires flexibility

By Chris Murray 

High oil prices seem to be the boogeyman on every economist’s mind this year, especially now as chaos erupts in the middle east.  The stratospheric rise in gasoline prices, in particular, is shaving GDP growth, fueling a growing trade deficit, and keeping consumer sentiment in the dumps.  Politicians, predictably, are generally doling out blame along party lines, and a host of reasons have been suggested to explain this bump in oil prices.  Popular scapegoats are the Chinese (first they took our manufacturing jobs, now they’re taking our oil?), the big oil companies (does American’s pain at the pump subsidize Lee Raymond’s stock options?), the car companies (who presumably hypnotized us into buying SUVs?), hedge funds (lightly regulated investment vehicles for the super wealthy are an irresistible target), or general “instability” in various Arab, African, and/or Russia-related countries. But perhaps more foreboding that the rise in oil prices has been American, consumers’ response to it.

            Generally, when the price of a product goes up, consumers use less of that product and the quantity demanded goes down.  How much the quantity demanded falls for a given increase in price is called the elasticity of demand.  Of major interest is the elasticity of demand for oil in the U.S because it effects how prices respond to supply shocks (like broken pipelines).  That is, if there is a sudden reduction in the supply (or expected future supply) of gasoline but demand is relatively elastic, the increase in price will be moderate.  However if the same shock occurs and the elasticity of demand is small, the price must skyrocket to nudge obdurate consumers into consuming less.  So how elastic has the demand been for gasoline?  While crude oil prices are up over 600% since July 1998 and over 200% since July 2002, oil consumption is up roughly 10% since 1998 and 5% since 2002.  Meanwhile 2006 is shaping up to be a record year for oil consumption, despite all-time high nominal prices.  That’s backwards: consumption should go down as prices rise, but it’s clear from the demand curves that America’s elasticity of demand is inverted.  Even adjusting for GDP growth, oil consumption has lagged GDP growth by only about 6% total since 2002 while prices rose over 200%.

            This astoundingly low elasticity of demand has several consequences.  First, it makes prices extremely volatile in the face of supply shocks: if a 200% increase in price made no substantial dent in consumption, how high will prices have to go if a new supply shock (say in Nigeria or Iran) knocks some supply offline and forces consumption down by 5% more?  Second, if demand is inelastic, producers maximize revenue by reducing output and raising price, while intransigent consumers bite the bullet at absorb the increases.  Hardly a soul in this country wants to see more US dollars flowing to Venezuela or Saudi Arabia, but if demand remains inelastic it’s in their financial interest to lower their oil output and raise prices.

            Sadly, supply shocks won’t end anytime soon.  Venezuela’s atrophying production under an incompetent government, the insanity in Nigeria, political grandstanding in Iran, and the recent chaos in Iraq all look set to continue for some time into the future.  However, several details may make the situation less bleak than it may appear at first glance.  The first is that demand is usually far more inelastic in the long run than in the short run.  Though it’s surprising that American consumers, given at least 3 years of rapidly rising oil prices, have done such a poor job of reducing consumption, hopefully this is merely a because habits, automobiles, and lifestyles are slowly changing but the effects of those changes take a long time to appear.  Second, high crude prices make substitute products, like ethanol and electricity supplied by coal, more competitive: this should increase supply, though certainly it takes many years for additional energy production to come online.  Lastly, the price volatility induced by inelastic demand works both ways: that is, just like small supply shocks can drive prices up sharply, small additional increases in supply can push prices down quickly.  Policies that could increase supply (like allowing more oil exploration, reducing the tariff on imported ethanol, and initiating trading in long-dated option contracts on oil) can make only a tiny dent in short- or medium-term supply.  Ultimately, prices have recently been most heavily influenced by consumers (in)elasticity of demand, which has been determined by the aggregate individual decisions of consumers far more than by any policy decree.  If these consumers continue to shrug off high energy prices, future supply shocks may force prices to a level which requires immediate and painful changes and forces economic growth to a halt.  The market is sending us a signal to reduce our oil consumption far more strongly than Al Gore ever could: we could use the flexibility to heed it.