What is an entry barrier?

New entry into an industry will increase the number of competitors, which in turn will increase rivalry in subsequent time periods. It is possible to estimate the likelihood that new entry will happen by assessing the size of the barriers to entry.

This is post nine in a series on industry analysis.

Historical perspective

Somewhat surprisingly, industrial organisation economists do not have a consensus view of what constitutes a barrier to entry. Here are a few seminal contributions:

• Bain (1956) defined entry barriers as anything that permitted incumbents to earn above-normal profits without attracting new entrants. Economies of scale and product differentiation would be examples of entry barriers according to this definition. I consider this definition is weak because defining a term by its consequences is very bad practice.
• Stigler (1968) defined entry barriers as costs that must be born by new entrants, but not by incumbents. This definition would exclude economies of scale if both the incumbent and the entrant would have similar cost in building a factory. To the extent that product differentiation is mainly supported by advertising, this will constitute a Stiglerian barrier to entry. A challenger to PepsiCo and Coca-Cola would have to spend more to get the same degree of impact of its advertising.
• Demsetz (1982) pointed out that both of these definitions are problematic. If there for instance would be a licencing requirement, that would not be counted as an entry barrier in any of the above definitions. 

Not much subsequent work has been done in economics on finding a consensus view. To make the the concept of barriers to entry valuable for industry analysis, we need a better definition of barriers to entry. The real world is never as clear-cut as in an economist's model. To do practical industry analysis we also need to take cognitive biases into account as well as realise that new entry can be a long process taking maybe ten years to happen in some cases. For an economist, such transition issues are less relevant because they tend to think about either short-run (kind of within 6 months) and long-run (kind of until end of time).

Michael Porter. Porter (1980) identified seven categories of entry barriers; economies of scale, product differentiation, capital requirements, switching costs, access to distribution channels, government policy, cost advantages independent of scale. However, the last entry barrier is a catch-all category so it can be argued that his list in fact is longer. A problem with this list is that it is very inclusive, which means that there is a tendency to see strong entry barrier in most industries.


Practical assessment of barriers to entry

This is a blog post and I will not try to come up with a better definition of barriers to entry here. However, I will provide some practical guidelines to analyse the threat of new entry into an industry.

It is preferable to focus on the potential new entrants' ability to survive in the focal industry over the medium term; around 5-10 years into the future. To be able to survive 5-10 years, the new entrant must deal with all aspects of business (e.g. minimum efficient scale in manufacturing, sufficiently different product supported my marketing and/or innovation). A new entrant that survives this long will be in business long enough to cause genuine trouble for the incumbents.

First, an assessment of the following three factors should be made. Second, an assessment of the likely retaliation of the incumbents should be made. The second assessment is more important when the three factors show that barriers to entry are low.

Factor 1. Regulation. Regulation can come in many forms: The government may require a licence to produce or distribute products. There might be mandatory quality standards to adhere to. Some industries are more regulated than other industries.

Factor 2. Large irrecoverable, fixed costs. Barriers to entry are higher if a larger proportion of cost is irrecoverable, fixed costs. The variable cost will (by definition) always be matched to sales, but if a large portion of the cost is fixed and cannot be recovered in the case of exit there is a barrier to entry. We can distinguish between two theoretical situations, which both can have barriers to entry:

• The new entrants' fixed costs are not any larger than the fixed cost born by the incumbents in the past (somehow properly expressed in today's value of money). According to Stiglitz this would not constitute a barrier to entry. However, Stiglitz fails to take the adjustment process into account. The incumbent might have incurred the costs over several decades (read real options), but the new entrant will have to spend the amount during a fairly short time period. If the costs are irrecoverable, the potential new entrant is subjecting himself to higher risk.
• The new entrants' fixed costs are higher than the fixed cost born by the incumbents in the past. This could be the case if there is a lot of proprietary production knowledge (experience curve), which is hard to copy, or if there is a lot of brand advertising. A new entrant would have to spend more money on advertising to get the same impact as the incumbents have. In this situation the barriers to entry are even higher.

Practically, each major cost component should be analysed separately, for instance:

• Production cost. There might be learning effects that makes it necessary for the new entrant to enter with large scale to quickly move down the learning curve. However, this is only the case if the learning effects can be proprietary. If the new entrant in fact can learn from the mistakes of the incumbents, the production cost of the new entrant might actually be lower than for the incumbents.
• Marketing cost. There is often a large fixed component especially in advertising. Advertising is almost exclusively an irrecoverable cost and the new entrant normally would be at a disadvantage. Most extant industries already have several competitors that advertise and to get the same impact a new entrant must spend substantially larger amounts in advertising. In some cases advertising costs can be partly recoverable. This happens with brand extensions, in which the corporate brand is advertised.
• Product innovation. A portion of product development costs is irrecoverable. However, certain knowledge could be gained that is useful for other applications. 

The list above is just an illustration of the analysis that has to be made.

Factor 3. Lower variable unit cost for the incumbents.  This factor is much less important, but could be relevant in specific situations. The incumbents might be able to source raw material at a lower cost. This could be the case when the incumbent is vertically integrated and it might be difficult for the new entrant to buy raw material. Or it could be because the incumbent have learnt to utilise raw material or direct labour more efficiently.

Example: Airlines. In the past, the industry was highly regulated. Foreign airlines were discriminated against, national airlines were forced to fly on unprofitable routes. This was largely changed by deregulation (US in the 1980s, Europe in the 1990s, and Asia in the 2000s). There are still some regulatory restrictions for international flights. Furthermore, most costs of running an airline are variable; the plane can be leased and fuel cost is a function of flight miles. There is very little expensive image advertising required. In other words, there is very little irrecoverable fixed costs involved. Finally, the variable costs are not lower for the incumbents, they have to buy jet fuel on the open market.

Retaliation of incumbents. The above mentioned three factors are not under the control of the the incumbents. Based on an analysis of the three factors an assessment of barriers to entry can be made. However, this is not the whole picture. The incumbents can also retaliate against any new entrant to an industry. Such retaliation is naturally making it more difficult for the new entrant to survive in the industry. Retaliation could take many forms, e.g. more advertising, new products, lower prices, better warranties.

When the telecommunication services industry was deregulated in the 1990s in Europe, the former national monopolists stared with 100% market share. When new entrants entered the industry it would have been possible for the incumbent to pre-emptively lower prices, but it would have been very costly to do so without price discriminating against certain customer groups. So for many years, the new entrants had lower prices than the incumbents. The result was that the incumbents lost market share every year, but the alternative of lowering prices would have been even more costly for the incumbents.

Incumbents are subject to cognitive limitations, just like all companies are. If a new entrant decides to enter in the most profitable industry segment, the incumbents might retaliate more aggressively. A useful strategy would instead be to enter in a less profitable industry segment to get a beach hold in the industry.

The only way for the analyst to understand the retaliation of incumbents is to understand the past behaviour of the incumbents. If they were aggressively retaliating in the past, they can be expected to do the same in the future. Potential new entrants will know this and be less inclined to enter.