Introduction to Modelling Search Costs Flashcards

1
Q

Sequential Search

A

Make a decision after every search

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Non-Sequential Search:

A

Make a decision to carry out a number of searches

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Switching Costs

A

There are switching costs of changing to a new supplier: Dual fuel example - once you have searched, you need to click through, or contact the switching company by phone or sign up online, to arrange the transfer, but no one needs to visit home.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Empirical Modelling of Search Costs - Sorensen (JPE 2000) Example

A

In this paper they examine the prediction that price dispersion will arise when there is a positive probability that a randomly chosen consumer knows only one price. There are differences in search costs among consumers who purchase products with different frequencies. It tests proposition arising from search theory models that consumers will have greater incentives to price shop for repeatedly purchased prescriptions ( for which the consumer is more likely to know the price) than rarely purchased prescriptions.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Search on the Internet

A

Internet search usually involves finding out about a number of suppliers at the same time (e.g. car insurance; cameras). e.g. use search engines.

  • Frictionless commerce would imply Bertrand equilibrium, with no one making any money.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Ellison- Ellison Obfuscation Model (Etrica, 2009)

A

They examine a straightforward product: computer memory. Small firms selling computer parts at pricewatch.com.

Idea: Demond for these products is likely to be extremely priced elastic, so firms set low prices for a basic product, but make more money on add-ons.

They found that total demand elasticity wrt price for low quality very high around -25. This would imply that there would be very low markups. Elasticity is a bigger number than for medium and high prices.

Low quality is effectively a loss leader, low prices for low quality leading to more sales of the higher quality products given their prices.

So, even with internet search firms can make money through “upsales”.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Switching Costs

A

Consumers make many decisions with potentially long term effects: e.g. choosing an electric toothbrush, choosing a fixed rate mortgage, selecting a mobile phone contract.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q
Hotelling Model (Basic - Switching Costs)
Assumptions
A

When Modelling switching behaviour, it makes sense to analyse this within a market where there are horizontally differentiated products.

Assumptions:

  • Firms are located at A and B
  • Consumers are uniformly distributed along the line.
  • Each consumer wants 1 unit of the product
  • “Transportation Cost” depends on the distance from the firm
  • Cost (for the consumer at x) of buying from firm A is Pa +t.x
  • Cost of buyinh from B is Pb+ (1-x)t

Consumers located to the left of Xa buy from A, and vice versa.
Xa is such that Pa + tXa = Pb + (1-x)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Quality Issues

A

For some goods, quality is readily perceived.
For other’s, they are not.
Different types of good, different degrees of information.
Information issues on both firm and consumer side.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Search Goods

A

The consumer can ascertain the features of the good before purchase. The problem is not knowing where/ at what price they may be found. E.g. Tyres (for a normal car); petrol; socks.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Experience Goods

A

Consumer must consume the product in order to ascertain its quality (Consumer May well know where to get the product) E.g. Haircut, Perfume.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Credence GOods

A

Even after consumption, you are not able to ascertain the quality. Maybe after prolonged consumption you can. For example, effectiveness of an anti-virus software; pensions etc.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Symmetric Limited Information on quality

A

Suppose a firm produces a good that is either low quality or high quality, but the firm cannot ascertain the true quality and neither can consumers.
For consumer, valuations rH > rL

For a randomly chosen example of the good, expected value = LanderRh + (1- Lander)Rl for the product:

Lander = the proability that the good is of high quality.

A risk neutral consumer would be willing to pay up to the amount for the product.

Here is there summetric information, and no particular problems exist.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Lemons Market - Akerlof

A

Suppose sellers of used cars know whether their car is of high quality or low quality through experience.

Say 1/2 of cars are high quality.

Then if all cars are offered for sale, market price would be mean value of high and low quality.

But then owners of high quality cars would not be willing to put them on sale because consumers are willing to pay less than they are worth to the sellers.

So all the cars on sale are low quality.
More generally if there is a range of qualities only the lowest survives, the market breaks done.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly