




































In ternationa l
Scholars
Journa ls

 

African Journal of Agricultural Marketing ISSN 2375-1061 Vol. 7 (3), pp. 001-008, March, 2019. Available online at 
www.internationalscholarsjournals.org © International Scholars Journals 

 

Author(s)                                           Author(s) retain the copyright of this article. 
 
 

 

Full Length Research Paper 

 

Determination of the sacrifice rate in Turkey, Brazil 
and Italy: A comparison among countries 

 
Tuba Başkonuş Direkçi 

 
Department of Economics, Faculty of Economics and Administrative Sciences, Gaziantep University, 

Turkey. E-mail: baskonus@gmail.com 
 

Accepted 10 September, 2018 
 
The ongoing research on output and inflation tradeoff contradiction is what most nations’ central banks are facing 
currently. After a long inflationary period, which started in 2001, Turkey has also become successful in overcoming 
and curbing down the inflation rate with fiscal and monetary policy measures. Although the general acceptance of 
output sacrifice for inflationary improvements is a common knowledge of today, the “Neo Classical School” of 
thought argues that market actors will not delay its expectations. The fundamental purpose of this study at hand is 
to comprehend the existence of output losses during the anti-inflationary periods between1990 and 2008 of Turkey. 
In this study, time periods considered will be divided into intervals for testing the existence of output losses in the 
measure of sacrifice ratios. This study will be very exploratory, especially for the lived experience in the post 2001 
period. Estimation of the study relies on the basic Phillips Curve hypothesis. Variables used in the research at hand 
are: inflation rate, output gap and public sector borrowing rate (PSBR). With the output gap, the intention is to 
measure the potential output loss. Sacrifice ratio will be measured as output loss, which is considered as a part of 
inflationary decline. Similar ratios will be calculated and compared for alternative countries. Relatively, a small 
sacrifice ratio for the post 2001 period of Turkey can be interpreted as successful central bank implementations. 
Italy’s findings when compared to Turkey’s results have the following parallel findings: if the sacrifice ratio for the 
disinflation period is positive, then the sacrifice ratio for the post disinflation periods is negative for both countries. 
Findings show that disinflation periods cause a restraint on the production of Brazil. Persistence coefficient for 
Brazil is much more on a lower level as compared to Turkey and Italy. 

 

Key words: Disinflation, sacrifice rate, monetary policy, output gap. 
 
 
INTRODUCTION 

 
For a long period of time, Turkish economy had the 
principal agenda of never dropping inflation rate, high 
inflation rate, policies that should or have been adopted 
to lower the inflation rate and the impact of these adopted 
policies on the economy itself. By using disinflationist 
policies for reducing inflation, it can result to creation of 
internal output gap increase. In other terms, while 
reducing inflation, the national output is reduced and due 
to this reason, the output is sacrificed. The sacrifice rate 
in laments term is the amount of total output that is lost to 
reduce the inflation rate (Çetinkaya and Yavuz, 2002). 
 
 

 
JEL classification: E31, E52, C22. 

 
 
 
 

 
When the world economy was looked at from a 

summarizing point of view, it can be stated that inflations 
have negative effect on nations, and due to this effect, 
attempts were made to reduce inflation through the 
disinflation programs that were frequently adopted. Many 
economists in that sense state that low inflation will 
increase earnings in the long term. Taylor (1983) and 
Sargent (1983) state the importance of disinflation 
periods in a supporting manner while they also discuss 
the recovery speed of disinflation period from different 
perspectives (Taylor, 1983; Sargent, 1983). Taylor in this 
sense states the position of “slow disinflation pace will 
reduce costs”, while Sargent states the position of “fast 
disinflation being less costly” meaning that it will cost less 
output gap. Jordan, within the same topic, discusses that 

file:///C:\Users\user\Documents\REPUBLICATION\AGRICULTURAL%20SCIENCES\AppData\Local\Temp\www.internationalscholarsjournals.org


 
 
 

 

on disinflation periods, with increasing output gap periods 
being high, the contrary situations convey no situation 
with symmetric gap reduction (Jordan, 1997). This in 
meaning conveys a finding that states the opposite of 
sacrifice rates do not generate utility rate environment. 
Conclusively, at these points, one can infer that it is a 
difficult task to reduce inflation and its effects on the 
economy.  

In this study, the 1990 to 2008 periods sacrifice rate for 
Turkish economy was evaluated. It should be known that 
as a result of the 2001 year (May period) adopted dis-
inflation oriented monetary policies, fractional drops in the 
inflation was experienced within the Turkish economy. 
Due to this reason, the research has adopted the view of 
this time period with pre and post economic situations. To 
give a summary of the Turkish economy, the inflation rate 
increase on average for the year 1999, 2000 and 2001 
was 68.8, 39.9 and 68.5%, respectively. After the 
execution of disinflation purposed monetary policies, the 
obtained data from Turkish Statistical Institute (TUIK) 
dropped within the inflation rate. To give quantifiable 
evidence to such drop, the inflation rate was 31.7, 16.65, 
8.49, 8.63, 9.61, 9.17 and 8.62% for the year 2003, 2004, 
2005, 2006, 2007, 2008 and 2009, respectively.  

The aim of this study is to analyze the 2001 year set 
IMF supported policies of “Shift to Strong Structural 
Economic Plan” adopted by the Republic of Turkey. In 
this way, by a yearly period, the country’s sacrifice ratio 
and “persistence or hysteria coefficient” could be calcula-
ted. This study does not only focus on Turkey and its 
performances, but on Italy and Brazil for comparative 
purposes of these experience analysis. When studying 
Turkey, which is the EU accession process, Italy was 
chosen for comparison being an EU country. Brazil was 
chosen because Turkey and Brazil have similar stories on 
their economy as regard inflation and crisis. 
 

 

LITERATURE REVIEW 

 

Looking at the related literature, almost all studies confirm 
the existence of sacrifice ratio from an economist 
perspective. On the other hand, the concept of cost and 
speed of disinflation have different approaches. The 
meaning of cost in this study is based on the rate of 
reduction of output while attempting to reduce inflation 
over reducing inflation rate. If the inflation reduces due to 
shrinking policies, the output gap increases more 
drastically; meaning that the inflation cost is more than 
what was anticipated. Also, if an attempt is made to 
reduce inflation rate with a short term policy, disinflation 
speed with an alternative approach is reduced with a long 
term policy. In this situation, disinflation pace should be 
defined in a slower manner. Taylor (1983) states that dis-
inflation may take a slow pace and thus, create a lower 
medium for cost. According to his studies, while inflation 
is reduced, the sacrificed income drop is higher than the 

 
 
 
 

 

inflation rate. Sargent (1983) in this sense states “fast 
pace disinflation have less costs”. Ball on the subject 
matter states that disinflation cost on average inflation 
rate is far more sensitive and affected (Ball, 1994). Ball in 
his study examined the sacrifice ratio factors, which were 
a result of disinflation. In the study, sacrifice ratio was 
defined as the ratio of loss output over the reducing trend 
inflation rate. Defining sacrifice ratio by developing a 
method for calculating single disinflation period is 
implemented for 65 periods of midlevel inflation holding 
OECD countries (Ball, 1994). As such, Ball’s results 
reflect that the sacrifice ratio was reducing as in the pace 
of disinflation. In this study, slowly reducing the inflation 
has raised the sacrifice ratio, but sudden policies of 
inflation reduced sacrifice ratios. At the same time, ratio 
also reduced with the wage setting institutions flexibility. 
Wage rigidity is found to be also increasing the sacrifice 
ratio. Openness of economy did not affect the ratio. Ball’s 
results also show that disinflation policy at the beginning 
and income policies have no certain relation with the 
ratio.  

Zhang (2001), in his study, looked at the sacrifice ratio 
in terms of its long term and empirical effects. His study 
reflects the term also called hysteric effect, which is a 
strong persistence. The empirical study focused on the 
1960 to 1990 unemployment data quarterly and also on 
G-7 countries. As a result of Zhang’s study, the calcula-
ted long term sacrifice ratio is greater than the non-
calculated long term sacrifice ratio. Sacrifice ratio and the 
beginning of inflation rate have been found to be in a ne-
gative relationship. However, long term sacrifice ratio and 
wage profitability have no relation between each other.  

Jordan (1997) looked at situations of disinflation and 
rapid inflation increase, and investigated them both. The 
article was about the central bank’s independence rate in 
terms of explaining sacrifice and benefice ratio fluctua-
tions. Ball (1994) also used the output gain approach for 
19 industrialized countries time periods between 1960 
and 1992. The purpose of the research was to look at 
how central bank independence ratio would define 
sacrifice and benefice ratio fluctuations. Possible flow 
pace, inflation rate change, nominal wage profitability and 
early period inflation rates were considered within the 
study. As a result, central banks with the higher indepen-
dence rate had higher sacrifice ratios, but inflations 
increased rate periods and benefice rate that were not 
high were observed. Daniels on his research with a 
similar research to Jordan’s 1997 study, looked at 
openness of an economy and relationship of sacrifice 
ratio based on central bank independence (Daniels et al., 
2005). In this study, Daniels found a positive relation 
between sacrifice ratio and openness of economy. As 
openness increases, the central bank’s positive effect on 
sacrifice ratio reduces.  

The study of Yay (2001) was one of the first studies 
that based their discussion on Turkey. This study first 
looked at the disinflationist policy costs theoretically and 



 
 
 

 

then investigated the different counties heterodox 
disinflation policies. It was observed that Argentina (1979 
to 1981 and 1985 to 1986), Brazil (1985 to 1986 and 
1994), Chile (1978 to 1982), Mexico (1987 to 1994), 
Uruguay (1978 to 1982 and 1990) and Israel (1982 to 
1983 and 1985) were enriching countries in terms of 
results of the study. Lastly, the study looked at the 1999 
and onward period with its IMF originating stability 
program.  

Another study whose discussion was based on Turkey 
was done by Çetinkaya and Yavuz (2002). The study co-
ntained sacrifice ratio and output gap measurements with 
alternative measurement. Based on 3 different periods (3  
× 3 matrix), two positive sacrifice ratios were obtained in 
this study. In the study, output loss could not be obtained 
through grand disinflation, whereas in this study, the 
relationship proposed to be obtained with the current 
framework literature is the relationship between dropping 
disinflation period, reducing outputs and reducing inflation 
rate. The relationship between the potential income used 
and the actual income difference between the year’s 
inflation rate series is the relation coefficient, which is 
actually the sacrifice ratio. These coefficients’ calculation 
is composed of two different methods. The first method 
includes a regression equation, which will provide a ratio. 
This method does not include between period changes, 
but includes a single constant change coefficient; thus, 
the method is heavily criticized. Other approaches include 
an approach where a different ratio can be calcu-lated for 
each year individually as a series. In this study, both 
methods will be adopted and individually calculated. After 
evaluating sacrifice ratio, within the disinflation periods, 
disinflation periods monetary policy affects sus-tainability 
and will be measured with “hysteria coefficient”. 
 

 
METHODOLOGY 
 
Model 

 
The evaluation of sacrifice ratio for the sake of reducing inflation is 
the main purpose of this paper; however, while producing the model 
to evaluate the sacrifice ratio for the Turkish economy between the 
1990 and 2008 periods, sacrifice ratio was first defined based on 
the well-known Phillips curve. The Gross Domestic Product (GDP) 
or the Gross National Product relationship with output was basically 
defined as (Okun, 1978; Gordon and King, 1982):  

 
(1) 

 
In Equation (1), y is the actual output level, yt* is the potential 

output level, πt is the actual inflation rate at period t, (πt – π t-1) is 

the first difference of πt  and  represents the error term. Within   
the equation, the sacrifice ratio is defined by , while the conducted 
regression analysis result can be accepted as a stable value. The 
sacrifice ratio is expected to be positive. The meaning of this 
expectation is that disinflationist periods in-between inflation rate 
increase (πt – πt-1), and it is the increasing difference that is 
observed between the actual and potential difference (yt – yt*). To 
elaborate more, if the in-between two periods inflation rate dropped 

 
  

 
 

 
as a result of the disinflation policies caused by actual shrinkage 
experienced within the economy and actual output dropping, the 
gap between potential and actual inflation will increase; as such, 
this is defined as the output gap. As a result, the defined rate of 
sacrifice is the output reduction of the inflation rate of one score. 
The higher the in-between periods inflation rate, the higher the 
output gap.  

However, the equality in finding sacrifice ratio based on its 
fundamentals vary in nature in the second model based on Ball’s 
work (Ball, 1994). While inflation is rising or demand is in sudden 
fluctuations, the constant assumption is not found appropriate, 
which demands us to adopt the second approach shown as:  
 

(2) 
 
 
From Equation (2), the output gap and the same period’s actual 
inflation rates’ drop is defined as disinflation period (πt – πt-1). By so 
doing, the sacrifice ratio is turned into a series, which allows us to 
view the general trend. This method was adopted from Ball’s 
approach by Jordan (1997), Bernanke et al. (1999) and Boschen 
and Weise (2001).  

Although the recommended equality will change annually, and 
based on the annual change i (Ball, 1994), this ratio will change 
from year to year f it is stable or not. Also, it will follow a similar 
structure at the same time (Zhang, 2001). The important issue for 
this method employed is not only the sacrifice ratio, but also the 
effect impact originating from the change. Due to this reason, the 
mentioned shortcomings and problems were considered and the 
following equation was adopted.  
 

 
(3)  

 
Here, coefficient represents the persistence effects’ impact. When 

is close to 1, the coefficients’ persistence level will increase as  
well. Zhang assumed coefficient as: 0 < <1. In the study of 
Blinder (1987), like Zhang (2001), the results showed long term 
effects could be obtained in places where monetary shock ori-
ginating disinflations were obtained (Çetinkaya and Yavuz, 2002).  

Ball (1994) and Zhang (2001) both looked at highly persistent 
effect also known as hysteric effect. Hysteric effect will be 
happening if constricting monetary policies affect the output, 
meaning that if there is disinflation, an ongoing effect will occur 
(Zhang, 2001). 

 

EMPIRICAL ANALYSES 
 
As mentioned earlier, the purpose of this study is to 
investigate if there are any output losses during anti-
inflationist monetary policy adoption in Turkey during the 
1990 to 2008 year periods and comparing analysis 
results with countries like Italy and Brazil in terms of 
economics. Due to this reason, the data used for analysis 
will be introduced, and the countries mentioned will be 
evaluated and calculated based on sacrifice ratios 
individually to draw a general comparison of the 
environment. 

 

Data 
 
For this study, Gross Domestic Product (GDP) data were 
obtained from OECD for Italy and Brazil,  apart from  Turkey 



 
 
 
 

 
Table 1. Disinflation periods in Republic of Turkey.  

 
Disinflation periods in the Republic of Turkey   

Period Beginning Ending 

I January 1992 December 1992 

II October 1994 December 1996 

III January 1998 January 2001 

IV February 2001   
 

TURKEY’S 1990-2008 SACRIFICE RATIO 
 

80 

 
60 

 
40 

 
20 

 
0 

 
-20 

 
-40  

90 92 94 96 98 00 02 04 06 08 
 

Figure 1. Turkey’s 1990-2008 sacrifice ratio graph. 
 

 

Turkey. The reason for such an approach is to lower the 
variance between the series and standardization. 
Consumer Price Index (CPI) was obtained for all three 
countries (2005=100). The GNP series for Italy was in her 
national currency with millions on current prices quarterly 
and were removed from seasonality, while Brazil’s GNP 
series was in billions and in her national currency as well. 
Turkey’s GNP series was obtained from Turkish Central 
Bank’s electronic data distribution center. To obtain the 
GNP series’ potential and actual values difference, firstly, 
the data’s natural algorithm was taken following a Hodrick 
Prescott Filter. The obtained Hodrick Prescott Filter trend 
values and natural algorithm values difference allows us 
to obtain the gap. CPI inflation series was obtained by 
executing natural algorithm, followed by taking the first 
level difference of the variables. 
 

 

Sacrifice ratio for Turkey 

 
 

 

another disinflationist policy. After the adopted 
“Introduction to 1q inflation rate experienced a significant 
decline. Due to the pre-2001 February dis-inflation 
periods, the analysis, in a short period timeline, will be 
looked at individually at the pre and post 2001 periods.  

The GDP data series, obtained for Turkey, originated 
from the Republic of Turkey Central Bank’s (TCMB) 
website, while CPI inflation rate for Brazil and Italy data 
series with the mentioned categories were obtained from 
the Organization for Economic Co-operation and 
Development’s (OECD) website. However, all data series 
used within the research are composed of quarterly data 
series.  

For the primary calculation of the sacrifice ratio of 
Turkey, Ball’s method was adopted as in Equation 2 
(Ball, 1994). This calculation allows the observation of 
each individual year sacrifice ratios, individually.  

 

(2) 
 

 
Within the 1990 to 2008 periods, Turkish economy 
frequently experienced disinflation policies. When Table 1 
was observed, the executed monetary policies can be 
seen until the fourth period. Following these periods, the 
Fourth period, February 2001 economic crisis demanded 

 
 
For the secondary equation, the denominator was formed 
by the output gap and the nominator was formed by the 
difference of the quarterly period’s inflation difference. 
The progression of the obtained sacrifice ratios over the 
years for Turkey can be observed in Figure 1. As can be 



 
 
 

 
Table 2. Regression analysis results among Turkey.  

 

Dependent variable 
 α Coefficient β Coefficient   

 

     
 

      
 

 TRYGAP TRDINF TRYGAP1 LM TEST ARCH 
 

1990:3-2008:4 MODEL I -0.075 (-1.18)   0.022 
 

 MODEL II -0.214(-0.308) 0.655 (7.65)  0.242 
 

 TRYGAP TRDINF TRYGAP1 LM TEST ARCH 
 

1990:3-2001:4 MODEL I -0.077(-0.988)   0.074 
 

 MODEL II 0.036(1.32) 0.837(10.46)  0.31 
 

 TRYGAP TRDINF TRYGAP1 LM TEST ARCH 
 

2002:1-2008:3 MODEL I 0.406(1.87)   0.67 
 

 MODEL II 0.21(1.62) 0.77(5.22)  0.059 
 

 
The coefficients parentheses are Z statistics from the GARCH analysis. 

 

 

seen from the graphic, over the years sacrifice ratio 
generally seemed stable, while on some periods, certain 
peaks were observed. In the year 1993 and 1999, the 
sacrifice ratio seemed to drop, while in 1995, 1997 and 
2007, high peaks were experienced.  

The analysis conducted for Turkey in the second 
section, as regards the country’s interests, contains the 
output gaps between two periods of inflation difference 
(TRDINF), the cited model that defined the regression 
difference between two periods of the inflation’s variable 
and the output gaps lag series regressions (Table 2). The 
conducted two regressions adopted the sum of ordinary 
least squares method. For the first model, the   

coefficient represents the sacrifice ratio, while for the 
second model the β coefficient represents the persistency 
coefficient.  

The conducted regression also contains disinflation 
periods referenced and are adopted for the entire variable 
sets. Following this procedure, the entire data set was 
divided into 2001:1 and 2001:2 periods and the models 
were conducted for these two set periods. In this way, the 
drops effect of the May 2001 “Introduction to Strong 
Economy” programs inflation rate on sacrifice ratio can be 
defined. The sacrifice ratio findings show that contrary to 
the literature, the effect is negative and is found with no-
relation.  

Due to the changing variance and autocorrelation 
problems with the conducted least sum of squares 
method, models were preferably analyzed with GARCH 
(1, 1) method. This way, variance problem was handled. 
With the conducted regression analysis covering all years 

with sacrifice ratio represented by coefficient, it was 
concluded contrary to the literature findings that the study 
reflected negative and non-consistent results. Persis-
tence coefficient which was represented by β coefficient, 
as expected is between the values of zero and one. As 
such, the results reflect a positive and statistically 
significant outcome. 

 
 

 

As persistence coefficient is closer to the other, one 
can say that there is a high level of persistency. The 
years between 1990 and 2008 periods have significant 
coefficients of 0.65, meaning that disinflation has an 
ongoing effect. The years between 1990 and 2001 
periods show that the regression model reflects no 
relation on the first model, but models one lag output gap 
series and in addition provides positive results (0.036) 
that are coherent with the related literature. While z sta-
tistic has a value of 1.32 and p value is 0.18, persistence  
coefficient  with the value of 0.837 is statistically 
significant. The last regression run for Turkey is for the 
2002 to 2008 year periods. The calculated sacrifice ratio 
value for the periods is 0.406. Z statistic stands out to be 
1.87 and the error value as 0.06. As a consequence, this 
is significant. Between the mentioned years, Turkey was 
within a sacrifice of 0.406% due to the inflation. The  
periods mentioned showed that coefficient is 0.77, 
which is also significant. The coefficient for each 3 ana-
lysis is significant and on average varies between 0.65 
and 0.83, which is on a persistent level. This result in 
terms of the analysis signifies disinflations existence for 
all the years considered.  
 

 

Sacrifice ratio for Italy 

 

The calculated sacrifice ratio using Ball’s (1994) 
approach can be seen in Figure 2. With the formula used  

as: , the reflected 
ratio results as compared to Turkey’s inflation cost is 
more constant for Italy.  

For Italy, in the years 1993 and 1995, there have been 
drastic increases on the sacrifice ratio. This finding as 
compared to Turkey’s sacrifice ratio is shown in Figure 1. 
The regression analysis executed for Turkey is also a-
dopted for Italy without any changes or alterations. Here, 



 

 

 
 Italy’s 1990-2008 year periods sacrifice ratio. 

  

300  
 

 
200  

 

 
100  

 

 
0  

 

 
-100  

 

 
-200  

90 92 94 96 98 00 02 04 06 08 
 

            Figure 2. Italy’s 1990-2008 year periods sacrifice ratio. 
 
 
 

 

Here, Italy’s output gap is calculated and Italy’s 1990:1 to 
2008:3 periods’ inflation and difference series with one 
lag regression models are generated. As in Turkey 
analysis, the first model independently included inflation 
series as a variable; while in the second model, both 
variables were included. All data obtained on Italy from 
the OECD database are in quarterly form. Due to Italy’s 
1996 disinflationist monetary policies, the conducted ana-
lysis first included all years within the frame time before 
analyzing the pre-1996 time period. Looking at the sacri-
fice graphic post, the 1996 period seems to be constant.  
For Italy, the entire data sets included regression analysis 
results showing the two models amid negative relation 
with regards to sacrifice ratio. For the second model, the 
persistency coefficient was included. This model also 
reflected negative results which were opposite to the 
literature findings. Persistence coefficient with 0.747 
values in the findings was parallel to that of the literature 
findings. For Italy, 1990:3 to 1995:4 periods in the first 
model reflected negative and non-significant results, 
though the persistency ratio was significant with 0.806 
value. Nonetheless, the last period results were different 
from the two periods. This implies that the sacrifice ratios 
of Italy between 1996:1 and 2008:4 periods, parallel to 
the literature, are positive and statistically significant. 
Here, the persistency rate is 0.62 in value and is 
significant. The persistency rate is also known as the 
hysteric effect. Hysteric effect in summary is the return to 
equilibrium after a shock in the economy is experienced. 
At this point, a suitable question that arises is “If the long 
term trend is formed by short term balances, will the long 
term economic shocks be persis-tent?” If the short term 
experienced effects influence the long term balance, then 
we can speak of hysteria effects’ existence. From this 
standpoint, this research signifies long term continuation 

 
 
 
 
 
of money policies as hysteric effect.  

In practice, if the persistency coefficient grows bigger, 
the persistence intensity grows stronger. For an actual 
output to reach a potential output, more time is required 
than anticipated. If the coefficient is zero, we can not be 
talking about persistency effect (Zhang, 2001). Italy’s 
findings as compared to Turkey’s results have the 
following parallel findings: if the sacrifice ratio for the 
disinflation period is positive, then the sacrifice ratio for 
the post disinflation periods is negative for both countries. 
Persistency ratio in Italy, as well as in Turkey, is 0.75 on 
average, which can be categorized as high on a neutral 
level. 
 

 

Sacrifice rate for Brazil 

 

For Brazil, which parted ways with the IMF on the year 
2005, same method was adopted in calculating the 
sacrifice ratio. The conducted regression results can be 
seen in Table 4. The inner parentheses (Garch modeling 
Z values) coefficients are the t statistics of the model, 
while the values under LM and GARCH headlines are the 
significance levels.  

The first model was evaluated through GARCH (1, 1) 
method in Table 3, while Brazil’s data set was obtained 
from OECD’s database. Quarterly data series were 
obtained at the beginning of the year 1995. Ball’s 
approach for sacrifice ratio can be seen in Figure 3 (Ball, 
1994). The sacrifice ratio holding white noise properties 
have fluctuating values between 1997 and 2000 periods 
and intensified values after 2006.  

Brazil’s data series unlike Turkey and Italy does not 
start from the year 1990. Therefore, Brazil’s analyses 
were not divided into periods but were rather analyzed in 



  
 
 

 
Table 3. Regression analysis results among Italy.  

 

Dependent variable 
 α Coefficient β Coefficient   

 

ITYGAP ITDINF ITYGAP1 LMTESTI ARCH 
 

 
 

1990:3-2008:4 
MODEL I -0.11 (-0.57)   0.075 (0.78) 

 

MODEL II -0.825 (-2.96) 0.747 (10.81) 4.92 (0.085) 0.11 (0.085)  

 
 

 
 ITYGAP ITDINF ITYGAP1 LMTESTI ARCH 

1990:3-1995:4 MODEL I -1.112 (-1.24)   1.17 (0.27) 

 MODEL II -1.27 (-2.38) 0.806 (7.9) 4.14 (0.12) 0.27 (0.60) 

 ITYGAP ITDINF ITYGAP1 LMTESTI ARCH 

1996:1-2008:4 MODEL I 0.311 (0.27)   0.31 (0.57) 

 MODEL II -0.496 (-1.55) 0.62 (5.84) 2.39 (0.30) 0.017 (0.89) 
 

The coefficients parentheses are Z statistics from the GARCH analysis values under the LM and ARCH tests are p values. First models 
for each 3 terms estimated with GARCH (1,1). 

 

 
Table 4. Regression analysis results among Brazil.  

 
Dependent variable BYGAP BDINF BYGAP1 LMTESTİ ARCH 

 

1996:3-2008:4 
I. MODEL 0.110(0.85)   0.267(0.60) 

 

II. MODEL 0.057(0.394) 0.437(3.53) 2.17(0.33) 5.14(0.16) 
 

  
 

          
 

BRAZIL 1996 – 2008 SACRIFICE RATIO GRAPH. 
 

40 
 
 

30 
 
 

20 
 
 

10 
 
 

0 
 
 

-10 
 
 

-20  
90 92 94 96 98 00 02 04 06 08 

 
Figure 3. Brazil 1996 – 2008 sacrifice ratio. 

 

 

a general fashion. The regression analysis conducted for 
Brazil reflected positive results in terms of sacrifice ratio, 
although it was insignificant. The findings showed that the 
disinflation periods caused a restraint on the production 
of Brazil. The second model also supported these 
findings, in that the persistence coefficient was positive 
and significant. As such, the persistence coefficient for 
Brazil is much more on a lower level as compared to 
Turkey and Italy. 

 
 

 

Conclusion 

 

Attempts to reduce inflation can result from the 
consequences of increasing costs which can be 
discussed. These costs are generally calculated with 
achieving a measure for calculating the output costs. With 
this research for Turkey, a trial was made to explore the 
Italy and Brazil’s 1990 and 2008 year periods, which 
reflected the disinflationist monetary policies effect on 



 
 
 

 

output. Consequently, the expected sign of the sacrifice 
ratios was positive. The approach here is to assume that 
the increase of periods among inflation periods, positively 
influence the output gap as well.  

Considering all three countries, especially the 1990 to 
2008 and 1990 to 2001 year periods for Turkey, the 
analysis showed that the sacrifice ratio was insignificant 
in result and negative in relation. The implemented 
disinflation program “Phase to Strong Economy” policy 
period results with positive sacrifice ratio between the 
2002 and 2007 year periods. Sacrifice ratio for the three 
periods have the same sign. Pre-2001 period on average 
is 0.055, while post-2001 periods have positive sign and 
an increase in rate around 40%. This result also provides 
us with the different approaches adopted between the pre 
and post 2001 year. The other adapted coefficient as a 
means to detect the disinflation polices intensity is the ( ) 

coefficient. coefficient reflects the persistency power, 
while , close to 1, represents the high level of persistency. 
Coefficients’ expected value is assumed to  

be . For Turkey, the persistence coefficient for 
the given years of analysis is within 0.77 level. This 
reflects the finding that disinflation policy effects last 
extendedly, that is, the policies narrowing effect last 
longer and there is a medium encouraging output gap 
process.  

The analysis done for Italy concludes with statistically 
significant results and the expected results of the sacrifice 
ratio is parallel to Turkey with regard to disinflation policy 
for the given years. Italy on average has sacrifice ratios 
between 0.31 and 0.49. Thus, persistence ratio is 
between 0.62 and 0.806, which is statistically significant. 
When compared to Turkey, these values are very high. 
With these findings, one can say that the disinflation 
policies impact takes a long term effect on Italy as well. 
Brazil’s sacrifice ratio is positive but found to be non-
significant. Thus, the persistence ratio value of 0.43 can 
be accepted as significant for Brazil, but comparing the 
persistence coefficient to Turkey and Italy, one can say 
that Brazils persistence ratio is rather low. This result also 
shows us that the executed disinflation policies of Brazil 
had rather shorter term impact on the economy when 
comparing it to countries like Turkey and Italy.Brazil, in 
July 1st 1999, planned to shift to an inflation targeting 
plan. The Central Bank obligated targeting was reported 
quarterly and targeting was decided once in two years. 
Targeting adoption was not stable, in that they can be in 
a changing manner. The IMF themed program, adapted 
by Turkey on the other hand, rely on cutting down 
government expenditures, while supporting strict 
monetary policies. When the results were compared, 
flexible inflation targeting showed more successful and 
promising results. For countries like Turkey and Italy with 
high persistence rate, strict monetary policies should be 
reconsidered, based on disinflation policies and a much 
more flexible targeting policy in terms of inflation targeting 

 
 
 
 

 

being adopted. Due to strict monetary policies being 
composed of narrowing policies, future market 
expectations are also narrowed. This in result creates an 
environment where persistence and sacrifice ratios are in 
an augmenting manner. 
 
 
REFERENCES 
 
Ball L (1994). “What determines the sacrifice ratio?” in N G Mankiw 

(ed.) Monetary Policy University of Chicago Press, pp. 155-82.  
Bank Independence,” 
http://www.springerlink.com/content/v01qh310145r0475/fulltext.pdf 
Bernanke  B,  Thomas  L,  Frederic  SM,  Adam  SP  (1999).  “Inflation  

Targeting: Lessons from the International Experience. Princeton, NJ: 
Princeton Univ. Press. 

Blinder AS (1987). Hard Head, Soft Heart: Tough-Minded Economics 
for a Just Society. Reading, Massachusetts: Addison-Wesley.  

Boschen J, Charles W (1999). ”The Ex Ante Credibility of Disinflation 
Policy and the Cost of Reducing Inflation.” J. Macroecon., Summer 
2001, 23(3): 323-347.  

Çetinkaya A, Arzu DY (2002). “Calculation Of Output-Inflation Sacrifice 
Ratio: The Case Of Turkey.” Research Department Working Paper 
No: 11, November, October 2002  

Daniels, Joseph PN, Farrokh VD (2005). “Openness, Central Bank 
Independence and the Sacrifice Ratio.” J. Money Credit Bank,. 37(2): 
371-379.  

Gordon RJ, Stephen K (1982). “The Output Cost of Disinflation in 
Traditional and Vector Autoregressive Models. ” Brookings Pap. Econ.  

Activity, 1: 205-242. 
Jordan  TJ  (1997). “Disinflation Costs, Accelerating Inflation Gains, and  

Central Bank Independence,” 
http://www.springerlink.com/content/v01qh310145r0475/ fulltext.pdf  

Kandiller R (2001). “Güçlü Ekonomiye Geçiş Programı ve Ötesi”, Dış 
Ticaret Dergisi, Yıl: 6, Sayı: 22: 55-56.  

Okun AM (1978). “Efficient Disinflation Policies.” Am. Econ. Rev. 68: 
348-352.  

Sargent TJ (1983). “The Ends of Four Big Inflations” in: Inflation: 
Causes and Effects, ed. by Robert E. Hall, University of Chicago 
Press, for the NBER http://www.nber.org/books/hall82-1, 41–97. 

Taylor JB (1983). Union Wage Settlements During a Disinflation. Am.  
Econ. Rev., 73(5): 981-993 .  

Yay GG (2001). “Enflasyonu Düşürme Politikalarının Maliyetleri: Teori, 
Uygulama Ve Türkiye”, İktisat İşletme ve Finans, 2001 Temmuz, 
16(184): 66-83. 

Zhang LH (2001). “Sacrifice Ratios with Long-Lived Effects”,Econ  
Papers, 2001:04, 

http://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww. 

econ.jhu.edu%2Fpdf%2Fpapers%2Fall4202001.pdf;h=repec:jhu:pap 

ers:446 


