colusion

E C O N O M Í A

TESIS de MAGÍSTER

I N S T I de Economía E
D
Instituto T U T O

DOCUMENTO
DE TRABAJO

2012

Indicadores de Colusión en el Mercado Farmacéutico Chileno

Víctor Serey.

www.economia.puc.cl

´
PONTIFICIA UNIVERSIDAD CATOLICA DE CHILE
INSTITUTO DE ECONOM´
IA
MAGISTER EN ECONOM´
IA

TESIS DE GRADO
MAGISTER EN ECONOM´
IA

´˜
Serey, Zuniga, V´
ıctorDaniel

Agosto, 2012

palabra

´
PONTIFICIA UNIVERSIDAD CATOLICA DE CHILE
INSTITUTO DE ECONOM´
IA
MAGISTER EN ECONOM´
IA

´
Indicadores de Colusion en el Mercado
´
Farmaceutico Chileno

por
´˜
V´
ıctor Daniel Serey Zuniga

´
COMISION:
ALEJANDRA TRAFERRI
AR´
ISTIDES TORCHE
´
TOMAS RAU

Agosto, 2012
Santiago, Chile

palabra

´
Indice
1. Introducci´nXXXXXXXXXXXXXXXXXXXXXXXXXXXX
o

1

2. Detectando un Cartel
¿Qu´ entendemos por colusi´n?
e
o
¿Qu´ sabemos de los carteles detectados?
e
Ambig¨edades entre la teor´ y la evidencia
u
ıa

4
4
5
7

3. El Caso de Colusi´n de las Farmacias
o
Contexto y desarrollo del mercado farmac´utico en Chile
e
Caracter´
ısticas de los productos cuestionados por la FNE
La acusaci´n y el fallo del TDLCo
Antecedentes adicionales

10
10
11
12
16

4. Metodolog´ Indicadores de Colusi´n
ıa:
o
Correlaci´n entre precio y demanda
o
Estabilidad de precios
Correlaci´n entre los precios
o
Correlaci´n entre cuotas de mercado
o

18
19
20
22
22

5. Descripci´n de los datos
o
Sobre la base de datos y los medicamentos
Sobre los precios considerados
Comportamiento a nivel agregadoSobre las cuotas de mercado

25
25
26
29
31

6. Resultados
CM1: Correlaci´n entre precio y cantidad
o
CM2: Estabilidad de precios
CM3: Correlaci´n entre precios
o
CM4 y CM5: Indicadores de cuotas de mercado
Resumen de resultados

34
35
42
48
49
57

7. Discusi´n
o
Usando la informaci´n inculpatoria
o
Sobre los costos
Extensiones del presente trabajo

59
60
66
688. Conclusi´n
o

70

Referencias

72

Anexos

75

palabra

Abstract

The aim of this paper is to study new evidence of collusion emerging from the pricing
behavior in the famous case of the Chilean pharmacies. For that purpose, I use a set of
collusive markers reviewed by Joe Harrington (2008) based on theoretical and empirical
literature on collusion. These markersexploit the fact that particular correlations between
prices, quantities and market shares are expected under collusion.
In order to evaluate these markers, I use the transactional database of sales, for 220 disputed drugs, covering from January 2006 to December 2008. These were reported by each of
the three questioned pharmacies (Cruz Verde, Farmacias Ahumada y Salcobrand) to the
authority.
Themain result is that almost half of the disputed drugs met simultaneously all the
markers. This result provides independent evidence favoring a collusive hypothesis which,
complemented with criminal evidence, shows that the pricing behavior is highly consistent
with collusive explanations. For drugs that do not exhibit a seasonal pattern of demand,
the classical result of Rotemberg & Saloner(1986) of negative correlation between price
and demand is confirmed. Likewise, for drugs following a seasonal pattern of demand, I
found agreement with the price behavior predicted by Haltiwanger & Harrington (1991).
It is also confirmed that, during the period identified by the authority like collusion, the
drugs prices show greater stability and more positive correlation compared to the behaviorduring 2006 (competive period). Market shares are also more stable during this period
for most of the drugs, while for others there is evidence of non-rational behavior for a
competitive regime.

iii

palabra

Resumen

El principal objetivo de esta tesis es estudiar nueva evidencia econ´mica que surge de los
o
indicadores de colusi´n compendiados de la literatura por Joe Harrington...