República de Colombia Departamento Nacional de Planeación

Dirección de Estudios Económicos

ARCHIVOS DE ECONOMÍA

The cost of disinflation in Colombia - A sacrifice Ratio Approach-

José Daniel REYES PEÑA Documento 243 13 de Noviembre de 2003. La serie ARCHIVOS DE ECONOMIA es un medio de la Dirección de Estudios Económicos, no es un órgano oficial del Departamento Nacional de Planeación. Sus documentos son de carácter provisional, de responsabilidad exclusiva de sus autores y sus contenidos no comprometen a la institución.

1

The Cost Of Disinflation In Colombia – A Sacrifice Ratio Approach –

José Daniel REYES PEÑA*. Departamento Nacional de Planeación.

The disinflation episode that Colombia has experienced since 1991 has been costly in

terms of the output and employment. This paper computes the sacrifice ratio not only for

this period but also for each disinflation episode in the Colombian economy since 1957.

In order to accomplish the objective mentioned above, a range of methodologies are

used to compare the sacrifice ratio estimated for each episode. The results show that the

current disinflation episode appears to be not as costly as the previous one, however an

explanation of why the sacrifice ratio of the current episode should be calculated since

1997 is proposed.

[This Draft: October 2003]

Keywords: Sacrifice Ratio and Disinflation Monetary Policy.

JEL Classification Numbers: E52, E58.

1. Introduction

Since 1991 the Colombian economy has shown a disinflationary path. This is the result of the

fundamental objective of the central bank, established in the 1991 constitution, of preserving the buying

power of the currency. Although the achievement of this goal by the central bank has been remarkable

(figures 1 and 2), there is an economic debate about its responsibility in the mediocre economic growth that

Colombia has had in the last five years. Colombia’s average GDP growth in the period 1990-1995 was 4,5%

while in the period 1996-2000 it was 0.92%1.

The relationship between inflation and output has been widely studied both empirically and

theoretically. The most useful empirical tool to measure this relation is the sacrifice ratio, concept that was

introduced by Okun (1978). He proposed to exam a family of Phillips curve models to derive the cost of

* I am grateful for comments and corrections from Gabriel Piraquive, Andrés Escobar, Natalia Salazar, Fernando Mesa, Carlos Castro and Claudio Karl. Specially, I would like to thank Laurence Ball and Lawrence Huiyan Zhang for their comments to my work. I also would like to thank Arzu Cetinkaya and Devirm Yavuz from the research department of the Central Bank of the Republic of Turkey. All errors are mine. Contact information: jreyes@contractor.iadb.org.

2

disinflation in terms of the percentage of output lost during a given period. Later, Gordon and King (1982)

used a vector autoregression models (VAR) to estimate the sacrifice ratio. After that, Ball (1994) proposed a

different technique for estimating the sacrifice ratio for each disinflation episode based on the movements of

trend output and trend inflation2. Finally, Cecchetti and Rich (1999) used a structural vector autorregresion

(SVAR) approach to create an estimation of the sacrifice ratio, based on a structural identification of

aggregate demand and aggregate supply shocks.

The relationship between inflation and output or unemployment has also been studied in Colombia

from different perspectives. Studies made by Clavijo (1994), Partow (1995), Sánchez and Nuñez (1998),

Misas and López (1999), and Birchenal (1999) have found evidence of the trade-off between inflation and

unemployment in the short run for the Colombian economy. Furthermore, there have been three papers which

have studied the sacrifice ratio: Sarmiento et. al. (1998) using Ball’s sacrifice ratio methodology analyzed the

desinflationary policies applied in nine countries, whose economic structure is similar to Colombia’s. Gómez

and Julio (2000), within a model of transmission mechanisms of monetary policy estimated the sacrifice ratio

under different kinds of assumptions on the formation of inflation expectations. Finally, Gómez (2002) used a

model with a wage-price system, a real block, and an inflation targeting rule, which resemble price setting in

the Colombian economy to estimate the sacrifice ratio for different specifications of wage indexation.

The main purpose of this paper is to shed some light on the methodological discussion about the true

sacrifice ratio of the Colombian economy. In this sense, it computes the sacrifice ratio in each disinflation

episode following Ball’s methodology. Special attention is given to the current disinflation episode and how

robust the sacrifice ratio is under alternative specifications of inflation trend and output trend. Alternatives

methodologies such as Zhang (2001), and Cecchetti and Rich (1999) will be explored. Furthermore, three

alternatives measures for the output gap are explored and applied to Zhang’s methodology.

This paper is divided in seven sections: Disinflation episodes are identified in section 2; section 3

uses Ball’s methodology to estimate the sacrifice ratio in each disinflation episode; section 4 applies Zhang

(2001) methodology in order to study long-lived effects of disinflation and to observe the effects of

disinflation on unemployment; section 5 test the robustness of the sacrifice ratio under different specifications

of output trend as well as inflation trend and explores the methodology proposed by Cecchetti and Rich

(1999); section 6 makes some comments about the cost of the current disinflation policy; and, finally, some

concluding remarks are presented in section 7.

1 In fact, in 1999 Colombia had its worst economic performance in all of its history, GDP growth was –4.20%. 2 Following Ball’s work, Zhang (2001) extended this methodology to consider possible long-lived effects over the output (i.e. hysteresis) in a disinflation episode. His work also intends to measure the sacrifice ratio in terms of unemployment.

3

2. Selecting Episodes

In order to select the episodes of disinflation in Colombia, Ball’s methodology is followed. In his

work he defined a disinflation period as: “any period that starts at an inflation peak and ends at a though

with annual rate at least two points lower that the peak”3. For doing this, three assumptions are required

accordingly:

Trend inflation is the centered nine-quarter moving average of inflation45.

An inflation peak is a quarter in which trend inflation is higher than the four previous and the four

following quarters.

An inflation though is a quarter in which trend inflation is lower than the four previous and the four

following quarters6.

Figure 1 and Figure 2 show the trend inflation in quarterly and annual data. Beginning in 1956 it can

be identified six complete disinflation episodes (1957-1961, 1963-1965, 1966-1968, 1974-1975, 1977-1978,

1980-1984) and the beginning of another (1991).

Figure 1. Trend Inflation and disinflation episodes(1956 - 2002)

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

1956

1958

1960

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

Data Source: DANE and Author's calculation.

Figure 2. Trend inflation and desinflation episodes(1978:1 - 2001:4)

0.0%5.0%

10.0%15.0%20.0%25.0%30.0%35.0%

I 78

I 79

I 80

I 81

I 82

I 83

I 84

I 85

I 86

I 87

I 88

I 89

I 90

I 91 I 9

2I 9

3I 9

4I 9

5I 9

6I 9

7I 9

8I 9

9I 0

0I 0

1Data Source: DANE and Author's calculation.

Looking at trend inflation between 1991 – 2001, it is notable that the current disinflation episode is

characterized by two sub-periods: the first one, which I called “moderate disinflation period”, began in the

second quarter of 1991 and ended in the third quarter of 1997; the second one, which I called “pronounced

disinflation period” began in the third quarter of 1997 and ended in the second quarter of 2001.

3 See Ball (1994) page 157. 4 Inflation is measured by the percentage change in the consumer price index (CPI). 5 This definition applies when there are quarterly data. If there are annual data, trend inflation is the centered eight quarters moving average of inflation. The four quarters of the year, two of the previous year and two of the next year.

4

6 For annual data, a peak (though) is a quarter in which trend inflation is higher (lower) than the previous and following year.

5

3. Standard Estimation Of Sacrifice Ratio

The sacrifice ratio has its origins in the Phillips curve, concept developed by A. W. Phillips in 1958.

He found a negative relationship between inflation (defined as the growth rate of wages) and the

unemployment rate for England for the period 1861-1957. This finding was broadly tested around the world

and most studies confirmed the hypothesis of the existence of the trade-off between the inflation rate and

unemployment rate in the short run.

In Colombia, Clavijo (1994), Partow (1995), Sánchez and Nuñez (1998), Misas and López (1999),

Birchenal (1999), and Gómez and Julio (2000) used different forms of the Phillips curve and found evidence

of this trade-off in the short run. Figures 3 and 4 show the relation between inflation rate and unemployment

rate in Colombia with quarterly and annual data. It is clear that the relationship between these variables is

negative, as it has been confirmed by the previous studies. Specifically, Gómez and Julio (2000) estimated a

sacrifice ratio of 0.7897 or 0.322 using two different assumptions on the formation of inflation expectations in

an augmented Phillips curve8. Gómez (2002) estimates the sacrifice ratio building a model to resemble price

setting behavior in Colombia under different specifications of wage indexation; his results indicate that when

indexation is backward looking disinflation implies output cost, when there is a decreasing weight of the

backward looking argument in wage indexation the sacrifices ratio is lower, and with credibility of the central

bank disinflation may lead to a negative sacrifice ratio.

Figure 3. Phillips Curve (1977:1 - 2002:1)

0.0%5.0%

10.0%15.0%20.0%25.0%30.0%35.0%40.0%45.0%

5.0% 8.0% 11.0% 14.0% 17.0% 20.0% 23.0%

Unemployment Rate

Infla

tion

Rat

e

Data Source: DANE and Author's calculation.

Figure 4. Phillips Curve (1977 - 2001)

0.0%5.0%

10.0%15.0%20.0%25.0%30.0%35.0%

5.0% 8.0% 11.0% 14.0% 17.0% 20.0% 23.0%

Unemployment Rate

Infla

tion

Rat

e

Data Source: DANE and Author's calculation.

7 For this scenario, he also found a confident interval of (0.5 , 1.785) for the sacrifices ratio. 8 A simple Phillips curve model can be represented as ttt

ett yy εαππ +−+= )( * , where tπ is the inflation

rate, ty is the actual output level and *ty is the “full employment” output level. Moreover 10 << α . In this

context, the sacrifice ratio is defined as α1

.

6

The method to compute the sacrifice ratio based on the Phillips curve has some implications that

makes it less useful for analyzing the cost of different disinflation episodes. This happens because it assumes

that the sacrifice ratio is symmetric on inflation. For these reasons, there are other studies that compute the

sacrifice ratio based on ad hoc estimates of the change in inflation and output losses. Mankiw (1991), for

example, used the Okun’s law for United States in order to compute the sacrifice ratio of Volcker disinflation.

In order to develop a systematic estimation of the sacrifice ratio in any disinflation episode, Ball

(1994) proposed a method to estimate it as the ratio of the change in trend inflation and the sum of output

losses over a disinflation episode. The change in trend inflation is the difference between inflation trend

(section 2) at the peak (πP) and at the though (πT). The output losses are the difference between the log of

actual output (yt) and the log of trend level of output (y*t). He assumed that:

Output is at its natural level at the start of disinflation episode (inflation peak).

Output returns to its potential level four quarters after the end of an episode.

Between this two points potential output growths log-linearity.

Then the sacrifice ratio for each disinflation episode is calculated as:

( )

PT

T

Pttt yy

SRππ −

⎥⎦

⎤⎢⎣

⎡−

=∑

+

=

4*

[1]

The output losses in any disinflation episode can be represented graphically as:

P T T+4

Potencial Output

Actual Output

yt

Time

Output Losses

Figure 5. Sacrifices Ratio by the Standard Method

This methodology cannot be used to compute the sacrifice ratio for the current disinflation episode

because equation [1] needs to know the time when this episode ends. The data showed in section 2 (figure 1

and 2) cannot be used to conclude if the disinflation episode has ended because the trend inflation is still

7

falling. For this reason the estimates, following Ball’s methodology, are only for the six complete

disinflations episodes showed in figure 1. The results are showed in table 1 for annual data, and table 2 for

quarterly data.

Table 1. Sacrifices Ratios in Colombia

1957-1984Table 2. Sacrifices Ratios in Colombia

1980:III -1984:1

Episode Length in Years

Initial Inflation

Change inInflation

SacrificesRatio Episode Length in

QuartersInitial

InflationChange inInflation

SacrificesRatio

1957 - 1961 4 14.3% 7.8% 0.25 80:III 84:I 14 27.2% 8.8% 0.921963 - 1965 2 20.3% 8.9% -0.021966 - 1968 2 13.4% 6.4% 0.17 Source: Author's calculations1974 - 1975 1 23.9% 2.3% 0.221977 - 1978 1 27.3% 5.1% -0.121980 - 1984 4 27.1% 8.1% 0.21

Source: Author's calculations

The sacrifice ratios lie within the interval (0.0 to 3.6) found by Ball (1994) for a sample of 28

disinflation episodes in 9 countries. The ratio is positive in four disinflation episodes and it suggests that the

cost had been moderate until the eighties. The interpretation of these sacrifice ratios is what percent the

current output has to fell, from its long run level, due to a reduction of 1% in trend inflation.

The sacrifice ratio calculated for 1980-1984 disinflation episode confirmed Ball’s results that annual

data understate output losses. Nevertheless, due to Colombia’s data availability, this cannot be tested for other

disinflation episodes.

Figures 6 and 7 show the difference between output trend and actual output in each disinflation

episode.

Figure 6. Trend output vs Actual Output1956-1984

7.1

7.2

7.3

7.4

7.5

7.6

7.7

7.8

1956

1958

1960

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

Log

Out

put

Data Source: DANE and Author's calculation.

Figure 7. Trend output vs Actual Output1980:3-1985:1

7.0

7.0

7.0

7.0

7.0

7.0

7.0

7.1

7.1

7.1

III 8

0

I 81

III 8

1

I 82

III 8

2

I 83

III 8

3

I 84

III 8

4

I 85

Log

Out

put

Data Source:Dane and Author's calculation.

8

There are two disinflation episodes with negative sacrifice ratios (1963-1965 and 1977-1978). This

means that in the period when inflation trend had been falling the actual output was higher that its trend. In

other words, it could be inferred that in these periods tight monetary policy did not reduce output level

because there were other factors that hit positively the economy (i.e. supply shocks)9.

Although the sacrifice ratio calculated in this section is the most common tool used to measure the

output loss due to a disinflation policy, it has some limitations. Among these limitations are:

• As Zhang (2001) argues, this methodology could tend to underestimate the true sacrifice ratio if

economic shocks have long live effects in the economy.

• Because Ball’s methodology assumes that the recession, in a disinflation episode, is totally cause by

a tight monetary policy it does not include the possible effects of supply or structural shocks.

Cecchetti and Rich (1999) proposed an alternative methodology to address this problem.

• As Ball (1994) admitted, his methodology is very sensible to the definition of trend output. Changes

over how it is estimated vary significantly the results of sacrifice ratio10.

9 Nevertheless this hypothesis must be tested in future studies. 10 On the other hand, changes over the methodology used to estimate trend inflation do not varies the calculates of sacrifice ratio.

9

4. Alternatives Estimations Of Sacrifice Ratio

This section uses Zhang’s (2001) methodology on Colombia data. The sacrifice ratio will be

estimated for the same disinflation episodes with the assumption that the shocks could have long live effects

over the output. This implies that the output cost, due to a disinflationary policy, could be greater than the one

shown in the previous section.

Other approach to the problem is to compute the sacrifice ratio in terms of unemployment. This

methodology also follows Zhang’s work.

4.1. Sacrifice Ratio under long-lived effects.

Zhang’s methodology takes into account the possible persistence effect of the shocks in an economy.

To do this, he made the following assumptions:

• Trend inflation is measured in the same way that Ball (1994). The assumption that output is at its

natural level at the start of the disinflation episode is also hold in this methodology.

• For each disinflation episode, the rate of growth of natural output is equal to the rate of growth of the

Hodrick-Prescott filter (HP filter) of the log of output. This assumption introduces the possibility of

some persistence effects in the output.

• He used two smoothing parameters of the HP filter (1600 and 16000) because the filter is affected by

recessions. For this reasons the sacrifice ratio could have been underestimated because of the

recession, due to a disinflationary policy, which could have affected its natural level11.

• Finally, the compute of the sacrifice ratio uses the same equation ([1]) that Ball proposed in 1994.

Figure 8 shows the graphic representation of this methodology. It compares the output loss as well as

the potential output with the methodology presented in the preceding section. Area A represents the output

loss under Ball’s methodology meanwhile area B is the output loss that must be added to area A if the

economy shows some kind of persistence. As a result, the short-term loss in a economy is area A+B. Area C

is the long term loss due to the persistence of the shocks in the economy. In this study, as well as in Zhang’s

work, I am only concentrated on the short-term loss.

11 This issue is very critical to the Colombian economy because in the current disinflation episode the output felt to its worst level in all historical economic data.

10

P T T+4

Actual Output

yt

Time

Old Potential

Figure 8. Sacrifices Ratio in a Economy with Persistence Effects

A

New Potential Output

BC

This approximation has an important implication for Colombian data because it does not need to

know when the disinflation episode ends due to the assumption that natural output grows at the rate of the HP

filter. In other words, it is possible to compute the sacrifice ratio for the current disinflation episode (figure 9)

with the available data12. Figure 10 shows the output gap in this episode using the rate of growth of HP filter

with a smooth parameter equal to 1,600.

Figure 9. Trend Inflation in the Actual Disinflation Episode

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

II 9

1

II 92

II 93

II 94

II 95

II 96

II 97

II 98

II 99

II 00

II 01

Data Source: DANE and Author's calculation.

Figure 10 .Output Gap in the Actual Disinflation Episode (Lambda =1600)

7.1

7.1

7.2

7.2

7.3

7.3

7.4

II 9

1

II 92

II 93

II 94

II 95

II 96

II 97

II 98

II 99

II 00

II 01

Log

Out

put

Data Source: DANE and Author's calculation.

This episode is particular since the output level is near its natural level during most part of the

episode. This implies that a tight monetary policy, established in order to reduce the inflation rate in the

beginning of the nineties, did not affect economic activity. ¿But what explains this atypical behavior of the

Colombian economy?, some insights to this questions are going to be explored in the sixth section.

12 For the current disinflation episode, the sacrifice ratio under this methodology is ( )

4:20004:1991

4:2000

4:1991

*

ππ −

⎥⎦

⎤⎢⎣

⎡−

=∑

=ttt yy

SR for

quarterly data.

11

The results for annual and quarterly data are showed in tables 3 and 4 respectively. The average of all

disinflation episodes is greater when the sacrifice ratio takes into account possible long –lived effects13.

Although there is one episode (1966-1968) where the sign changed, compared to the application of Ball

methodology, all episodes show a greater effect (in absolute value) with respect to those showed in table 1.

Even thought the magnitude of the sacrifice ratios changed using this alternative methodology, they coincide

in the fact that the three most important disinflation episodes in Colombia are: 1980-1984, 1974-1975 and

1957-196114.

Table 3. Sacrifices Ratios in Colombia with Long-Lived Effects

1957-1984Table 4. Sacrifices Ratios in Colombia with Long-Lived Effects

1980:III -2001:1

Episode Length in Years

Initial Inflation

Change inInflation

SacrificesRatio Episode Length in

QuartersInitial

InflationChange inInflation

SacrificesRatio

(HP 1600)

SacrificesRatio

(HP 16000)

1957 - 1961 4 14.3% 7.8% 0.25 80:III 84:I 14 27.2% 8.8% 2.11 2.271963 - 1965 2 20.3% 8.9% -0.33 91: II 01:IV 43 29.5% 22.3% 0.89 1.011966 - 1968 2 13.4% 6.4% -0.331974 - 1975 1 23.9% 2.3% 0.63 Source: Author's calculations1977 - 1978 1 27.3% 5.1% -0.761980 - 1984 4 27.1% 8.1% 1.831991 - 2001 10 29.6% 21.9% 0.92

Source: Author's calculations

These results show that sacrifice ratios are greater with quarterly data than with annual data. As it

was expected, the sacrifice ratio with a natural output level calculated with a smooth parameter of 1,600 is

smaller than the another one calculated with a smooth parameter of 16,000, since the first one is very affected

by recessions, specially the last recession.

The current disinflation episode does not appear to be the most important one. This result is

something curious because this period had the worst economic recession in Colombia, however it can be

explained by the fact that the output was closer to its potential at the beginning of episode and because we do

not still know when this episode will end. Moreover, the fall in inflation is larger in the current disinflation

episode than in the previous one.

4.2. Sacrifice Ratio from unemployment data.

Zhang (2001) used this methodology to explore the sacrifice ratio as the cost of having a higher

unemployment rate due to a disinflation policy. His methodology has the following assumptions:

13 In fact, the average for annual data following Ball’s methodology is 0.12 meanwhile with Zhang’s methodology is 0.32.

12

• At the beginning of the disinflation episode (inflation peak), the unemployment rate is at its natural

level (or NAIRU15).

• The NAIRU is constant through the disinflation episode16.

The differences between the actual unemployment rate and its natural level is the unemployment

loss. In this context the sacrifice ratio can be constructed as follows:

( )

TP

T

Pttt UU

SRππ −

⎥⎦

⎤⎢⎣

⎡−

=∑

+

=

4*

[2]

Due to unemployment data availability, this exercise is estimated only for quarterly data17. For this

reason, only two episodes are analyzed: 1980:I – 1984:I and the current disinflation episode. The difference

between the actual unemployment rate and the natural level in each disinflation episode is showed in figures

11 and 12.

Figure 11. Actual Unemployment vs Natural Rate of

Unemployment (1980:III - 1984:I)

5%

10%

15%

III 8

0

IV 8

0

I 81

II 8

1

III 8

1

IV 8

1

I 82

II 8

2

III 8

2

IV 8

2

I 83

II 8

3

III 8

3

IV 8

3

I 84

Une

mpl

oym

ent R

ate

Data Source: DANE and Author's calculation.

Figure 12. Actual Unemployment vs Natural Rate of

Unemployment (1991:II - 2000:IV)

5%

10%

15%

20%

II 9

1

II 92

II 93

II 94

II 95

II 96

II 97

II 98

II 99

II 00

Une

mpl

oym

ent R

ate

Data Source: DANE and Author's calculation.

From figure 11 it is clear that unemployment was not affected in the episode of the eighties

meanwhile the current disinflation episode appears to have a greater impact in the unemployment rate.

However, the sacrifice ratio is smaller in the second episode because the change in inflation was higher in that

period. Table 5 shows the calculates of sacrifice ratio.

14 The set of episodes showed in tables 3 and 4 do not include the current disinflation episode because it was not taken into account with Ball’s methodology. 15 Non Accelerated Inflation Rate of Unemployment. 16 The author also calculated the sacrifices ratio using the standard method. He assumes that NAIRU grows log-linearly in the disinflation episode. Nevertheless, he recognized that it may not be a good idea to apply this method to the unemployment data because there are two negatives sacrifice ratios. For these reason I did not use this method for the Colombian economy. 17 Unemployment data corresponds to National Households Surveys for seven cities from 1977 to 2000. The cities which are include are: Bogotá, Barranquilla, Cali, Medellín, Bucaramanga, Cali y Pasto.

13

Episode Length in Quarters

Initial Inflation

Change inInflation

SacrificesRatio

80:III 84:I 14 27.2% 8.8% 3.4291: II 00:IV 39 29.5% 21.2% 3.34

Source: Author's calculations

Table 5. Sacrifices Ratio with Unemployment Data

For the two disinflation episodes analyzed the sacrifice ratio is positive. Its interpretation is that

when the inflation trend went down for 1%, the unemployment losses ( Ut –U*t ) go up by 3.42% in the

eighties and 3.34% in the current disinflation episode.

For the disinflation episode from 1980 to 1984, this methodology appears appropriate because the

natural rate of unemployment was very close to the rate estimated by Núñez and Bernal (1998), and Henao

and Rojas (1998), moreover the level of unemployment rate at the beginning of the current disinflation

episode was very close to 10%. On the other hand, this argument does not appears to be very convincing for

the current episode given the increasing literature about the existence of hysteresis effects in the

unemployment rate18.

18 See Guataquí (2000), Reyes (2002) y Sánchez et. al. (2002).

14

5. Some Changes Over The Methodologies And An Econometric Approach

One of the concerns on Ball’s methodology is the lack of robustness to changes over how the

potential output is computed. In this section, I will change the assumptions about trend output as well as trend

inflation to see how robust are the preceding estimates for Colombia.

Another concern is on the effects of supply shocks over the methodology. Cecchetti and Rich (1999)

proposed the estimation of a structural VAR system for this problem. In the second part of this section, I will

apply this technique for Colombia.

5.1. Changes over the Methodologies.

5.1.1. Trend Output.

Three alternative measures of the output gap are used. The first one is the difference between

capacity utilization and its average over time, the second one is using a potential output constructed by the

estimation of aggregate production function for the Colombian economy, and the third one is to apply the

“one-side” HP filter to avoid the potential negative bias in Zhang’s methodology.

The capacity utilization is a close indicator of economic performance. In this context, a proxy of the

output gap is constructed as the difference between actual capacity utilization and its average over time. The

data corresponds to the survey on the opinion of the entrepreneurs and businessmen gathered by Fedesarrollo.

Figure 13 illustrates the proxy of output gap and each disinflation episode.

Figure 13. Capacity Utilization . Proxy of Output Gap.

60.0%

65.0%

70.0%

75.0%

80.0%

I 81

I 82

I 83

I 84

I 85

I 86

I 87

I 88

I 89

I 90

I 91 I 9

2I 9

3I 9

4I 9

5I 9

6I 9

7I 9

8I 9

9I 0

0I 0

1I 0

2

Data Source: Fedesarrollo.

15

The standard methodology of the sacrifice ratio uses a statistic approach for output trend. In the

second approach, I am going to use a economic approach to estimate a potential output. Therefore, I use the

estimation of the potential output for the Colombian economy using a Constant Elasticity of Substitution

(CES) production function19. Figure 14 and 15 show the potential output as well as the output gap estimated

using this methodology.

Figure 14. Actual Output vs Potential Output

7.07.17.17.27.27.37.37.4

I 81

I 82

I 83

I 84

I 85

I 86

I 87

I 88

I 89

I 90

I 91 I 9

2I 9

3I 9

4I 9

5I 9

6I 9

7I 9

8I 9

9I 0

0I 0

1I 0

2Data Source: DANE and Author's calculation.

Figure 15. Output Gap from a CES Production Function.

-0.0-0.0-0.0-0.0-0.00.0

I 81

I 82

I 83

I 84

I 85

I 86

I 87

I 88

I 89

I 90

I 91 I 9

2I 9

3I 9

4I 9

5I 9

6I 9

7I 9

8I 9

9I 0

0I 0

1I 0

2

Data Source: DANE and Author's calculation.

Finally, as Zhang shows in his paper, the rate of growth of the trend output could be negatively

affected by the recession during the disinflationary episode. As a result, the methodology is affected by a

negative bias in the estimates of the sacrifice ratio. A way to overcome the bias is with the “one sided” Hp

filter, that uses the growth of rate of the HP filter only up to the beginning point of the episode. This approach

could be very useful to estimate the current sacrifice ratio because the recession in 1998 affected the output

trend as it is shown in figure 16.

Figure 16. "one side" HP filter vs Huiyan's Methodology

7.1

7.2

7.2

7.3

7.3

7.4

7.4

I 90

I 91 I 92

I 93

I 94

I 95

I 96

I 97

I 98

I 99

I 00

I 01

Log

Out

put

"one side" Hp filter

Huiyan's Methodology

Data Source: DANE and Author's calculation.

Table 6 shows the estimates for each of these three alternatives. The first and the third approaches

support one of the previous section’s conclusions; the sacrifice ratio of the eighties appears to be greater that

19 This method estimates an aggregate production function. I calculated the natural levels for the inputs (capital and labor), which I used to estimate the potential output for Colombian economy. For more details see Reyes (2003) or appendix A.

16

the sacrifice ratio of the current disinflation episode. According to the second approach, the current

disinflation has a larger sacrifice ratio. There is an important change between the estimates of the sacrifice

ratio in the eighties with Zhang’s methodology and with the “one sided” methodology. This is because the

growth rate of the output trend in the original methodology was 0.05% while with the “one sided”

methodology it was 0.09%.

Quarterly data

Episode Length in Initial Change inQuarters Inflation Inflation Cap. Uti P. F. One Side

80:III 84:I 14 27.2% 8.8% 2.81 1.65 8.1791: II 01:IV 39 29.5% 21.2% 0.74 2.04 3.60

Source: Author's calculations

Sacrifices Ratio

Table 6. Alternative calculations of Sacrifices Ratio

Using annual data, the three approaches supported the idea that the sacrifice ratio in the eighties was

larger that the current sacrifice ratio. All episodes kept the sign they had with Zhang’s methodology except

the episode from 1974 to 1975 where the sign was negative with the production function approach.

Annual DataLength in Initial Change in

Years Inflation Inflation Cap. Uti P. F. One Side HP

1963 - 1965 2 20.3% 8.9% --- --- -0.201966 - 1968 2 13.4% 6.4% --- --- -0.101974 - 1975 1 23.9% 2.3% --- -0.69 1.291977 - 1978 1 27.3% 5.1% --- -0.72 -0.491980 - 1984 4 27.1% 8.1% 0.66 0.76 2.711991 - 2001 10 29.6% 21.9% 0.16 0.50 1.39

Source: Author's calculations.

Table 7. Alternative calculations of Sacrifices Ratio

Sacrifices Ratio

Note: Empty spaces indicate that there was not available data for this episodes due to differents approaches.

Episode

5.1.2. Trend Inflation.

I use three measures of trend inflation. Instead of constructing trend inflation as a centered nine-

quarters moving average, I use two alternatives approaches: five and four quarters moving average. The

results are showing in table 8 for annual data20. This exercises shows little difference between the estimates of

20 This exercise was not done for quarterly data because there was only one episode to compare.

17

sacrifice ratio under Ball’s methodology. In other words the standard methodology appears no to be sensible

to changes in the method used to obtain trend inflation21.

Annual DataLength in Initial Change inQuarters Inflation Inflation 9 quarters 5 quarters 4quarters

1957 - 1961 4 14.3% 7.8% 0.25 0.20 ---1963 - 1965 2 20.3% 8.9% -0.02 -0.01 -0.011966 - 1968 2 13.4% 6.4% 0.17 0.13 0.121974 - 1975 1 23.9% 2.3% 0.22 0.19 ---1977 - 1978 1 27.3% 5.1% -0.12 -0.05 ---1980 - 1984 4 27.1% 8.1% 0.21 0.19 0.17

Source: Author's calculations.

Table 8. Alternative calculations of Sacrifices Ratio

Sacrifices Ratio

Note: Empty spaces indicate that the episode was lost due the different approaches of trend inflation.

Episode

5.2. An Econometric Approach to Sacrifice Ratio.

This subsection develops one of the methods proposed by Cecchetti and Rich (1999). In their paper,

they derived estimates of the sacrifice ratio from different kinds of structural VAR models22. They found that

the estimates are very imprecise and that the degree of imprecision increases with the complexity of the model

used. This subsection makes a brief summary of the first model proposed by Cecchetti and Rich (1999) and its

applications for Colombian data.

This approach consists in the estimation of sacrifice ratio from a SVAR23 that includes output (yt)

and inflation (πt). The system can be written as:

tntntttt XXXXX ε+Β++Β+Β+Β= −−− .......22110

With: [3]

⎥⎦

⎤⎢⎣

⎡∆∆

=−

−−

it

itit

yX

π ⎥

⎦

⎤⎢⎣

⎡=Β ii

ii

i bbbb

2221

1211 ⎥⎥⎦

⎤

⎢⎢⎣

⎡=

πε

εε

t

yt

t ⎥⎦

⎤⎢⎣

⎡=Β

00021

012

0 bb

ni ,....,2,1,0=

21 I also try to obtain the inflation rate as the percentage change in the GDP deflator. In this case, the results do not change and mostly all disinflation episodes remains equal as well as the sacrifice ratios. 22 In his comment to Ball’s original work, Cechetti suggested the use of this methodology in order to take into account the possible effects of a supply shocks over the path of output and inflation. 23 “A SVAR can be viewed as a dynamic simultaneous equation model with identifying restrictions based on economic theory” Cecchetti and Rich (1994), page 7.

18

Where εt is a vector of structural errors that can be expressed as the innovation pertaining to an

aggregate supply ( ytε ) and aggregate demand ( πε t ). Due to the assumption that each shock comes from a

different source, it is required that [ ] IE tt ='εε .

In the context of the estimation of the sacrifice ratio, it is useful to present the system as its vector

moving average representation (VMA) because it uniquely identified the impulse responses of the system to

the structural shocks. In particular, we are interested in the dynamic output response to a demand shock (i.e.

monetary shock) in a given number of periods. Equation [4] shows the VMA representation for this system.

( )∑∞

=−−− ==+++=

022110 .........

itititttt LAAAAAX εεεεε

[4]

With ⎥⎦

⎤⎢⎣

⎡= ii

ii

i aaaa

A2221

1211

For an explanatory purpose it is useful to present [4] as:

∑∑

∑∑∞

=−

∞

=−

∞

=−

∞

=−

+=∆

+=∆

022

021

012

011

iit

i

i

yit

it

iit

i

i

yit

it

aa

aay

π

π

εεπ

εε [5]

Cecchetti and Rich (1999) use [5] to construct an estimation of the sacrifice ratio. The sum of each

element ia22 measures the effect of demand shock on its level. In the same way, the sum of each element

ia12 measures the effect of demand shock on output level. At this point, it is important to remember that we

are interested in the cumulative effect of output. For this reason, and in the same spirit of Ball’s original work,

the sacrifice ratio can be written as:

⎟⎠

⎞⎜⎝

⎛

⎟⎟⎠

⎞⎜⎜⎝

⎛

=⎟⎠

⎞⎜⎝

⎛

⎟⎠

⎞⎜⎝

⎛++⎟

⎠

⎞⎜⎝

⎛+⎟

⎠

⎞⎜⎝

⎛

=

∂∂

⎟⎟⎠

⎞⎜⎜⎝

⎛∂

∂

=

∑

∑∑

∑

∑∑∑∑

=

= =

=

===

+

=

+

τ

τ

τ

τ

πτ

τ

π

ε

επ

ετπ

022

0 012

022

012

1

012

0

012

0...

)(

i

i

i

i

j

i

i

i

i

i

i

i

i

i

t

t

j t

jt

a

a

a

aaay

S [6]

19

Equation [6] requires a time horizon ( )τ in which the impact of a demand shock is going to be

analyzed. The numerator measures the cumulative output loss while the denominator is the difference in the

level of inflation after τ periods. Figure 17 illustrates the relationship between the sacrifice ratio and the

structural impulse response function for output and inflation.

Source: Cecchetti and Rich (1999)

Output

Time

Area=

Inflation

Time

⎟⎠

⎞⎜⎝

⎛ ∑=

τ

022

i

ia

⎟⎟⎠

⎞⎜⎜⎝

⎛∑ ∑

= =

τ

0 012

i

i

j

ia

τ012a 0

22a

τ

Figure 17. Sacrifices Ratio under a SVAR Approach

The estimation of the structural system was done following Blanchard and Quah’s (1989)

methodology. For this reason, the restriction imposed on the system was that aggregate demand shocks do not

have permanent effects on the level output. In particular, the following is imposed ∑∞

=

=0

12 0i

ia .

Figures 18 and 19 show the impulse responses of output and inflation to a monetary policy shock.

The estimation of equation [5], for the Colombia economy, is equal to 8,55. This compute24 assumes a τ equal

to twenty quarters25. Appendix B shows the main steps of the estimation of this SVAR. The economic

interpretation behind this value is that the GDP should fall 8.5 percent due to a reduction of one percent in the

inflation rate.

According to this methodology, a tight monetary policy has had important temporal effects over the

real activity. In particular, this methodology suggests that an important part of the economic recession since

1970 is due to this type of policies. Nevertheless, as Cecchetti and Rich (1999) admitted in their work, this

methodology seems not to be appropriated for estimate a precise sacrifice ratio; in fact they found that the

degree of imprecision increases with the complexity of the model used. Further investigation of this idea for

the Colombian economy may be warranted.

24 It was used quarterly seasonal adjusted data from 1977 to 2001. 25 This value is used by Cecchetti and Rich (1999) in their paper.

20

Figure 18. Dynamic Response of Output to a Monetary Policy Shock

0.000.000.000.000.000.010.010.010.01

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Data Source: Author's calculation.

Figure 19. Dynamic Response of Inflation to a Monetary Policy Shock

0.000.000.000.000.000.010.010.010.01

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Data Source: Author's calculation.

21

6. The Current Disinflation Episode

In this section I will explore some hypothesis on how the current disinflation episode could be

understood. The main purpose of this section is to come up with an answer to the paradox of having the

greatest fall in inflation since 1990 without the usual effects on output that one might expect under the

Keynesian theory26. Finally, I will estimate the sacrifice ratio without taking into account those years where

observe output was above potential output.

The hypothesis is basically as follows: In the beginning of the nineties, almost at the same time as

when the central bank began a tight monetary policy in order to induce a reduction of the inflation rate,

Colombia decided to open its economy (goods & capitals); these two features had two opposite impacts on the

aggregate demand: on one hand, not only national consumption but also national investment began to

increase27 due to the new requirements of the productive system as well as the progressive changed in the

consumers tastes28. On the other hand, the monetary policy had a negative impact in the aggregate demand

given the central bank goal.

These two forces were in opposite directions and their impacts over the aggregate demand (i.e.

output) depended on which one was larger. From the beginning of the nineties to 1997, the positive impact

prevailed over the negative one making the output unaffected by the tight monetary policy29. The increases of

the total consumption as well as total investment was financed with debt, this fact made firms and household

accumulate liabilities, which increases their financial vulnerability30. This situation turned unsustainable in

1998 when the economy began to enter in a recessive phase. Since that year, the negative impact began to be

more significant than the positive impact. In fact, since that year output has been below its potential level31

(see figure 10).

Due to the fact that monetary policy only had a negative impact over the output since 1998 and

because the main assumptions in Ball’s methodology is that recession is the result of a tight monetary policy,

I estimated the sacrifice ratio for the pronounced disinflation period. In order to do this, I use Zhang’s

methodology since it does not require to know when the end of the recession will be. The estimated value is

3.6 using a smoothing parameter of 1.600 or 3.7 using a smoothing parameter of 16.000.

26 In other words, I am concerned with figuring out why the sacrifice ratio in the current disinflation episode is not the most significant one. Moreover, this question becomes truly important when in 1999 Colombia had its worst economic performance in all its history. 27 In fact, in the first middle of the nineties the consumption and investment grew at the average annual rate of 4.3% and 8.4%, meanwhile in the eighties these rates were 3.1% and 3.5% respectively. 28 This was the result of the imported goods which came to Colombia due to the elimination of the barriers of external commerce. One example of this is the complete transformation of the vehicles used in the cities. 29 This period corresponds to what I called “moderate disinflation period” in section two. 30 This explanation is proposed by Echeverry (2001).

22

These values are comparable with those values showed in table 4 for the whole episode. As it was

expected these value are greater than those calculated in section 4. According to this estimation, the tight

monetary policy of the central bank through the decade of nineties has made output fell 3.7% from its natural

level, for each percent of reduction in trend inflation.

31 This period corresponds to what I called “pronounced disinflation period” in section two.

23

7. Concluding Remarks

This paper examines the output cost of disinflation in Colombia for each disinflation episode from

1957 to 2002 following the methodologies proposed by Ball (1994), Zhang (2001), and Cecchetti and Rich

(1999). In order to take into account the sensitivity of the results to the assumption on the construction of

trend output and trend inflation different procedures are used. Finally, the sacrifice ratio of the current

disinflation episode is estimated without taking into account the years where output is above its potential

level.

As it was expected, the sacrifice ratios for Colombia appear to be sensitive to the way the output

trend is estimated meanwhile they are not sensitive to the method used in the construction of inflation trend. It

is important to emphasize the fact that similar sacrifice ratios are obtained with Zhang’s methodology as well

as with different ways used in estimating output trend. In particular, the capacity utilization seems to be a

good variable to obtain the output trend in the disinflation episode. Moreover, estimating the sacrifice ratio

from unemployment data seems to be appropriated for the disinflations of the eighties but not for the

disinflation of the nineties because there is a lot of discussion about the changes of the natural unemployment

rate. Table 9 shows the results of these approaches. The result of using the Cecchetti and Rich’s methodology

appears somewhat higher than the previous findings.

Table 9. Sacrifice Ratios in each disinflation episode in the Colombian economy.

A Q A Q

Ball (GDP) 0.25 -0.02 0.17 0.22 -0.12 0.21 0.92

Zhang (GDP) 0.25 -0.33 -0.33 0.63 -0.76 1.83 2.11 0.921 0.891

Zhang (U) 3.42 3.34

Capacity utilization 0.66 2.81 0.16 0.74

Production Function -0.69 -0.72 0.76 1.65 0.50 2.04

One Side HP -0.20 -0.10 1.29 -0.49 2.71 8.17 1.39 3.6

9 quarters 0.25 -0.02 0.17 0.22 -0.12 0.21

5 quarters 0.20 -0.01 0.13 0.19 -0.05 0.19

4quarters -0.01 0.12 0.17

Mean2 0.24 -0.10 0.03 0.31 -0.38 0.84 3.13 0.68 2.13 Standard Desviation2 0.03 0.14 0.20 0.65 0.32 0.94 2.90 0.64 1.43

1. Using a smooth parameter of the HP filter of 16002. Zhang (U) is not incluided.Note: The interpretation of this variables is what percent the current output has to fell (if it has postive sign) from its long run level, due to a reduction of 1% in trend Inflation

1991 - 20011957 - 1961 1963 - 1965 1966 - 1968 1974 - 1975 1977 - 1978 1980 - 1984

Cha

nges

ove

r O

utpu

t Gap

Stan

dard

M

etho

dolo

gies

Cha

nges

ove

r In

flatio

n Tr

end

24

The results show that the current disinflation episode has not been the most important one due to its

particular characteristics. Although, in this episode the output fell to its worst level and the decline in the

inflation rate was remarkable, at the beginning of the episode the output was above its natural level for a long

time. In this paper, it is argued that this happened because there were two economic forces playing opposites

roles in the Colombian economic performance in the nineties: the first one is the inflow of capital to the

economy and the change in consumers tastes, meanwhile the second one is a tight monetary policy conducted

by the central bank. This situation affected the true sacrifices ratio during this episode, in the spirit of Ball ‘s

(1994) methodology. To overcome this situation the sacrifice ratio had to be computed only for those years

when the monetary policy had effects on the economic activity. In this case, the sacrifices ratio was higher

than for any other disinflation episode that Colombia has experienced since 1950.

In summary, the objective of this paper was to shed some light on the methodological puzzle of

having an acquired measure of the sacrifice ratio in each disinflation episode in the Colombian economy. This

issue is crucial for policymakers, since this tool would enable them to assess the consequences of previous

and current monetary policies. Nevertheless, there is a wide area to be explored in this topic. What the long-

run benefit analysis of disinflation is and what the role of fiscal policy should be in each disinflation episode,

are issues which further research is warranted.

25

8. References

Ball, L. (1994). “What Determines the Sacrifice Ratio?” in Mankiw G. (ed.), “Monetary Policy”, Chicago,

National Bureau of Economic Research.

Birchenall, J. A. (1999). “La Curva de Phillips, la crítica de Lucas y la Persistencia de la Inflación en

Colombia”. Archivos de Macroeconomía No. 102. Departamento Nacional de Planeación.

Blanchard, O. J. and D. Quah. (1989). “ The Dynamic Effects of Aggregate Demand and Supply

Disturbances” The American Economic Review. Vol. 79, No.4

Cecchetti, S. G and R. W. Rich. (1999). “ Structural Estimates of the U.S. Sacrifice Ratio”, Research and

Market Analysis Group. Federal Reserve Bank of New York.

Clavijo, S. (1994). “Inflación o Desempleo: ¿Acaso hay Escogencia en Colombia?”. Archivos de

Macroeconomía No. 30. Departamento Nacional de Planeación.

Echeverry, J. (2001). “Memorias de la Recesión de Fin de Siglo en Colombia: Flujos, Balances y Política

Anticíclica”. Boletines de Divulgación Económica No. 7. Departamento Nacional de Planeación.

Henao, M. and N. Rojas. (1998). “La Tasa Natural de Desempleo en Colombia”, Archivos de

Macroeconomía No. 89. Departamento Nacional de Planeación.

Gómez, J. (2002). “Wage Indexation, Inflation Inertia, and the Cost of Disinflation”. Borradores Semanales

de Economía, No. 198. Banco de la República.

Gómez, J. and J. M. Julio. (2000). “Transmission Mechanisms and Inflation targeting: the case of

Colombia’s Disinflation”. Borradores Semanales de Economía No. 168. Banco de la República.

Gordon, R. J. and S. R. King. (1982). “ The output cost of disinflation in traditional and vector

autoregressive models. Brooking papers on Economic Activity 1: 205-42.

Guataquí, J. C.. (2000), “Estimaciones de la tasa natural de desempleo en Colombia. Una revision”.

Borradores de investigación, No. 2. Universidad del Rosario.

Mankiw, N. G. (1991), “Macroeconomics”. New York: Worth Publishers.

26

Misas, M. and E. López. (1999). “Un Examen Empírico de la Curva de Phillips en Colombia”. Borradores

Semanales de Economía No. 117. Banco de la República.

Núñez, J. and R. Bernal. (1998). “El Desempleo en Colombia: Tasa Natural, Desempleo Cíclico y

Estructural y la Duración del Desempleo, 1976 – 1998”. Archivos de Macroeconomía No. 97. Departamento

Nacional de Planeación.

Okun, A. M. (1978). “Efficient disinflationary policies”, American Economic Review 68 (May): 348 –52.

Pantula, S. G. (1989). “Testing for Unit Roots in Time Series Data”. Econometric Theory, 5, 256-271.

Partow, Z. (1995). “Una Revisión de la Literatura sobre los Costos de la Inflación”, Borradores Semanales de

Economía No. 032. Banco de la República.

Reyes, J. (2001). “Evidencia de Histéresis en el Desempleo en Colombia (1977 – 2000)”. Master thesis,

Universidad de los Andes.

Reyes, J. (2003). “PIB Potencial bajo una Función de Producción tipo CES”, mimeo, Departamento Nacional

de Planeación.

Sánchez, F. and J. Nuñez. (1998). “La Curva de Salarios para Colombia. Una Estimación de las Relaciones

entre el Desempleo, la Inflación y los Ingresos Laborales, 1984 – 1996”. Archivos de Macroeconomía No. 80.

Departamento Nacional de Planeación.

Sánchez, F., L.M. Salas and O. A. Nupia (2003). “ Histéresis en el Desempleo Colombiano”, mimeo,

Departamento Nacional de Planeación.

Sarmiento, E., A. Pontón and C. Cardona (1998). “ Evidencia sobre desinflaciones: Experiencia

Internacional”, Borradores Semanales de Economía No. 102. Banco de la República.

Zhang, L. (2001). “ Sacrifice Ratio with Long – Lived Effects”. Working Paper 446, The Johns Hopkins

University.

27

APPENDIX A

Potential Output from a CES production function.

This exercise incorporates an estimate of an aggregate production function for the Colombian

economy. GDP is produced by two inputs: Labor (L) and capital (K)32. The specific functional form of the

production function is shown in equation [A1].

[ ] ρηρρ αα/)1( −−− −+= ttt LKAY [A1]

Where 0<α<1 is the share parameter and ρ>−1 determines the degree of substitutability of the inputs.

A is the factorial productivity parameter and η>0 is the returns to scale parameter. This function can be

transformed either into a Cobb-Douglas function, a function of substitutes goods or a function of

complementary goods depending on ρ 33.

Due to the functional form [A1], the estimation uses non linear least squares34. Since it is more useful

to treat the productivity as a time varying parameter instead of time invariant one, it is obtained as a residual

of the production function.

Finally, it is necessary to built the potential level of inputs (L and K) in order to put them into the

production function and to estimate potential output. For the stock of capital, it was assumed that the natural

capacity utilization corresponds to the average of the data that is above its historical mean level. On the other

hand, the natural level of labor was estimated using the natural rate of unemployment over the population

who are willing to work. In this context, the natural rate of unemployment corresponds to the average of the

data that is below its historical level35.

The potential output is estimated from the CES production function. For further details and results

see Reyes (2003).

32 The capital stock was corrected by capacity utilization indicator from Fedesarrollo. 33 When 1−→ρ is the case of perfect substitution, when ∞→ρ is the case of no substitution, and when

0=ρ is the case of unit elasticity of substitution (Cobb-Douglas case).

34 The estimated equation was: tttt LKAY εααρη ρρ +−+−= −− ])1(ln[lnln .

35 Due to data availability, this exercises was done only for quarterly data. For annual data it was assumed that the natural rate of unemployment was the same that for quarterly data.

28

APPENDIX B

Drift Trend Unit root Lags Autocorrelation Conclusion Drift Trend Unit

root Lags Autocorrelation Conclusion

Model 31.26 -1.73

(-2.8)-1.09

(-3.45)4 No

Model 30.67 -1.34

(-2.8)-6.87(-3.4)

3 No I(0)

Model 2-0.13(2.54)

-0.14(-2.5)

4 NoModel 2

Model 1-1.14

(-1.95)4 No I(1)

Model 1

Drift Trend Unit root Lags Autocorrelation Conclusion Drift Trend Unit

root Lags Autocorrelation Conclusion

Model 31.22 1.05

(2.8)-1.21

(-3.45)4 No

Model 32.61 -1.1

(-2.79)-4.07

(-3.45)5 No I(0)

Model 21.14

(2.54)-1.08(-2.8)

4 No I(1)Model 2

Model 1*2.72(-1.9)

4 NoModel 1

I used the procedure suggested by Doldado et al (1990).The critic value at 5% is in parenthesis.

* In the Dickey-Fuller test it is mandatory that the parameter of unit root to be less or equal to cero. When the estimation is greater that zero it is necesary to use another test. In this case I use the second model, even though the drift is no estadistically different form zero

Augmented Dickey - Fuller t-statistic[Inflation]

Augmented Dickey - Fuller t-statistic[D(Inflation)]

Augmented Dickey - Fuller t-statistic[Output]

Augmented Dickey - Fuller t-statistic[D(Output)]

Level Trend conclusion

Inflation 0.54 I(1)D (Inflation) 0.21 I(0)

Output 0.16 I(1)D (Output) 0.10 I(0)

Critical values*.

10% 5% 1%Nivel 0.35 0.46 0.74Tendencia 0.12 0.15 0.22* These critical values do not depend on the number of data.

KPSS test

29

Lags Ho HaTrace

t-statistic*Critical

Value (5%)Trace

t-statistic*

Critical Value (5%)

Trace t-statistic*

Critical Value (5%)

r=0 r>=1 21.78 19.96 5.00 15.41 14.94 25.32r=1 r=2 3.31 9.24 1.56 3.76 3.06 12.25

r=0 r>=1 24.88 19.96 8.44 15.41 19.22 25.32r=1 r=2 6.48 9.24 -2.00 3.76 3.55 12.25

r=0 r>=1 32.90 19.96 8.39 15.41 15.88 25.32r=1 r=2 6.54 9.24 1.85 3.76 2.71 12.25

r=0 r>=1 38.12 19.96 11.28 15.41 16.50 25.32r=1 r=2 5.54 9.24 3.48 3.76 3.50 12.25

r=0 r>=1 17.12 19.96 4.04 15.41 8.00 25.32r=1 r=2 2.62 9.24 1.03 3.76 2.88 12.25

* This statistic was adjusted for small sample according to Cheung and Lai (1993).Note: I followed the so-called Pantula principle (Pantula (1989)).

4

5

Johansen procedure for cointegration analysis -Determination of the cointegration rank-

Cidrift Model

1

2

3

Cimean Model Drift Model

Lags Akaike (aic)

Schwarz (sc)

Hanna-Quinn(hq)

0 -19.4829 -19.4829 -19.4829

1 -19.4453 -19.335 -19.4008

2 -19.4369 -19.2162 -19.3479

3 -19.4789 -19.1478 -19.3453

4 -19.9758 -19.5343 -19.7977

5 -19.9374 -19.3856 -19.7148

6 -19.9824 -19.3202 -19.7152

Model selection criteria for the VAR.

Note: For each of the lags I used the same sample period in order to make them comparable.

Lags Pormanteau Test

Adjusted Pormanteau

Test

Lambda 1(Kurtosis)

Lambda 2(skewness)

Lambda 3(Join)

0 209.79792(0.0000)

241.40843(0.0000)

9.49090(0.00869)

28.40368(0.0000)

37.89459(0.0000)

1 223.43639(0.0000)

254.56886(0.0000)

4.57165(0.10169)

30.34441(0.0000)

34.91606(0.0000)

2 150.12655(0.00002)

172.42386(0.0000)

4.41119(0.11018)

39.23275(0.0000)

43.64394(0.0000)

3 127.89375(0.0007)

147.58335(0.00001)

3.09765(0.21250)

28.90988(0.0000)

32.00753(0.0000)

4 79.58155(0.36694)

92.85799(0.09163)

0.05465(0.97305)

9.67836(0.00791)

9.73301(0.04517)

579.16627(0.23696)

91.55104(0.05088)

0.30663(0.85786)

2.06640(0.35587)

2.37302(0.66751)

6 63.33090(0.57035)

73.63052(0.24269)

0.29665(0.86215)

7.03695(0.02964)

7.33360(0.11927)

P-value in parenthesis.

Autocorrelation Multivariate Normality

Error Diagnostic

C DLY(-1) DLY(-2) DLY(-3) DLY(-4) DLY(-5) DINF(-1) DINF(-2) DINF(-3) DINF(-4) DINF(-5)

DLY 0.0041 -0.1701 -0.1141 -0.0046 0.5016 0.1542 -0.1898 0.0713 -0.1003 0.0541 -0.0178(-1.4951) (-1.5421) (-1.1594) (-0.0450) (5.0142) (1.4010) (-1.8159) (0.7834) (-1.1643) (0.60739) (-0.1957)

DINF -0.0093 0.0683 0.2602 0.2581 0.2022 0.1895 0.0081 -0.0958 0.2052 -0.4809 -0.0912(-3.338) (0.6124) (2.6154) (2.48126) (1.999) (1.7027) (0.0764) (-1.0408) (2.3568) (-5.3441) (-0.9915)

DLY: First difference of Output. DINF: First Difference of Inflation.Note: t-statistics in parenthesis.

Var Estimates

ARCHIVOS DE ECONOMIA No Título Autores Fecha 1 La coyuntura económica en Colombia y Venezuela Andrés LANGEBAEK Octubre 1992 Patricia DELGADO Fernando Mesa Parra 2 La tasa de cambio y el comercio colombo-venezolano Fernando Mesa Parra Noviembre 1992 Andrés LANGEBAEK 3 ¿Las mayores exportaciones colombianas de café redujeron Carlos Esteban POSADA Noviembre 1992 el precio externo? Andrés LANGEBAEK 4 El déficit público: una perspectiva macroeconómica. Jorge Enrique RESTREPO Noviembre 1992 Juan Pablo ZÁRATE Carlos Esteban POSADA 5 El costo de uso del capital en Colombia. Mauricio OLIVERA Diciembre 1992 6 Colombia y los flujos de capital privado a América Latina Andrés LANGEBAEK Febrero 1993 7 Infraestructura física. “Clubs de convergencia” y crecimiento José Dario URIBE Febrero 1993 económico. 8 El costo de uso del capital: una nueva estimación (Revisión) Mauricio OLIVERA Marzo 1993 9 Dos modelos de transporte de carga por carretera. Carlos Esteban POSADA Marzo 1993 Edgar TRUJILLO CIRO Alvaro CONCHA Juan Carlos ELORZA 10 La determinación del precio interno del café en un modelo Carlos Felipe JARAMILLO Abril 1993 de optimización intertemporal. Carlos Esteban POSADA Edgar TRUJILLO CIRO 11 El encaje óptimo Edgar TRUJILLO CIRO Mayo 1993 Carlos Esteban POSADA 12 Crecimiento económico, “Capital Carlos Esteban POSADA Junio 1993 humano” y educación: la teoría y el caso colombiano posterior a 1945 13 Estimación del PIB trimestral según los componentes del gasto. Rafael CUBILLOS Junio 1993 Fanny Mercedes VALDERRAMA 14 Diferencial de tasas de interés y flujos de capital en Colombia Andrés LANGEBAEK Agosto 1993 (1980-1993) 15 Empleo y capital en Colombia: nuevas Adriana BARRIOS Septiembre 1993 estimaciones (1950-1992) Marta Luz HENAO Carlos Esteban POSADA Fanny Mercedes VALDERRAMA Diego Mauricio VÁSQUEZ 16 Productividad, crecimiento y ciclos en la economía Carlos Esteban POSADA Septiembre 1993 colombiana (1967-1992) 17 Crecimiento económico y apertura en Chile y México y Fernando MESA PARRA Septiembre 1993 perspectivas para Colombia. 18 El papel del capital público en la producción, inversión y Fabio SÁNCHEZ TORRES Octubre 1993 el crecimiento económico en Colombia. 19 Tasa de cambio real y tasa de cambio de equilibrio. Andrés LANGEBAEK Octubre 1993 20 La evolución económica reciente: dos interpretaciones Carlos Esteban POSADA Noviembre 1993

alternativas. 21 El papel de gasto público y su financiación en la coyuntura Alvaro ZARTA AVILA Diciembre 1993 actual: algunas implicaciones complementarias. 22 Inversión extranjera y crecimiento económico. Alejandro GAVIRIA Diciembre 1993

ARCHIVOS DE ECONOMIA No Título Autores Fecha Javier Alberto GUTIÉRREZ 23 Inflación y crecimiento en Colombia Alejandro GAVIRIA Febrero 1994 Carlos Esteban POSADA 24 Exportaciones y crecimiento en Colombia Fernando MESA PARRA Febrero 1994 25 Experimento con la vieja y la nueva teoría del crecimiento Carlos Esteban POSADA Febrero 1994 económico (¿porqué crece tan rápido China?) 26 Modelos económicos de criminalidad y la posibilidad de Carlos Esteban POSADA Abril 1994 una dinámica prolongada. 27 Regímenes cambiarios, política macroeconómica y flujos Carlos Esteban POSADA Abril 1994 de capital en Colombia. 28 Comercio intraindustrial: el caso colombiano Carlos POMBO Abril 1994 29 Efectos de una bonanza petrolera a la luz de un modelo Hernando ZULETA Mayo 1994 de optimización intertemporal. Juan Pablo ARANGO 30 Crecimiento económico y productividad en Colombia: Sergio CLAVIJO Junio 1994 . una perspectiva de largo plazo (1957-1994) 31 Inflación o desempleo: ¿Acaso hay escogencia en Colombia? Sergio CLAVIJO Agosto 1994 32 La distribución del ingreso y el sistema financiero Edgar TRUJILLO CIRO Agosto 1994 33 La trinidad económica imposible en Colombia: estabilidad Sergio CLAVIJO Agosto 1994

cambiaria, independencia monetaria y flujos de capital libres

34 ¿’Déjà vu?: tasa de cambio, deuda externa y esfuerzo Sergio CLAVIJO Mayo 1995 exportador en Colombia. 35 La crítica de Lucas y la inversión en Colombia: Mauricio CÁRDENAS Septiembre 1995 nueva evidencia Mauricio OLIVERA 36 Tasa de Cambio y ajuste del sector externo en Colombia. Fernando MESA PARRA Septiembre 1995 Dairo ESTRADA 37 Análisis de la evolución y composición del Sector Público. Mauricio Olivera G. Septiembre 1995 Manuel Fernando Castro Q. Fabio Sánchez T. 38 Incidencia distributiva del IVA en un modelo del ciclo de vida. Juan Carlos PARRA OSORIO Octubre 1995 Fabio José SÁNCHEZ T. 39 Por qué los niños pobres no van a la escuela? Fabio SÁNCHEZ TORRES Noviembre 1995 (Determinantes de la asistencia escolar en Colombia) Jairo Augusto NÚÑEZ M. 40 Matriz de Contabilidad Social 1992. Fanny M. VALDERRAMA Diciembre 1995 Javier Alberto GUTIÉRREZ 41 Multiplicadores de Contabilidad derivados de la Matriz Javier Alberto GUTIÉRREZ Enero 1996 de Contabilidad Social Fanny M. VALDERRAMA G. 42 El ciclo de referencia de la economía colombiana. Martin MAURER Febrero 1996 María Camila URIBE S. 43 Impacto de las transferencias intergubernamentales en la Juan Carlos PARRA OSORIO Marzo 1996 distribución interpersonal del ingreso en Colombia. 44 Auge y colapso del ahorro empresarial en Colombia: Fabio SÁNCHEZ TORRES Abril 1996 1983-1994 Guillermo MURCIA GUZMÁN Carlos OLIVA NEIRA 45 Evolución y comportamiento del gasto público en Colombia: Cielo María NUMPAQUE Mayo 1996 1950-1994 Ligia RODRÍGUEZ CUESTAS

ARCHIVOS DE ECONOMIA No Título Autores Fecha 46 Los efectos no considerados de la apertura económica en el Fernando MESA PARRA Mayo 1996 mercado laboral industrial. Javier Alberto GUTIÉRREZ 47 Un modelo de Financiamiento óptimo de un aumento Alvaro ZARTA AVILA Junio 1996

permanente en el gasto público: Una ilustración con el caso colombiano.

48 Estadísticas descriptivas del mercado laboral masculino y Rocío RIBERO M. Agosto 1996 femenino en Colombia: 1976 -1995 Carmen Juliana GARCÍA B. 49 Un sistema de indicadores líderes para Colombia Martín MAURER Agosto 1996 María Camila URIBE Javier BIRCHENALL 50 Evolución y determinantes de la productividad en Colombia: Fabio SÁNCHEZ TORRES Agosto 1996 Un análisis global y sectorial Jorge Iván RODRÍGUEZ Jairo NÚÑEZ MÉNDEZ 51 Gobernabilidad y Finanzas Públicas en Colombia. César A. CABALLERO R Noviembre 1996 52 Tasas Marginales Efectivas de Tributación en Colombia. Mauricio OLIVERA G. Noviembre 1996 53 Un modelo keynesiano para la economía colombiana Fabio José SÁNCHEZ T. Febrero 1997 Clara Elena PARRA 54 Trimestralización del Producto Interno Bruto por el lado Fanny M. VALDERRAMA Febrero 1997 de la oferta. 55 Poder de mercado, economías de escala, complementariedades Juán Mauricio RAMÍREZ Marzo 1997

intersectoriales y crecimiento de la productividad en la industria colombiana.

56 Estimación y calibración de sistemas flexibles de gasto. Jesús Orlando GRACIA Abril 1997 Gustavo HERNÁNDEZ 57 Mecanismos de ahorro e Inversión en las Empresas Públicas Fabio SÁNCHEZ TORRES Mayo 1997 Colombianas: 1985-1994 Guilllermo MURCIA G. 58 Capital Flows, Savings and investment in Colombia: 1990-1996 José Antonio OCAMPO G. Mayo 1997 Camilo Ernesto TOVAR M.

59 Un Modelo de Equilibrio General Computable con Juan Pablo ARANGO Junio 1997 Competencia imperfecta para Colombia. Jesús Orlando GRACIA Gustavo HERNÁNDEZ Juan Mauricio RAMÍREZ 60 El cálculo del PIB Potencial en Colombia. Javier A. BIRCHENALL J. Julio 1997 61 Determinantes del Ahorro de los hogares. Explicación Alberto CASTAÑEDA C. Julio 1997 de su caída en los noventa. Gabriel PIRAQUIVE G. 62 Los ingresos laborales de hombres y Rocío RIBERO Agosto 1997

mujeres en Colombia: 1976-1995 Claudia MEZA 63 Determinantes de la participación laboral de hombres y Rocío RIBERO Agosto 1997

mujeres en Colombia: 1976-1995 Claudia MEZA

64 Inversión bajo incertidumbre en la Industria Colombiana: Javier A. BIRCHENALL Agosto 1997 1985-1995

65 Modelo IS-LM para Colombia. Relaciones de largo plazo y Jorge Enrique RESTREPO Agosto 1997 fluctuaciones económicas.

66 Correcciones a los Ingresos de las Encuestas de hogares y Jairo A. NÚÑEZ MÉNDEZ Septiembre 1997 distribución del Ingreso Urbano en Colombia. Jaime A. JIMÉNEZ CASTRO

67 Ahorro, Inversión y Transferencias en las Entidades Fabio SÁNCHEZ TORRES Octubre 1997 Territoriales Colombianas Mauricio OLIVERA G. Giovanni CORTÉS S.

ARCHIVOS DE ECONOMIA No Título Autores Fecha 68 Efectos de la Tasa de cambio real sobre la Inversión Fernando MESA PARRA Octubre 1997 industrial en un Modelo de transferencia de precios. Leyla Marcela SALGUERO Fabio SÁNCHEZ TORRES 69 Convergencia Regional: Una revisión del caso Javier A. BIRCHENALL Octubre 1997 Colombiano. Guillermo E. MURCIA G. 70 Income distribution, human capital and economic Javier A. BIRCHENALL Octubre 1997 growth in Colombia. 71 Evolución y determinantes del Ahorro del Gobierno Central. Fabio SÁNCHEZ TORRES Noviembre 1997 Ma. Victoria ANGULO 72 Macroeconomic Perforrmance and Inequality in Colombia: Raquel BERNAL Diciembre 1997 1976-1996 Mauricio CÁRDENAS Jairo NÚÑEZ MÉNDEZ

Fabio SÁNCHEZ TORRES

73 Liberación comercial y salarios en Colombia: 1976-1994 Donald ROBBINS Enero 1998

74 Educación y salarios relativos en Colombia: 1976-1995 Jairo NÚÑEZ MÉNDEZ Enero 1998

Determinantes, evolución e implicaciones para Fabio SÁNCHEZ TORRES la distribución del Ingreso

75 La tasa de interés “óptima” Carlos Esteban POSADA Febrero 1998 Edgar TRUJILLO CIRO 76 Los costos económicos de la criminalidad y la violencia en Edgar TRUJILLO CIRO Marzo 1998 Colombia: 1991-1996 Martha Elena BADEL 77 Elasticidades Precio y Sustitución para la Industria Juán Pablo ARANGO Marzo 1998 Colombiana. Jesús Orlando GRACIA Gustavo HERNÁNDEZ 78 Flujos Internacionales de Capital en Colombia: Ricardo ROCHA GARCÍA Marzo 1998 Un enfoque de Portafolio Fernando MESA PARRA 79 Macroeconomía, ajuste estructural y equidad en Colombia: José Antonio OCAMPO Marzo 1998 1978-1996 María José PÉREZ Camilo Ernesto TOVAR Francisco Javier LASSO 80 La Curva de Salarios para Colombia. Una Estimación de las Fabio SÁNCHEZ TORRES Marzo 1998 Relaciones entre el Desempleo, la Inflación y los Ingresos Jairo NÚÑEZ MÉNDEZ Laborales: 1984- 1996. 81 Participación, Desempleo y Mercados Laborales en Colombia. Jaime TENJO G. Abril 1998 Rocio RIBERO M. 82 Reformas comerciales, márgenes de beneficio y Juán Pablo ARANGO Abril 1998 productividad en la industria colombiana Jesús Orlando GRACIA Gustavo HERNÁNDEZ Juán Mauricio RAMÍREZ 83 Capital y Crecimiento Económico en un Modelo Alvaro ZARTA AVILA Mayo 1998.

Dinámico: Una presentación de la dinámica Transicional para los casos de EEUU y Colombia

84 Determinantes de la Inversión en Colombia: E videncia sobre Clara Helena PARRA Junio 1998. el capital humano y la violencia.

85 Mujeres en sus casas: Un recuento de la población Piedad URDINOLA C. Junio 1998. Femenina económicamente activa

86 Descomposición de la desigualdad del Ingreso laboral Fabio SÁNCHEZ TORRES Junio 1998. Urbano en Colombia: 1976-1997 Jairo NÚÑEZ MÉNDEZ

ARCHIVOS DE ECONOMIA No Título Autores Fecha 87 El tamaño del Estado Colombiano Indicadores y tendencias: Angela CORDI GALAT Junio 1998.

1976-1997

88 Elasticidades de sustitución de las importaciones para la Gustavo HERNÁNDEZ Junio 1998. economía colombiana.

89 La tasa natural de desempleo en Colombia Martha Luz HENAO Junio 1998. Norberto ROJAS

90 The role of shocks in the colombian economy Ana María MENÉNDEZ Julio 1998. 91 The determinants of Human Capital Accumulation in Donald J. ROBBINS Julio 1998.

Colombia, with implications for Trade and Growth Theory 92 Estimaciones de funciones de demanda de trabajo dinámicas Alejandro VIVAS BENÍTEZ Julio 1998.

para la economía colombiana, 1980-1996 Stefano FARNÉ Dagoberto URBANO

93 Análisis de las relaciones entre violencia y equidad. Alfredo SARMIENTO Agosto 1998. Lida Marina BECERRA 94 Evaluación teórica y empírica de las exportaciones Fernando MESA PARRA Agosto 1998. no tradicionales en Colombia María Isabel COCK

Angela Patricia JIMÉNEZ 95 Valoración económica del empleo doméstico femenino Piedad URDINOLA C. Agosto 1998.

no remunerado, en Colombia, 1978-1993

96 Eficiencia en el Gasto Público de Educación. María Camila URIBE Agosto 1998. 97 El desempleo en Colombia: tasa natural, desempleo cíclico Jairo NÚÑEZ M. Septiembre 1998. y estructural y la duración del desempleo: 1976-1998. Raquel BERNAL S. 98 Productividad y retornos sociales del Capital humano: Francisco A. GONZÁLEZ R. Noviembre 1998. Microfundamentos y evidencia para Colombia. Carolina GUZMÁN RUIZ Angela L. PACHÓN G. 99 Reglas monetarias en Colombia y Chile Jorge E. RESTREPO L. Enero 1999. 100 Inflation Target Zone: The Case of Colombia: 1973-1994 Jorge E. RESTREPO L. Febrero 1999. 101 ¿ Es creíble la Política Cambiaria en Colombia? Carolina HOYOS V. Marzo 1999. 102 La Curva de Phillips, la Crítica de Lucas y la persistencia Javier A.BIRCHENALL Abril 1999.

de la inflación en Colombia.

103 Un modelo macroeconométrico para la economía Javier A.BIRCHENALL Abril 1999. Colombiana Juan Daniel OVIEDO

104 Una revisión de la literatura teórica y la experiencia Marcela ESLAVA MEJÍA Abril 1999. Internacional en regulación

105 El transporte terrestre de carga en Colombia Marcela ESLAVA MEJÍA Abril 1999. Documento para el Taller de Regulación. Eleonora LOZANO RODRÍGUEZ

106 Notas de Economía Monetaria. (Primera Parte) Juan Carlos ECHEVERRY G. Abril 1999. 107 Ejercicios de Causalidad y Exogeneidad para Ingresos Mauricio BUSSOLO Mayo 1999.

salariales nominales públicos y privados Colombianos Jesús Orlando GRACIA (1976-1997). Camilo ZEA

108 Real Exchange Rate Swings and Export Behavior: Felipe ILLANES Mayo 1999. Explaining the Robustness of Chilean Exports. 109 Segregación laboral en las 7 principales ciudades del país. Piedad URDINOLA Mayo 1999. 110 Estimaciones trimestrales de la línea de pobreza y sus relaciones Jairo NÚÑEZ MÉNDEZ Mayo 1999 con el desempeño macroeconómico Colombiano: (1977-1997) Fabio José SÁNCHEZ T.

ARCHIVOS DE ECONOMIA No Título Autores Fecha 111 Costos de la corrupción en Colombia. Marta Elena BADEL Mayo 1999 112 Relevancia de la dinámica transicional para el Alvaro ZARTA Junio 1999

crecimiento de largo plazo: Efectos sobre las tasas de interés real, la productividad marginal y la estructura de la producción para los casos de EEUU y Colombia..

113 La recesión actual en Colombia: Flujos, Balances y Juan Carlos ECHEVERRY Junio 1999 Política anticíclica

114 Monetary Rules in a Small Open Economy Jorge E. RESTREPO L. Junio 1999

115 El Balance del Sector Público y la Sostenibilidad Juan Carlos ECHEVERRY Junio 1999 Fiscal en Colombia Gabriel PIRAQUIVE Natalia SALAZAR FERRO Ma. Victoria ANGULO Gustavo HERNÁNDEZ Cielo Ma. NUMPAQUE Israel FAINBOIM Carlos Jorge RODRIGUEZ 116 Crisis y recuperación de las Finanzas Públicas lecciones Marcela ESLAVA MEJÍA Julio 1999 de América Latina para el caso colombiano.

117 Complementariedades Factoriales y Cambio Técnico Gustavo HERNÁNDEZ Julio 1999

en la Industria Colombiana. Juan Mauricio RAMÍREZ 118 ¿Hay un estancamiento en la oferta de crédito? Juan Carlos ECHEVERRY Julio 1999

Natalia SALAZAR FERRO 119 Income distribution and macroeconomics in Colombia. Javier A. Birchenall J. Julio 1999. 120 Transporte carretero de carga. Taller de regulación. Juan Carlos ECHEVERRY G. Agosto 1999.

DNP-UMACRO. Informe final. Marcela ESLAVA MEJÍA Eleonora LOZANO RODRIGUEZ

121 ¿ Se cumplen las verdades nacionales a nivel regional? Nelly.Angela CORDI GALAT Agosto 1999.

Primera aproximación a la construcción de matrices de contabilidad social regionales en Colombia.

122 El capital social en Colombia. John SUDARSKY Octubre 1999. La medición nacional con el BARCAS Separata N° 1 de 5

123 El capital social en Colombia. John SUDARSKY Octubre 1999. La medición nacional con el BARCAS Separata N° 2 de 5

124 El capital social en Colombia. John SUDARSKY Octubre 1999. La medición nacional con el BARCAS Separata N° 3 de 5

125 El capital social en Colombia. John SUDARSKY Octubre 1999.

La medición nacional con el BARCAS Separata N° 4 de 5

126 El capital social en Colombia. John SUDARSKY Octubre 1999.

La medición nacional con el BARCAS Separata N° 5 de 5

127 The Liquidity Effect in Colombia Jorge Enrique RESTREPO Noviembre 1999.

128 Upac: Evolución y crisis de un modelo de desarrollo. Juan Carlos ECHEVERRI G. Diciembre 1999. Orlando Gracia B. Piedad Urdinola 129 Confronting fiscal imbalances via intertemporal Juan Carlos ECHEVERRY G. Diciembre 1999.

Economics, politics and justice: the case of Colombia Verónica Navas-Ospina

ARCHIVOS DE ECONOMIA No Título Autores Fecha 130 La tasa de interés en la coyuntura reciente en Colombia. Jorge Enrique RESTREPO Diciembre 1999.

Edgar TRUJILLO CIRO

131 Los ciclos económicos en Colombia. Evidencia empírica: Jorge Enrique RESTREPO Enero 2000. (1977-1998) José Daniel REYES PEÑa

132 Colombia'natural trade partners and its bilateral trade Hernán Eduardo VALLEJO Enero 2000. performance: Evidence from 1960 to 1996

133 Los derechos constitucionales de prestación y sus Luis Carlos SOTELO Febrero 2000. implicaciones económico- políticas. Los casos del derecho a la salud y de los derechos de los reclusos

134 La reactivación productiva del sector privado colombiano Luis Alberto ZULETA Marzo 2000. (Documento elaborado para el BID)

135 Geography and Economic Development: Fabio JOSÉ SÁNCHEZ T. Marzo 2000. A Municipal Approach for Colombia. Jairo NÚÑEZ MÉNDEZ

136 La evaluación de resultados en la modernización Eduardo WIESNER DURÁN Abril 2000. del Estado en América Latina. Restricciones y Estrategia para su desarrollo.

137 La regulación de precios del transporte de carga por Marcela ESLAVA MEJÍA Abril 2000. carretera en Colombia.

138 El conflicto armado en Colombia. Yuri GORBANEFF Julio 2000. Una aproximación a la teoría de juegos. Flavio JÁCOME

139 Determinación del consumo básico de agua potable Juan Carlos JUNCA SALAS Noviembre 2000. subsidiable en Colombia.

. 140 Incidencia fiscal de los incentivos tributarios Juan Ricardo ORTEGA Noviembre 2000.

Gabriel Armando PIRAQUIVE Gustavo Adolfo HERNÁNDEZ Carolina SOTO LOSADA Sergio Iván PRADA Juan Mauricio RAMIREZ

141 Exenciones tributarias: Gustavo A. HERNÁNDEZ Diciembre 2000 Costo fiscal y análisis de incidencia Carolina SOTO LOSADA

Sergio Iván PRADA Juan Mauricio RAMIREZ

142 La contabilidad del crecimiento, las dinámicas transicionales y Alvaro ZARTA AVILA Febrero 2001 el largo plazo: Una comparación internacional de 46 países y una presentación de casos de economías tipo: EEUU, Corea del Sur y Colombia.

143 ¿Nos parecemos al resto del mundo? Juan Carlos ECHEVERRY G. Febrero 2001

El Conflicto colombiano en el contexto internacional. Natalia SALAZAR FERRO Verónica NAVAS OSPINA

144 Inconstitucionalidad del Plan Nacional de Desarrollo: Luis Edmundo SUÁREZ S. Marzo 2001 causas, efectos y alternativas. Diego Mauricio AVILA A. 145 La afiliación a la salud y los efectos redistributivos Hernando MORENO G. Abril 2001 de los subsidios a la demanda. 146 La participación laboral: ¿qué ha pasado y qué Mauricio SANTA MARÍA S. Abril 2001 podemos esperar? Norberto ROJAS DELGADILLO 147 Análisis de las importaciones agropecuarias en la Gustavo HERNÁNDEZ Mayo 2001 década de los Noventa. Juan Ricardo PERILLA 148 Impacto económico del programa de Desarrollo Gustavo A. HERNÁNDEZ Mayo 2001

alternativo del Plan Colombia Sergio Iván PRADA Juan Mauricio RAMÍREZ

ARCHIVOS DE ECONOMIA No Título Autores Fecha

149 Análisis de la presupuestación de la inversión de la Nación. Ulpiano AYALA ORAMAS Mayo 2001 150 DNPENSION: Un modelo de simulación para estimar Juan Carlos PARRA OSORIO Mayo 2001 el costo fiscal del sistema pensional colombiano. 151 La oferta de combustible de Venezuela en la frontera Hernando MORENO G. Junio 2001 con Colombia: una aproximación a su cuantificación 152 Shocks fiscales y términos de intercambio en el caso Ómer ÖZAK MUñOZ. Julio 2001 colombiano. 153 Demanda por importaciones en Colombia: Igor Esteban ZUCCARDI Julio 2001

Una estimación.

154 Elementos para mejorar la adaptabilidad del Mauricio SANTA MARÍA S. Agosto 2001 mercado laboral colombiano. Norberto ROJAS DELGADILLO 155 ¿Qué tan poderosas son las aerolíneas colombianas? Ximena PEÑA PARGA Agosto 2001

Estimación de poder de mercado de las rutas colombianas.

156 Elementos para el debate sobre una nueva reforma Juan Carlos ECHEVERRY Septiembre 2001 pensional en Colombia. Andrés ESCOBAR ARANGO César MERCHÁN HERNÁNDEZ

Gabriel PIRAQUIVE GALEANO Mauricio SANTA MARÍA S. 157 Agregando votos en un sistema altamente Francisco GUTIÉRREZ S. Octubre 2001 desistitucionalizado. 158 Eficiencia -X en el Sector Bancario Colombiano Carlos Alberto CASTRO I Noviembre 2001 159 Determinantes de la calidad de la educación en Colombia. Alejandro GAVIRIA Noviembre 2001 Jorge Hugo BARRIENTOS 160 Evaluación de la descentralización municipal. Fabio SÁNCHEZ TORRES Noviembre 2001 Descentralización y macroeconomía 161 Impuestos a las transacciones: Implicaciones sobre el bienestar Rodrigo SUESCÚN Noviembre 2001 y el crecimiento. 162 Strategic Trade Policy and Exchange Rate Uncertainty Fernando MESA PARRA Noviembre 2001 163 Evaluación de la descentralización municipal en Colombia. Alberto MALDONADO C. Noviembre 2001 Avances y resultados de la descentralización Política en Colombia 164 Choques financieros, precios de activos y recesión Alejandro BADEL FLÓREZ Noviembre 2001 en Colombia. 165 Evaluación de la descentralización municipal en Juan Gonzalo ZAPATA Noviembre 2001 Colombia. ¿Se consolidó la sostenibilidad fiscal de los Olga Lucía ACOSTA municipios colombianos durante los años noventa. Adriana GONZÁLEZ 166 Evaluación de la descentralización municipal en Ma. Mercedes MALDONADO Noviembre 2001 Colombia. La descentralización en el Sector de Gonzalo VARGAS FORERO Agua potable y Saneamiento básico. 167 Evaluación de la descentralización municipal en Edgar GONZÁLEZ SALAS Diciembre 2001 Colombia. La relación entre corrupción y proceso de descentralización en Colombia. 168 Evaluación de la descentralización municipal en Carmen Helena VERGARA Diciembre 2001 Colombia. Estudio general sobre antecedentes, Mary SIMPSON

diseño, avances y resultados generales del proceso de descentralización territorial en el Sector Educativo.

169 Evaluación de la descentralización municipal en Edgar GONZÁLEZ SALAS Diciembre 2001 Colombia. Componente de capacidad institucional.

ARCHIVOS DE ECONOMIA No Título Autores Fecha 170 Evaluación de la descentralización municipal en Iván JARAMILLO PÉREZ Diciembre 2001 Colombia. Evaluación de la descentralización en

Salud en Colombia.

171 External Trade, Skill, Technology and the recent Mauricio SANTA MARÍA S. Diciembre 2001 increase of income inequality in Colombia

172 Seguimiento y evaluación de la participación de los Dirección de Desarrollo Diciembre 2001 resguardos indígenas en los ingresos corrientes de la Territorial Nación para el período 1998 y 1999.

173 Exposición de Motivos de la Reforma de la Ley 60 de Dirección de Desarrollo Diciembre 2001 1993. Sector Educación y Sector Salud Social 174 Transferencias, incentivos y la endogenidad del gasto Eduardo WIESNER DURÁN Enero 2002. Territorial. Seminario internacional sobre Federalismo

fiscal - Secretaría de Hacienda de México, CEPAL, ILPES, CAF - Cancún, México. 18-20 de Mayo de 2000

175 Cualificación laboral y grado de sindicalización Flavio JÁCOME LIÉVANO Enero 2002. 176 OFFSETS: Aproximación teórica y experiencia Nohora Eugenia POSADA Febrero 2002. Internacional. Yaneth Cristina GIHA TOVAR

Paola BUENDÍA GARCÍA Alvaro José CHÁVEZ G. 177 Pensiones: conceptos y esquemas de financiación César Augusto MERCHÁN H. Febrero 2002. 178 La erradicación de las minas antipersonal sembradas Yilberto LAHUERTA P. Marzo 2002. en Colombia - Implicaciones y costos- Ivette María ALTAMAR 179 Economic growth in Colombia: A reversal of "Fortune"? Mauricio CÁRDENAS S. Marzo 2002. 180 El siglo del modelo de desarrollo. Juan Carlos ECHEVERRY G Abril 2002. 181 Metodología de un Modelo ARIMA condicionado Juan Pablo HERRERA S. Abril 2002. para el pronóstico del PIB. Gustavo A. HERNÁNDEZ D. 182 ¿Cuáles son los colombianos con pensiones César Augusto MERCHÁN H. Abril 2002.

privilegiadas?

183 Garantías en carreteras de primera generación. José Daniel REYES PEÑA. Abril 2002 Impacto económico.

184 Impacto económico de las garantías de la Nación José Daniel REYES PEÑA. Abril 2002 en proyectos de infraestructura.

185 Aproximación metodológica y cuantitativa Ricardo PÉREZ SANDOVAL Abril 2002 de los costos económicos generados por el Andrés VERGARA BALLÉN problema de las drogas ilícitas en Colombia Yilberto LAHUERTA P (1995 - 2000) 186 Tendencia, ciclos y distribución del ingreso Juan Carlos ECHEVERRY G. Abril 2002. en Colombia: una crítica al concepto de Andrés ESCOBAR ARANGO

"modelo de desarrollo" Mauricio SANTA MARÍA S.

187 Crecimiento y ciclos económicos. Efectos de los choques Igor Esteban ZUCCARDI H. Mayo 2002. de oferta y demanda en el crecimiento colombiano

188 A general equilibrium model for tax policy Thomas Fox RUTHERFORD. Mayo 2002. analysis in Colombia. The MEGATAX model. Miles Kenneth. LIGHT

189 A dynamic general equilibrium model for tax Thomas Fox RUTHERFORD. Mayo 2002. policy analysis in Colombia. Miles Kenneth. LIGHT Gustavo HERNÁNDEZ

190 Sistema Bancario Colombiano: Alejandro BADEL FLÓREZ. Junio 2002.

¿Somos eficientes a nivel internacional?

ARCHIVOS DE ECONOMIA No Título Autores Fecha 191 Política para mejorar el servicio de transporte público DNP: DIE- GEINF Junio 2002. urbano de pasajeros. 192 Two decades of economic and social development Carlos Eduardo VÉLEZ Junio 2002.

in urban Colombia: a mixed outcome Mauricio SANTA MARÍA, Natalia MILLAN Bénédicte DE LA BRIERE World Bank (LAC/PREM)

193 ¿Cuáles colegios ofrecen mejor educación en Colombia? Jairo NÚÑEZ MÉNDEZ Junio 2002. Roberto STEINER

Ximena CADENA Renata PARDO CEDE, U. de los Andes

194 Nuevos enfoques de política regional en América Latina: Edgard MONCAYO J. Julio 2002.

El caso de Colombia en perspectiva histórica. Las nuevas teorías y enfoques conceptuales sobre el

desarrollo regional. ¿Hacia un nuevo paradigma? Separata 1 de 7 195 Nuevos enfoques de política regional en América Latina: Edgard MONCAYO J. Julio 2002.

El caso de Colombia en perspectiva histórica. Las políticas regionales: Un enfoque por generaciones

Separata 2 de 7 196 Nuevos enfoques de política regional en América Latina: Edgard MONCAYO J. Julio 2002.

El caso de Colombia en perspectiva histórica. Un mundo de geometría variable: Los territorios que ganan y los que pierden.

Separata 3 de 7 197 Nuevos enfoques de política regional en América Latina: Edgard MONCAYO J. Julio 2002.

El caso de Colombia en perspectiva histórica. Enfoques teóricos y evidencias empíricas sobre el desarrollo regional en Colombia.

Separata 4 de 7 198 Nuevos enfoques de política regional en América Latina: Edgard MONCAYO J. Julio 2002.

El caso de Colombia en perspectiva histórica. Las políticas regionales en Colombia.

Separata 5 de 7 199 Nuevos enfoques de política regional en América Latina: Edgard MONCAYO J. Julio 2002.

El caso de Colombia en perspectiva histórica. Tendencias del desarrollo regional en Colombia. -Polarización, apertura y conflicto-

Separata 6 de 7 200 Nuevos enfoques de política regional en América Latina: Edgard MONCAYO J. Julio 2002.

El caso de Colombia en perspectiva histórica. Marco conceptual y metodológico para el diseño de una nueva generación de políticas de desarrollo regional en Colombia.

Separata 7 de 7 201 Viabilidad de los servicios públicos domiciliarios Mauricio SANTA MARÍA Agosto 2002 en la ciudad de Santiago de Cali. Francisco BERNAL Carlos David BELTRÁN David VILLALBA 202 Optimal enforcement: Finding the right balance Jaime Andrés ESTRADA Agosto 2002 203 Does corporate governance matter for developing Paula ACOSTA MÁRQUEZ Agosto 2002 countries? An overview of the Mexican case.

ARCHIVOS DE ECONOMIA No Título Autores Fecha 204 Reflexiones sobre el proceso de paz del gobierno de Andrés Camilo LEGUÍZAMO Agosto 2002 PASTRANA y las FARC-Ep: (1998-2002) 205 Contratación pública en Colombia y teoría Económica. Yuri GORBANEFF Septiembre 2002. 206 Does planning pay to perform in infrastructure? Daniel TORRES GRACIA Septiembre 2002.

Deconstructing the babylon tower on the planning/ performance relationships in energy, telecommunications and transport sectors – colombian case.

207 A dynamic analysis of household decision making in urban Fabio SÁNCHEZ TORRES Octubre 2002. Colombia, 1976-1998 Jairo Núñez Méndez Changes in household structure, human capital and its returns, and female labor force participation .

208 Inversión pública sectorial y crecimiento Alvaro A. PERDOMO S. Octubre 2002. Económico: Una aproximación desde la Metodología VAR. 209 Impacto macroeconómico y distributivo del Impuesto de Ömer ÖZAK MUÑOZ. Octubre 2002. seguridad democrática. Oscar Mauricio VALENCIA

210 Empleo informal y evasión fiscal en Colombia. Jairo A. NÚÑEZ MÉNDEZ Octubre 2002.

211 Diagnóstico del programa de reinserción en Colombia: Maria Eugenia PINTO B. Noviembre 2002.

mecanismos para incentivar la desmovilización voluntaria Andrés VERGARA BALLÉN individual. Yilberto LAHUERTA P.

212 Economías de escala en los hogares y pobreza. Francisco Javier LASSO V. Noviembre 2002. Tesis para optar el título de Magíster en Teoría y

Política Económica de la Universidad Nacional de Colombia.

213 Nueva metodología de Encuesta de hogares. Francisco Javier LASSO V. Noviembre 2002. ¿Más o menos desempleados?

214 Una aproximación de la Política Comercial Estratégica Ricardo E. ROCHA G.. Diciembre 2002. para el ingreso de Colombia al ALCA. Juan Ricardo PERILLA

Ramiro LÓPEZ SOLER

215 The political business cycle in Colombia Allan DRAZEN Enero 2003. on the National and Regional level. Marcela ESLAVA University of Maryland 216 Balance macroeconómico de 2002 y Dirección de Estudios Enero 2003. Perspectivas para 2003. Económicos 217 Women workers in Bogotà ‘s Informal sector: Jairo G. ISAZA CASTRO Febrero 2003. Gendered impact of structural adjustment Policies in the 1990s.

Tesis para optar el título de Magíster en Estudios de Desarrollo del Instituto de Estudios Sociales de The Hague- Holanda.

218 Determinantes de la duración del desempleo en Carlos E. CASTELLAR P. Marzo 2003. el área metropolitana de Cali 1988-1998. José Ignacio URIBE G. (Documento elaborado por profesores del Depar- tamento de Economía de la Universidad del Valle)

219 Conflicto, violencia y actividad criminal en Colombia: Fabio SÁNCHEZ TORRES Marzo 2003. Un análisis espacial. Ana María DÍAZ Michel FORMISANO 220 Evaluating the impact of SENA on earnings and Alejandro GAVIRIA URIBE Abril 2003. Employment. Jairo A. NÚÑEZ MÉNDEZ

221 Un análisis de la relación entre inversión extranjera y Erika Bibiana PEDRAZA Abril 2003. Comercio exterior en la economía colombiana.

ARCHIVOS DE ECONOMIA No Título Autores Fecha 222 Free Trade Area of the Americas. An impact Miles Kenneth. LIGHT Abril 2003. Assessment for Colombia. Thomas Fox RUTHERFORD 223 Construcción de una Matriz de Contabilidad Social Gustavo Adolfo HERNÁNDEZ Mayo 2003. Financiera para Colombia. 224 Elementos para el análisis de Incidencia tributaria. Andrés ESCOBAR Mayo 2003. Gustavo HERNÁNDEZ Gabriel PIRAQUIVE Juan Mauricio RAMIREZ 225 Desempeño económico por tipo de4 firma: Erika Bibiana PEDRAZA Mayo 2003. Empresas nacionales vs. Grandes y pequeñas receptoras De inversión extranjera. 226 El balance estructural del Gobierno Central en Colombia. Natalia SALAZAR Junio 2003. Diego PRADA 227 Descentralización y Equidad en América Latina: Eduardo WIESNER Junio 2003. Enlaces Institucionales y de Política 228 Ciclos económicos y mercado laboral en Colombia: Fabio SÁNCHEZ TORRES Julio 2003. ¿quién gana más, quién pierde más? 1984-2000. Luz Magdalena SALAS Oskar NUPIA 229 Efectos de un acuerdo bilateral de libre comercio con Direcciones de Estudios Julio 2003.

Estados Unidos Económicos y de Desarrollo Empresarial del DNP

230 Pobreza, crimen y crecimiento regional en Colombia. Ricardo Ernesto ROCHA G. Agosto 2003. (Versión para comentarios) Hermes Fernando MARTÍNEZ

231 Contracciones leves y profundas: Jorge E. SÁENZ CASTRO Agosto 2003. Efectos asimétricos sobre la pobreza Juan Pablo HERRERA S. El caso colombiano 1984-2000. Oscar E. GUZMÁN SILVA

232 Sistema de modelos multivariados para la proyección Carlos Alberto CASTRO I. Septiembre 2003. del Producto Interno Bruto

233 Yet another lagging, coincident and leading index for Carlos Alberto CASTRO I. Septiembre 2003. The Colombian economy.

234 Posibles implicaciones de la legalización del consumo, Andrés VERGARA BALLÉN Septiembre 2003. Producción y comercialización de las drogas en Colombia. Yilberto LAHUERTA P. Sandra Patricia CORREA

235 Impactos económicos generados por el uso de minas Yilberto LAHUERTA P. Septiembre 2003. antipersonal en Colombia.

236 ¿Cuánto duran los colombianos en el desempleo y en el Hermes Fernando MARTÍNEZ Septiembre 2003. Empleo?: Un análisis de supervivencia.

237 Barreras a la entrada en el mercado de compras del Fernando J. ESTUPIÑAN Octubre 2003. Sector público. Un análisis de estructura de mercado en la perspectiva De la negociación del Area de Libre Comercio de las Américas.

238 Relative labor supply and the gender wage Gap: Diego F. ANGEL-URDINOLA Octubre 2003. Evidence for Colombia and the United States. Quentin WODON

239 The gender wage Gap and poverty in Colombia. Diego F. ANGEL-URDINOLA Octubre 2003.

. Quentin WODON 240 The impact on inequality of raising the minimum wage: Diego F. ANGEL-URDINOLA Octubre 2003.

Gap- narrowing and reranking effects. Quentin WODON

ARCHIVOS DE ECONOMIA No Título Autores Fecha 241 Inversión y restricciones crediticias en la década de Catalina DELGADO G. Octubre 2003.

los 90 en Colombia.

242 Metodologías de estimación del balance estructural: Luis Edgar BASTO M.. Noviembre 2003. Una aplicación al caso colombiano.

243 The cost of disinflation in Colombia : José Daniel REYES P.. Noviembre 2003. -A sacrifice Ratio Approach-