Fiscal communication and its effects on monetary policy in

Brazil: an analysis based on natural language processing

UFRPE-UAST, Department of Economics. PhD in Applied Economics and Permanent Professor.

ORCID: https://orcid.org/0000-0002-9578-1510. E-mail: diego.pitta@ufrpe.br.

Comunicação fiscal e seus efeitos na política monetária no Brasil: uma

análise baseada em processamento de linguagem natural

D i e g o P i t t a d e J e s u s

C á s s i o d a N ó b r e g a B e s a r r i a

We l l i n g t o n C h a r l e s L a c e r d a N o b r e g a

Submission date: August 14, 2025

Approval date: October 10, 2025

UFPB, Department of Economics. PhD in Applied Economics and Permanent Professor. ORCID:

https://orcid.org/0000-0002-3704-0343. E-mail: cassiodanobrega@yahoo.com.br.

UFPE-CAA, Department of Economics. PhD in Applied Economics and Permanent Professor.

ORCID: https://orcid.org/0000-0003-0634-9396. E-mail: wellington_charles@hotmail.com.

Abstract

The purpose of this article is to investigate how the tone of the fiscal authority's communications

can influence monetary policy decisions. To this end, a fiscal policy sentiment index was

developed using machine learning techniques, using the monthly public debt reports issued by

the National Treasury as a source of information. The sentiment index was used as an

explanatory variable in two approaches to achieve the central objective of the article. In the

first, a traditional version of the central bank's reaction function was estimated using classic

econometric techniques. In the second, the analysis was expanded towards a Dynamic

Stochastic General Equilibrium model (DSGE), in order to estimate central bank reaction

functions and, with this, produce inferences about the effect of fiscal policy sentiment in the

behavior of monetary policy. The main results suggest that fiscal policy sentiment has

influenced the monetary policy decision-making process in Brazil, indicating a possible

scenario of fiscal dominance. In this sense, this article contributes an unprecedented approach

to an important topic in public finance by reinforcing the fundamental role of communication

and coordination between monetary and fiscal authorities.

Keywords: Machine Learning. Textual analysis. Natural Language Processing. Fiscal

Dominance.

JEL Code: C02, C63, E62

Resumo

O objetivo deste artigo é investigar como o tom da comunicação da autoridade fiscal pode

influenciar as decisões de política monetária. Para tanto, foi desenvolvido um índice de

sentimento da política fiscal utilizando técnicas de machine learning, tendo como fonte de

informação os relatórios mensais da dívida pública emitidos pelo Tesouro Nacional. O índice

de sentimento foi utilizado como variável explicativa em duas abordagens para alcançar o

objetivo central do artigo. Na primeira, foi estimada uma versão tradicional da função de reação

do banco central utilizando técnicas econométricas clássicas. Na segunda, a análise foi

expandida para um Modelo de Equilíbrio Geral Estocástico Dinâmico (DSGE), a fim de estimar

as funções de reação do banco central e, com isso, produzir inferências sobre o efeito do

sentimento da política fiscal no comportamento da política monetária. Os principais resultados

sugerem que o sentimento da política fiscal tem influenciado o processo de tomada de decisão

da política monetária no Brasil, indicando um possível cenário de dominância fiscal. Nesse

sentido, este artigo contribui com uma abordagem inédita a um importante tópico das finanças

públicas, reforçando o papel fundamental da comunicação e coordenação entre as autoridades

monetárias e fiscais.

Palavras-chave: Machine Learning. Análise Textual. Processamento de Linguagem Natural.

Dominância Fiscal.

JEL Code: C02, C63, E62

1. Introduction

The analysis of monetary and fiscal policy instruments as means of economic stimulus or

stabilization is a topic of debate among academics, market professionals, and economic

policymakers. This discussion gained prominence with the seminal work of Sargent and

Wallace (1981), which highlighted the importance of coordination between monetary and fiscal

policies for achieving a stable economy. The argument is that, given the relationship between

their instruments, the lack of coordination can lead to reduced monetary policy efficiency or

even contradictory outcomes.

Monetary dominance is characterized by a scenario in which the monetary authority determines

the amount of revenue needed to meet the fiscal authority's requirements, resulting in stable

issuance of bonds and currency. Thus, greater control over inflation is achieved because public

debt is primarily financed through primary surpluses rather than the issuance of currency or

public bonds. On the other hand, under a fiscal dominance regime, the fiscal authority generates

a primary surplus independently of the need to stabilize the debt-to-GDP ratio, and the monetary

authority loses control over the price level, as it is compelled to generate the seigniorage

revenue necessary for government solvency (Nobrega, Maia, and Besarria, 2020).

Concerns about the performance of public finances have been a topic of debate in Brazil, raising

questions about the government's stance on fiscal policy management and the sustainability of

public accounts. More recently, during one of the longest monetary tightening cycles in the

history of the inflation-targeting regime in the country, several breaches of the spending cap

were observed—such as Constitutional Amendments No. 113/2021, No. 109/2021, and No.

1/2022—leading to an increase in public spending and signaling the exhaustion of the so-called

"New Fiscal Regime" established in 2016. In this context, the debate on the importance of

monetary and fiscal policy coordination has gained prominence, as noted in the 251st Minutes

of the Monetary Policy Committee (Copom), which states:

“The Committee reiterated the various channels through which fiscal policy can

influence inflation, not only through its direct effects on aggregate demand but also via

asset prices, the degree of economic uncertainty, inflation expectations, and the neutral

interest rate. The Committee assessed that changes in quasi-fiscal policies or the

reversal of structural reforms that result in a less efficient allocation of resources could

weaken the effectiveness of monetary policy”. (Minutes of the 251st meeting,

paragraph 12, December 2022).

In recent years, this topic has gained prominence, and a significant number of studies have

sought to investigate the existing dominance regime and the manner in which the interaction

between monetary and fiscal policies occurs. Notable contributions include the works of Issler

and Lima (2000), Schymura (2015), Tanner and Ramos (2003), Fialho and Portugal (2005),

Ázara (2006), Aguiar (2007), Gadelha and Divino (2008), Junior (2010), Ornellas (2011),

Araujo and Besarria (2014), Ferreira et al. (2015), and Nobrega, Maia, and Besarria (2020),

Sánchez and Maldonado (2024), among others. One aspect that has not yet been empirically

tested, and which has been highlighted as a potential tool to improve coordination between

fiscal and monetary policies, is communication. In the words of the former president of the

Central Bank of Brazil (BCB), Roberto Campos Neto:

“If communication is effective, we can do less while achieving greater impact, as

communication operates through an efficiency channel. Increasingly, we observe that

this also holds true in the fiscal domain: if policymakers clearly convey their actions in

a way that enables economic agents to understand the debt convergence process, it

becomes possible to increase spending at a lower cost.” (CAMPOS NETO, 2022).

The purpose of this article is to investigate how the tone of fiscal authority communications can

influence monetary policy decisions. To achieve this objective, two main strategies are

employed: The first estimates central bank reaction functions using simple equations, including

the tone (sentiment) of fiscal policy as one of the arguments in the equation, through an

Ordinary Least Squares (OLS) model and a Generalized Method of Moments (GMM) model;

The second follows a more recent approach of using Dynamic Stochastic General Equilibrium

(DSGE) models to estimate reaction functions and, thus, produce inferences about the behavior

of monetary policy.

The variable that measures fiscal policy sentiment was created using Machine Learning through

the Natural Language Processing technique to analyze the sentiments contained in the fiscal

reports produced by the National Treasury. For the text mining process, we first used a

traditional lexicon-based method, specifically the Loughran and McDonald (2011) dictionary.

Additionally, we employed the time-variant dictionary method by Lima, Godeiro, and Mohsin

(2019), which uses Machine Learning techniques for dictionary construction. After

constructing these new variables, they were incorporated into the Central Bank's reaction

function.

Following the seminal article by Taylor (1993)

, several authors have sought to estimate

reaction functions for different economies to capture and understand the behavior of central

banks. Most of these authors proceeded with the estimation of single equations (single equation

estimation), although more recently, some have moved toward estimating reaction functions in

the context of DSGE models. This is the case for Smets and Wouters (2007), Lubik and

Schorfheide (2007), and Finocchiaro and Heideken (2013), among others

Additionally, the literature has progressed toward introducing elements that could more

accurately reflect how central banks react to changes in the economic environment. One such

element has been allowing for interest rate smoothing by including the lagged interest rate level

in the reaction function. Another has been considering inflation expectations instead of past

inflation, as in the original rule. This latter modification was implemented to reflect the forward-

looking behavior of central banks in decision-making and, in particular, the recognition that

monetary policy affects the economy with a certain lag.

Some other authors have proposed introducing additional variables into the reaction function.

One of these variables has been the inclusion of fiscal variables in the Central Bank's reaction

function. Authors such as Canzoneri, Cumby, and Diba (2001) and Kumhof, Nunes, and

Yakadina (2010) added fiscal variables, such as public debt and the primary surplus, alongside

traditional measures. The inclusion of fiscal variables in the Central Bank's reaction function

Taylor (1993) brought to the debate the fundamental issue between policy rules and discretion in the

implementation of monetary policy, highlighting the potential credibility gains achieved with economic agents

when the government follows clear rules in combating inflation. The Taylor rule, as originally proposed, describes

the Central Bank's reaction to inflation through its deviations from the pre-established target and business cycles.

Later, Clarida, Galí, and Gertler (1998) proposed a forward-looking version of the original rule, in which the

reaction occurred based on inflation expectations.

Regarding studies applied to the Brazilian economy, see the works of Minella et al. (2003), Aragón and

Medeiros (2013), among others.

makes it possible to verify how the conduct of monetary policy is affected by the performance

of fiscal policy.

However, so far, no study has tested the inclusion of variables related to the sentiment of the

fiscal authority in the Central Bank's reaction function. The inclusion of fiscal policy sentiment

can illustrate how monetary policy behaves when aware of the fiscal authority's perspectives

on the state of the country's public accounts. In this sense, if the coefficient of the sentiment

variable is significant in the Central Bank's reaction function, it can be suggested that the

monetary authority's perception of the fiscal scenario, as captured by the tone of the fiscal

authority's publications, is a relevant variable in decision-making regarding the interest rate.

Furthermore, it may indicate that fiscal policy dominates the actions of monetary policy.

Thus, the main contribution of this article is the inclusion of a polarity index that measures the

sentiment of the monetary policy manager regarding the fiscal environment in the Central

Bank's reaction function, thereby providing an alternative approach to testing the fiscal

dominance hypothesis. The creation of the fiscal policy sentiment polarity variable itself can

also be considered a contribution, as no prior study for Brazil has developed such a variable

using text mining techniques on National Treasury publications. Therefore, this article

contributes with an innovative approach to an important topic in public finance by emphasizing

the fundamental role of communication and coordination between monetary and fiscal

authorities.

Overall, the results show that the inclusion of the fiscal sentiment index derived from a fixed

dictionary did not demonstrate relevance in the Central Bank's reaction function in the OLS and

GMM models. However, the index created from a time-variant dictionary has an impact on the

monetary authority's reaction function. Furthermore, in the DSGE model, the inclusion of fiscal

policy sentiment significantly increases the marginal density compared to a basic model without

the index. Therefore, the results suggest evidence of the occurrence of fiscal policy dominance

in the conduct of monetary policy actions.

In addition to this Introduction, the article contains four (4) other sections. Section 2 presents a

description of the methodology used, including the Central Bank's reaction function and the

DSGE model. Section 3 discusses the main results obtained. Finally, Section 4 outlines the main

conclusions and limitations of the article.

2. Methodology

2.1 Central Bank Reaction Function

Taylor (1993) argued that the behavior of the American Central Bank could be described by a

simple rule that linked changes in the interest rate to deviations in inflation and output from

their potential. In the present article, in particular, the monetary policy rule is forward-looking

and will be modified with the inclusion of fiscal policy sentiment, and it can be described as:



   󰇛 󰇜  

(1)

or in its estimated form:



     

(2)

where the variables with a circumflex are in the form of deviations from the Hodrick-Prescott

(HP) filter;  is the nominal interest rate;  is the inflation rate;  is the real output of the

economy;  is the fiscal policy sentiment;  is a shock that captures the non-systematic

components in the monetary policy rule; and 󰇛󰇜󰇛󰇜

󰇛󰇜.

The monetary policy rule presented above was estimated using OLS and GMM. According to

Silva and Besarria (2018), regarding the OLS method, it is noted that it may produce biased and

inconsistent estimates in the presence of endogeneity. In this case, GMM is used as an

alternative method. It is emphasized that the adequacy of the statistical inference generated by

this method is linked to the exogeneity and relevance of the instruments adopted. In other

words, the instruments must be orthogonal to the residuals and strongly correlated with the

endogenous variables included. Furthermore, the efficiency of the estimators is directly related

to the identification analysis of the selection of instrumental variables. To select the set of

instruments, the selection criteria described in Andrews (1999) were used, while the over-

identification hypothesis was addressed using the J-test.

Table 1 summarizes the variables, sources, frequency, and main treatments applied before

estimation.

Table 1 – Description, Source, and Treatment of the Data

Variable

Description

Source

Frequency

Treatment

Selic Rate

Basic interest rate

of the Brazilian

economy

Central Bank

of Brazil

Quarterly

Log transformation;

deviation from HP-

filter trend

Inflation Rate

IPCA accumulated

quarterly inflation

Central Bank

of Brazil

Quarterly

Accumulated

quarterly

GDP

Real Gross

Domestic Product

Central Bank

of Brazil

Quarterly

Seasonally adjusted;

deviation from HP-

filter trend

Public Debt

General

Government Gross

Debt (% of GDP)

Central Bank

of Brazil

Quarterly

Deviation from trend

Primary Balance

Government

primary result (% of

GDP)

Central Bank

of Brazil

Quarterly

Seasonally adjusted;

deviation from

average

Public Debt

Cycle

Cyclical component

of public debt

Own

elaboration

(HP filter)

Quarterly

HP filter (λ = 1600)

Fiscal Policy

Sentiment Index

(SFF, SFV, SFL)

Sentiment indices

from Treasury

Monthly Public

Debt Reports

Own

elaboration

(web scraping

and NLP)

Monthly

(aggregated to

quarterly)

Rescaled to [−1,1];

lags used as

instruments

Inflation

Expectations

Expected inflation

for the following

quarter

BCB – Focus

Survey

Quarterly

Quarterly average;

forward-looking

Primary Balance

Expectation

Market expectation

for the government

primary result (% of

GDP)

BCB – Prisma

Fiscal

Database

Quarterly

Deviation from

historical mean;

aligned to observed

series

Nominal Balance

Expectation

Market expectation

for the government

nominal result (%

of GDP)

BCB – Prisma

Fiscal

Database

Quarterly

Deviation from

historical mean

Public Debt

Expectation

Market expectation

for GG Gross Debt

(% of GDP)

BCB – Prisma

Fiscal

Database

Quarterly

Deviation from

average projection;

aligned to observed

series

Note: Own elaboration.

Furthermore, for the estimation using the GMM method, we employed a comprehensive set of

instruments, including real income, the Selic rate, and inflation expectations, as well as lags of

fiscal sentiment and GDP. Additional fiscal variables, such as the primary balance, public debt,

public debt cycle, nominal balance expectation, and public debt expectation, were also

incorporated to capture broader fiscal conditions and isolate the effect of fiscal communication

on the Central Bank’s reaction function. Quarterly data spanning from the first quarter of 2003

to the second quarter of 2021 were used in the estimation.

All series were seasonally adjusted prior to estimation to remove short-term fluctuations and

align temporal frequency. The GDP and investment series were expressed as deviations from

their long-term trends estimated through the Hodrick–Prescott (HP) filter. Inflation was

measured as deviations from the inflation target. Fiscal variables, such as the primary balance

and public debt, were normalized to represent deviations from their respective averages, while

sentiment indices were rescaled to range between −1 and 1. Inflation, nominal balance, and

public debt expectations were incorporated as forward-looking indicators, reflecting the

anticipatory behavior of economic agents.

2.2 Textual Estimation Procedure

In this section, the methodology for constructing the sentiment indices () will be presented.

The process was carried out similarly to the work of Jesus and Besarria (2022); however, the

authors constructed sentiment indices using BCB communications. In this research, the fiscal

policy sentiment index was obtained from the texts of the Federal Public Debt Monthly Reports.

The publication of these reports began in November 2000 in Portuguese and in March 2003 in

English. In the present work, the English version of the report was chosen, primarily because

the most notable and widely accepted dictionary used in sentiment analysis was developed in

this language, as proposed by Loughran and McDonald (2011).

After collecting the debt reports from the National Treasury website through web scraping,

several steps were undertaken to process the set of documents to extract as much information

as possible from the linguistic corpus, thereby minimizing the loss of information resulting from

sample manipulation. Before performing the lexicographic analysis on the documents, a series

of transformations were applied to the original text. The text is first divided into a sequence of

substrings (tokens), with all characters converted to lowercase letters. Additionally, stop words

were removed, as they do not add relevant information to the analysis.

Each





aims to capture some of the narrative information in the report at time , for each

document in our sample. This measure transforms thousands of words into a single number. To

obtain each fiscal policy sentiment series





, we used three approaches: one that measures

sentiments using dictionaries with fixed lexicons, another that uses machine learning models to

construct a time-variant dictionary, and another that employs unsupervised machine learning.

According to Shapiro, Sudhof, and Wilson (2020), there are two general methodologies for

quantifying sentiment in text. The first is known as the lexical methodology. This approach

relies on predefined lists of words, called lexicons or dictionaries, with each word assigned

a score for the emotion of interest. Typically, these scores are simply 1, 0, and -1 for

positive, neutral, and negative, but some lexicons include more than three categories.

Typical applications of this approach measure the emotional content of a given text corpus

based on the prevalence of negative vs. positive words in the corpus. These word-matching

methods are called bag-of-words (BOW) methods because the contextual characteristics of

each word, such as its order in the text, grammatical class, co-occurrence with other words,

and other context-specific features in the text where the word appears, are ignored.

Among this type of method, the dictionary created by Loughran and McDonald (2011)

(hereafter, LM) stands out. The authors developed lists of negative and positive words

selected to be appropriate for financial texts. They demonstrate that their dictionaries are

superior for classifying economic and financial texts compared to other dictionaries, such

as that of Apel and Grimaldi (2012) and the Harvard Psychosociological Dictionary, which

tend to miscategorize neutral words in a financial/economic context (e.g., taxes, costs,

capital, expense, liability, risk, surplus, and depreciation). The LM dictionaries contain

2,355 negative words and 354 positive words. Therefore, for constructing the sentiment

indices using the fixed-dictionary approach, we used the LM dictionary.

Shapiro, Sudhof, and Wilson (2020) state that the second and more nascent approach

employs machine learning (ML) techniques to build complex models that probabilistically

predict the sentiment of a given text set. One of the applications of ML models is in the

construction of time-variant dictionaries. Lima, Godeiro, and Mohsin (2019) used this

approach to create a time-variant dictionary method.

According to Lima, Godeiro, and Mohsin (2019), the assumption of a time-invariant dictionary

does not seem realistic in documents that introduce new words over time or if the vocabulary

used during periods of recession differs from that used during periods of economic expansion.

The authors emphasize that even if the vocabulary were constant over time, the predictive

power of certain words might vary; in other words, the relevance of words changes over time.

However, the existing literature does not address this effect, and, as a result, the resulting

predictors do not reflect the most predictive textual information found in the documents at a

given moment. In the current context of the Covid-19 pandemic, the use of a time-variant

dictionary is essential, as new terms have become relevant in the communications of monetary

and fiscal authorities. Therefore, for constructing sentiment indices using time-variant

dictionaries, we employed the approach developed by Lima, Godeiro, and Mohsin (2019).

Thus, using the methodology proposed by the authors to construct the time-variant dictionary,

we first created a vector of time series,





, where each element of the vector shows time series

observations of the frequency with which each word (or combination of words) appears in the

monthly debt report up to time . Therefore, this step transforms the words into numerical values

without using a pre-specified (fixed) dictionary. This numerical representation is high-

dimensional and sparse; thus, dimensionality reduction must be employed. In the second step,

we used supervised machine learning to select the most predictive time series (words) 

 .

The elastic net model was chosen to perform this step:

  󰆒

󰆒

(3)

where

  

is the forecast horizon

  

are estimated by minimizing the following objective

function:







󰇛 󰆒

󰆒󰇜



(4)

where





is a



vector of predetermined predictors, such as lags of





as well as

traditional structured data predictors, and  and  are the





and





norms,

respectively. After the dimensionality reduction step and the selection of the most

predictive time series 

  using the elastic net model above, the set of selected words

(

) for each period



is the one with the highest predictive power regarding the variable

of interest . This set of words is considered our dynamic sentiment dictionary for

period



To construct the sentiment index, we calculate the aggregate contribution of the words selected

in the dynamic dictionary for each monthly debt report. Specifically, the sentiment variable





is obtained as a function of the frequency of occurrence of the selected words 

 in the report,

weighted by the prediction coefficients () estimated in the elastic net model. These

coefficients capture the statistical relationship between the occurrence of the words and the

variation of the variable of interest over time. In the present article, our response variable for

constructing the time-variant dictionary will be the General Government Gross Debt (DBGG)

as a proportion of GDP.

Finally, both dictionary approaches calculate the sentiment index as the difference between

positive and negative words, divided by the sum of positive and negative words, as proposed

by Hubert and Labondance (2018):

  

  

(5)

Thus, we obtain the sentiment measure, , which ranges between -1 and 1.

To enhance the robustness of the analysis, we also constructed a sentiment index using an

unsupervised machine learning model, in this case, the Latent Dirichlet Allocation (LDA)

model. The LDA model allows the discovery of latent topics within a set of documents.

According to Lin and He (2009), LDA learns the topic distribution for each document and

the word distribution for each topic, where the discovered topics are represented as lists of

words with associated probabilities, indicating the importance of each word within the

topic. Venugopalan and Gupta (2022) state that the resulting topics should be interpreted

semantically and with sentiment polarities. The semantic analysis of the topics seeks to

identify the general themes. Subsequently, the polarity analysis assigns sentiments

(positive, negative, neutral) to the topics based on the qualitative analysis of the most

representative words for each topic.

According to Lin and He (2009), for constructing the sentiment index from the topics of the

LDA model, it is necessary to calculate the topic proportions in each document, where each

document has a distribution over the topics, indicating the relevance of each topic to the

document. Next, it is necessary to assign a sentiment value to each document: Using the topic

proportions and the sentiment polarity associated with each topic, a sentiment score is

calculated for each document

Thus, the present study proposes to construct three sentiment indices:

SFF

: Sentiment Index of fiscal reports constructed using a fixed-lexicon dictionary

(LM);

For example, if a document has 70% of a negative topic and 30% of a positive topic, the sentiment score

will reflect this combination.

SFV

: Sentiment Index of fiscal reports constructed using a time-variant lexicon

dictionary with supervised machine learning;

SFL

: Sentiment Index of fiscal reports constructed from topics using unsupervised

machine learning.

2.3 DSGE Model

For the present section, we use the base model from the work of Jesus, Besarria, and Maia

(2020). The modification introduced by this article to that model is the creation of another

version of the interest rate rule. Thus, two DSGE models will be estimated. The first is the base

model with the traditional interest rate rule, as in the work of Jesus, Besarria, and Maia (2020)

(basic model). The second includes a modified interest rate rule with the creation of a fiscal

sentiment variable within the model, incorporated into the Taylor rule.

2.3.1 Fiscal Policy

The role of the fiscal authority in the economy is to collect taxes and issue bonds to finance its

public investment and spending. The government’s tax revenue () is composed of taxes levied

on household consumption (), labor income (), and capital income (). The government’s

intertemporal budget constraint establishes that the current level of public debt must equal the

present value of future primary surpluses, ensuring fiscal solvency in the long run. In

operational terms, the government’s flow constraint in each period can be written as:

  

(6)

where  represents the government’s real primary surplus;  is the real value of public debt

󰇡

 󰇢. The variable  is the total amount of public bonds,  

󰆒

󰆒󰆒′.

The primary surplus is given by the difference between the government's total revenue and total

expenditure during the same period:

  



(7)

where:

   

󰆒󰆒

(8)

 

󰆒

󰆒

󰆒

󰆒

󰆒󰆒 

󰆒󰆒 

󰆒󰆒 

󰆒󰆒 



󰆒󰇛

󰆒

󰆒󰆒󰇜

(9)

Being 

󰆒

󰆒

󰆒

󰆒

󰆒󰆒 

󰆒󰆒 

󰆒󰆒 

󰆒󰆒  the government's revenue from household

income; 



󰆒 the government's revenue from taxing the return on physical capital owned by

patient households; 󰇛

󰆒

󰆒󰆒󰇜 the government's revenue obtained from the total household

consumption, and consequently, government consumption is also taxed. Public investment ()

is considered an exogenous shock.

The variable that aims to represent fiscal policy sentiment will be defined in the model based

on the fiscal environment. In this case, an increase/decrease in government debt induces a

negative/positive polarity in the variable that measures the sentiment of the fiscal authority

Therefore, the variable that will measure the net negativity of the authority is given by:

   

(10)

If the difference in public debt is positive/negative, we will have higher/lower net negativity,

resulting in a more pessimistic/optimistic fiscal policy sentiment;  represents a shock

arising from abrupt changes in the fiscal environment.

2.3.2 Monetary Policy

The monetary authority adopts an inflation target and sets the interest rate through a rule

proposed by Taylor (1993). As indicated, two versions of the interest rate rule will be

considered. A basic version, where the central bank looks only at the lagged interest rate level

and deviations in future developments and GDP, which we will call the “basic rule”, while in

the second version, we include deviations in inflation expectations at horizon   from its

steady-state level. We will call this version of the instrument rule the extended version. The

basic version can be represented by the following expression:



 

 󰇛 󰇜 



(11)

It is worth noting that we chose public debt because it is the fiscal variable that already takes into account the

performance of the others.

−

This rule specifies that the current nominal interest rate depends on an inertial or lagged

component (

); the deviation of expected inflation from the target set by the monetary

authority; the output gap, represented by the deviation of output from its steady-state value;

and, finally, an i.i.d. monetary policy shock,. The subscripts  e  are integers that can take

any value. The extended version can be represented by:



 

 󰇛󰇜 

󰇛󰇜

(12)

where





is the variable that measures fiscal sentiment.

2.4 Bayesian Estimation

In this section, we discuss our methodology for estimating and evaluating the models. The

solution of the DSGE model was obtained through a first-order Taylor approximation of the

equilibrium conditions around the non-stochastic steady-state value. Given the model solution

as a state-space representation and a vector of observable variables, the models were estimated

using Bayesian techniques. Specifically, a Metropolis-Hastings algorithm, which is a Markov

Chain Monte Carlo (MCMC) method, was employed to obtain the posterior probability

distribution of the parameters. Two independent chains were generated, each consisting of

400,000 draws, using the Metropolis-Hastings algorithm. The average acceptance rate across

the two chains was approximately 40%, and convergence was assessed using the methods

proposed by Brooks and Gelman (1998). The first 180,000 draws were discarded to ensure

independence from initial conditions. The statistics of interest were then calculated based on

the joint ergodic posterior probability distribution of the structural parameters.

For the estimation, three variables were used for each quarter: real GDP, nominal interest rate,

and household consumption. These variables were chosen because they are the most relevant

endogenous variables. The variables were used in natural logarithms and seasonally adjusted.

The cyclical component of the variables was obtained using the Hodrick-Prescott filter and

covers the quarterly period from the first quarter of 2003 to the second quarter of 2021. The

model was estimated using Dynare within the Matlab software.

2.5 Calibration and Prior Distributions

Some parameters were fixed during the estimation process, while others were estimated.

For the parameters that were kept fixed, we chose to use values from the related literature

Christiano and Eichenbaum (1992), Lim and McNelis (2008), Silva, Paes, and Ospina

(2014), Cavalcanti et al. (2018), Wesselbaum (2017). Table 5 in Appendix (A) presents a

brief description of these parameters.

For the estimated parameters, we opted to use prior distributions similar to those employed

in the related literature. For the parameters indicating the degree of substitution between

private consumption and the consumption of public goods and services, e , we used a

prior beta distribution with a mean of 0.50, consistent with the value found for Brazil by

Ferreira and Nascimento (2005), Santana, Cavalcanti, and Paes (2012), and Bezerra et al.

(2014)

, with a standard deviation of 0.02 for both. For the parameters of the Taylor rule,

we used prior distributions and values for the hyperparameters commonly found in the

literature (Smets and Wouters, 2003). The parameter governing the central bank's response

to price changes, ϕπ, was set to 1.5, satisfying the Taylor principle. For the coefficient

measuring the central bank's response to the output gap,





, we used a prior normal

distribution with a mean of 0.125 (Carvalho, Silva, and Silva, 2013).

For the parameter indicating the share of physical capital in the production function, similar

to Cavalcanti et al. (2018), we adopted a prior normal distribution with a mean of 0.30 and

a standard deviation of 0.05. Finally, for all autoregressive parameters, we employed a

prior beta distribution with a mean of 0.95 and a standard deviation of 0.02.

3 Results

3.3 Estimations of the Reaction Function

For the estimation of the monetary policy rule, in addition to the traditional variables (interest

rate, inflation gap, and output gap) and sentiment indices, other fiscal and fiscal expectations

variables were added. The objective of adding these variables is to try to isolate the effects of

fiscal communication in the Central Bank's reaction function. The additional variables are:

This value found for Brazil can be considered conservative. Bailey (1971) and Aschauer (1985) found values

between 0.23 and 0.42 for the United States.

Government Primary Balance, Government Nominal Balance, Public Debt, Public Debt Cycle,

Primary Balance Expectation, Nominal Balance Expectation, Public Debt Expectation.

Table 2 presents the estimations of the Central Bank's reaction function using OLS and GMM.

Columns 2 and 3 show the OLS estimates of the reaction function, while columns 4 and 5

present the reaction function estimates obtained using the GMM method. The first includes the

fiscal sentiment indices in the reaction function, while the second restricts the interest rate

response to variations in the sentiment indices.

As in the work of Silva and Besarria (2018), in general, the results obtained from the estimation

of the reaction function, regardless of the method, show a high degree of smoothing in the

interest rate dynamics, indicating that the Central Bank makes gradual changes to the interest

rate. Regarding the coefficient related to inflation expectations, it is observed that it is

statistically significant and greater than one, indicating that the Central Bank satisfies the Taylor

principle by increasing the real interest rate in response to deviations in expected inflation.

Table 2 - Estimation of the Reaction Function

Parameters in Structural Form



















0,9371

[0,0303]

0,8824

[0,0300]

0,9168

[0,0228]

0,8412

[0,0284]

0,9371

0,8824

0,9168

0,8412

Inflation

Gap

0,2569

[0,0385]

0,2085

[0,0378]

0,2636

[0,0394]

0,2289

[0,0398]

4,0843

1,7730

3,1683

1,4414

GDP Gap

0,0012

[0,0001]

0,0035

[0,0001]

0,0031

[0,0001]

0,0048

[0,0001]

0.0191

0,0102

0,0421

0,0302

Primary

Balance

0,0811

[0,0466]

0,1472

[0,0457]

0,0707

[0,0466]

0,0971

[0,0458]

1,2893

1,2517

0,8498

0,6115

Public

Debt

-0,0219

[0,0107]

-0,0221

[0,0115]

-0,0237

[0,0106]

-0,0269

[0,0120]

-0,3482

-0,1879

-0,2849

-0,1694

Public Debt

Cycle

0,0017

[0,0121]

0,0094

[0,0113]

0,0046

[0,0113]

0,0111

[0,0120]

0,0270

0,0799

0,0553

0,0669

Primary

Balance

Expectation

0,0046

[0,0653]

0,0018

[0,0570]

-0,0176

[0,0609]

-0,0494

[0,0570]

0,0731

0,0153

-0,2115

-0,3111

Nominal

Balance

Expectation

-0,1172

[0,0448]

-0,0525

[0,0432]

-0,1352

[0,0405]

-0,1129

[0,0395]

-1,8633

-0,4464

-1,6250

-0,7110

Public Debt

Expectation

-0,0191

[0,0117]

-0,0211

[0,0122]

-0,0116

[0,0091]

-0,0038

[0,0111]

-0,3037

-0,1794

-0,1394

-0,0239

SFF

-0,6345

[0,3821]

-0,5319

[0,4101]

-5,3954

-3,3495

SFV

0,7875

[0,1936]

0,6836

[0,2020]

6,6964

4,3048

SFL

0,2087

[0,1003]

0,0889

[0,0955]

1,7747

0,5599

Note: The terms in brackets represent the standard deviations of the estimated coefficients.

Regarding the sentiment indices, the OLS and GMM estimates indicate that the SFF variable

does not have statistical significance. The SFL variable showed significance only in the OLS

model. On the other hand, the SFV sentiment index demonstrated statistical significance in both

models, with a positive sign in the models. Thus, when the tone of the National Treasury's

communications is optimistic, the Central Bank increases the interest rate, and when the tone is

more pessimistic, the monetary authority reduces the interest rate. It is also worth noting that

the weight assigned to variations in fiscal policy sentiment was higher than the weight related

to inflation expectations.

These results indicate that fiscal policy sentiment significantly affects the conduct of monetary

policy, more specifically through the SFV index. It is worth noting that this sentiment index

uses public debt as the dependent variable in its construction, while the SFF index relies on a

fixed financial dictionary. Therefore, these results make it evident that the fiscal environment

(represented by the sentiment index) is a relevant variable in the conduct of monetary policy.

Consequently, such results may suggest the possibility of a fiscal dominance scenario. Since

the SFV index was constructed based on public debt as the dependent variable, it reflects the

market's perception of fiscal sustainability.

If the results show that the SFV significantly influences monetary policy decisions, this may

indicate that the central bank is adjusting its policy based on fiscal conditions. In other words,

when the SFV reflects fiscal deterioration, the central bank may choose not to raise interest

rates, even if inflation is rising, to avoid worsening the government's fiscal situation.

This subordination of monetary policy to fiscal needs is a clear indication of fiscal dominance,

where the need to stabilize public finances takes precedence over controlling inflation or other

monetary policy goals.

Despite these results, the estimation of simple equations presents some issues. As highlighted

by Lubik and Schorfheide (2007), Finocchiaro and Heideken (2013), and Silva and Besarria

(2018), these estimations suffer from endogeneity when estimated by OLS and may exhibit bias

when estimated by GMM, due to sample size and bias related to the use of stages in two-stage

GMM and iterative GMM estimations, which is proportional to the number of moment

conditions in instrumental variable models. Moreover, in practice, finding good instruments to

implement the GMM method is not trivial. Invalid or weak instruments represent a serious

challenge for reliable inference and may compromise the estimates.

3.4 Estimations of DSGE Models

This subsection presents the results of the estimation of the DSGE models. Table 5 (Appendix)

shows the mean values, standard deviations, and the corresponding lower (HPD inf) and upper

(HPD sup) bounds of the 95% Highest Posterior Density (HPD) credibility interval for the

estimated parameters using the Bayesian inference technique for the two types of models

estimated.

As found in the work of Silva and Besarria (2018), it is observed that the estimated

parameters show little variation between the two models, with posterior means very close

between the two. The estimation results reveal that Brazilian data provide little information

regarding the amount of labor supplied by patient households in the production of

intermediate goods. A similar result was identified by Finocchiaro and Heideken (2013) in

their estimations of this parameter for the United Kingdom and Japan, where the posterior

mean value was exactly equal to the prior value.

For the parameters defining the maximum borrowing capacity of impatient households and

entrepreneurs, the values were lower than the prior mean. These estimates may reflect the

fact that households and firms in the country face more restrictive access to credit compared

to developed countries.

The parameters for preference and technology shocks were higher than the prior mean,

revealing that these shocks are more persistent than initially hypothesized by the prior

distribution hyperparameters. Finocchiaro and Heideken (2013) obtain similar results in

their estimations of shock processes for the United States, United Kingdom, and Japan.

Regarding the parameters of the Central Bank's reaction function, the parameter measuring the

Central Bank's response to changes in inflation expectations was positive and greater than one,

satisfying the Taylor principle. Similarly, the parameter measuring the response to output

deviations was positive. Both parameters suggest the behavior of a Central Bank operating

under a flexible inflation targeting regime, assigning weight to both inflation and the real side

of the economy. Regarding fiscal policy sentiment, the mean of the parameter reflecting the

Central Bank's response, , was positive and significant. Therefore, as in the case of the simple

equation models, there are indications that the BCB explicitly considered fiscal policy

sentiment, i.e., the fiscal environment, in its reaction function during the period analyzed.

According to Silva and Besarria (2018), a convenient tool in Bayesian analysis is the use of

estimates to compare alternative models. One such method is to use the marginal density of the

data associated with each model and compare them, subsequently choosing the model best

supported by the data. One way to obtain the marginal density of the data, from the joint

posterior distribution, is to use Geweke's (1999) estimator, the modified harmonic mean

estimator. Table 3 presents the values for the marginal density of the data (in log) computed

using this estimator.

Table 3 - Comparison Between Models

Specification

Marginal Data Density

Log Bayes Factor

Without fiscal policy sentiment

147.6836

With fiscal policy sentiment

220.4311

72.7475

Note: Own elaboration.

Based on the model evaluation criteria created by Kass and Raftery (1995), we found some

evidence in favor of the model that includes fiscal policy sentiment. Therefore, based on the

estimations, it can be said that there is evidence, albeit limited, that the BCB explicitly

considered fiscal policy behavior in its interest rate decision-making process.

An analysis of the impulse response functions (Appendix B) indicates that the inclusion of fiscal

policy sentiment in the reaction function alters the transmission of monetary policy on GDP,

household consumption, and labor supply, with the exception of the interest rate. It is also

evident that the increase in the interest rate brought typical recessionary effects to the economy,

showing that the positive interest rate shock caused a reduction in consumption, labor supply,

and aggregate demand. These effects were observed regardless of whether the Central Bank

included fiscal policy sentiment in the reaction function or not.

The results obtained from the estimations of the DSGE models, although limited, align with

those found in the OLS and GMM model estimations. In other words, the results indicate that

fiscal policy sentiment is a relevant variable in the monetary authority's decision-making

process regarding the interest rate. Thus, this new investigative approach suggests a high

likelihood of the occurrence of the fiscal dominance phenomenon during the analyzed period.

These findings converge with those reported in the works of Ázara (2006), Junior, Garcia-

Cintado, and Junior (2021), Ornellas and Portugal (2011), and Nobrega, Maia, and Besarria

(2020).

4 Conclusion

The recent fiscal situation in Brazil raises the question of whether, and to what extent, the

Central Bank of Brazil has reacted to this scenario during the period from 2003 to 2021. This

article investigates this issue through two main strategies. The first involves estimating central

bank reaction functions using simple equations, including fiscal policy sentiment as one of the

arguments in the equation. The second develops a Dynamic Stochastic General Equilibrium

(DSGE) model and uses it to produce inferences about the behavior of monetary policy in

response to fiscal authority sentiment.

The results suggest that the central bank incorporated the fiscal authority's sentiment into its

monetary policy decisions. By reacting positively to the tone of the fiscal authority's

communication, the Central Bank of Brazil may have adopted a stance of increasing or

decreasing the interest rate when the fiscal outlook was more optimistic or negative,

respectively, indicating potential fiscal dominance.

Despite the promising results, future versions need to address certain issues to make the findings

more robust. First, constructing a fiscal sentiment variable using Portuguese-language

dictionaries could test whether significant changes occur in the results. For future research, it

would be worthwhile to use the dictionary by Machado (2019). Another point is that, besides

the dictionary by Lima, Godeiro, and Mohsin (2019), there are other dictionaries that use

machine learning, so testing these alternatives would also be important to provide greater

robustness to the results.

It is also recommended that future research not only test the inclusion of fiscal authority

sentiment but also examine fiscal policy uncertainty. In the present study, we discussed whether

the sentiment or tone of fiscal policy affects the monetary authority's interest rate decisions, but

fiscal policy uncertainty may also play a role—or even more so—than sentiment. Therefore,

we recommend including an index of fiscal policy uncertainty, particularly the Macroeconomic

Uncertainty Index (IIM) created by Besarria et al. (2021), which is an uncertainty index

constructed using Natural Language Processing of the Federal Public Debt Monthly Reports.

References

Aguiar, M. T. de. (2007). Dominância Fiscal e a Regra de Reação Fiscal: Uma Análise

Empírica para o Brasil [Disertación de maestría, Universidade de São Paulo - USP].

Repositorio de la USP. http://www.teses.usp.br/teses/disponiveis/12/12140/tde-19102007-

124240/pt-br-php

Andrews, D. W. (1999). Consistent moment selection procedures for generalized method of

moments estimation. Econometrica, 67(3), 543–563.

Apel, M., & Grimaldi, M. (2012). The information content of central bank minutes (Riksbank

Research Paper Series, No. 92). Sveriges Riksbank.

Aragón, E. K. d. S. B., & Medeiros, G. B. de. (2013). Testing asymmetries in central bank

preferences in a small open economy: A study for Brazil. Economía, 14(2), 61–76.

Araujo, J. M., & Besarria, C. da N. (2014). Relações de dominância entre as políticas fiscal e

monetária: uma análise para economia brasileira no período de 2003 a 2009. Revista de

Economia, 40(1).

Aschauer, D. (1985). Fiscal policy and aggregate demand. American Economic Review, 75(1),

117–127. http://www.jstor.org/stable-/1812707

Ázara, A. de. (2006). Dominância Fiscal e Suas Implicações Sobre a Política Monetária no

Brasil: Uma Análise do Período 1999-2005 [Disertación de maestría, Fundação Getúlio

Vargas - FGV]. Repositorio de la FGV.

http://bibliotecadigital.fgv.br/dspace/handle/10438/2022

Bailey, M. (1971). National income and the price level: Aggregate supply and aggregate

demand model. Mcgraw-Hill.

Banco Central do Brasil. (2022, diciembre). Ata do COPOM (251$^{a}$ Reunião).

Besarria, C. C. da N., Aragón, E. K. S. B., & Nóbrega, W. C. L. (2021). Incerteza

macroeconômica e seus efeitos fiscais: Uma análise a partir de precessamento natural e

modelos dinâmicos estocásticos de equilíbrio gerla (dsge). Cadernos de Finanças Públicas,

21(1).

Bezerra, A., Pereira, R. A. C., Campos, F. A. O., & Callado, M. C. (2014). Efeitos de

crescimento e bem-estar da recomposição dos investimentos públicos no Brasil. Pesquisa e

Planejamento Econômico, 44(3), 579–607. http://repositorio.ipea.gov.br/handle/11058/3651

Brooks, S., & Gelman, A. (1998). General methods for monitoring convergence of iterative

simulations. Journal of Computational and Graphical Statistics, 7(4), 434–455.

Campos Neto, R. (2022). Coordenação de política monetária e fiscal foi principal ponto na

conversa com Haddad [Entrevista con E. Taiar, L. Garcia y A. Ribeiro]. Valor Econômico.

https://valor.globo.com/financas/noticia/2022/12/15/campos-ponto-principal-na-conversa-

com-haddad-foi-coordenao-de-poltica-monetria-e-fiscal.ghtml

Canzoneri, M. B., Cumby, R. E., & Diba, B. T. (2001). Is the price level determined by the

needs of fiscal solvency? American Economic Review, 91(5), 1221–1238.

Carvalho, D., Silva, M., & Silva, I. (2013). Efeitos dos choques fiscais sobre o mercado de

trabalho brasileiro. Revista Brasileira de Economia, 67(2), 177–200.

Cavalcanti, A., Vereda, L., Doctors, R. B., Lima, F. C., & Maynard, L. (2018). The

macroeconomic effects of monetary policy shocks under fiscal rules constrained by public

debt sustainability. Economic Modelling, 71, 184–201.

https://www.sciencedirect.com/science/article/pii/S0264999317302468

Christiano, L., & Eichenbaum, M. (1992). Current real business cycle theories and aggregate

labor market fluctuations. American Economic Review, 82(3), 430–450. [link suspeito

removido]

Clarida, R., Gali, J., & Gertler, M. (1998). Monetary policy rules in practice: Some

international evidence. European Economic Review, 42(6), 1033–1067.

Ferreira, L. A. M., Godeiro, L. L., de Sá Porto, M. C., & Pereira, P. C. F. (2015). Dominância

fiscal ou dominância monetária no Brasil: uma análise do regime de metas de inflação

[Ponencia]. Universidade Federal de Uberlândia.

Ferreira, P., & Nascimento, L. (2005). Welfare and growth effects of alternative fiscal rules

for infrastructure investment in Brazil (Ensaios Econômicos, No. 604). Fundação Getúlio

Vargas. https://bibliotecadigital.fgv.br/dspace/bitstream/handle/10438/422/1996.pdf

Fialho, M. L., & Portugal, M. S. (2005). Monetary and fiscal policy interactions in brazil: An

application of the fiscal theory of the price level. Estudos Econômicos (São Paulo), 35(4),

657–685.

Finocchiaro, D., & Heideken, V. V. (2013). Do central banks react to house prices? Journal of

Money, Credit and Banking, 45(8), 1659–1693.

https://onlinelibrary.wiley.com/doi/abs/10.1111/jmcb.12065

Gadelha, S. R. d. B., & Divino, J. A. (2008). Dominância fiscal ou dominância monetária no

Brasil? uma análise de causalidade. Economia Aplicada, 12, 659–675.

Geweke, J. (1999). Using simulation methods for bayesian econometric models: inference,

development, and communication. Econometric Reviews, 18(1), 1–73.

Hubert, P., & Labondance, F. (2018). Central bank sentiment [Working paper]. Narodowy

Bank Polski. https://www.nbp.pl/badania/seminaria/14xi2018.pdf

Issler, J. V., & Lima, L. R. (2000). Public debt sustainability and endogenous seigniorage in

Brazil: Time-series evidence from 1947–1992. Journal of Development Economics, 62(1),

131–147.

Jesus, D. P. de, & Besarria, C. da N. (2022). Narrativas do banco central e previsões

macroeconômicas: Usando análise textual de machine learning [Ponencia]. 50º Encontro

Nacional de Economia - ANPEC, Fortaleza - CE.

Jesus, D. P. de, Besarria, C. da N., & Maia, S. F. (2020). The macroeconomic effects of

monetary policy shocks under fiscal constrained: An analysis using a dsge model. Journal of

Economic Studies.

Junior, C. J. C., Garcia-Cintado, A. C., & Junior, K. M. (2021). A modern approach to

monetary and fiscal policy. International Review of Economics Education, 37, 100232.

Junior, K. M. (2010). Há dominância fiscal na economia brasileira? uma análise empírica para

o período do governo Lula. Indicadores Econômicos FEE, 38(1).

Kass, R., & Raftery, A. (1995). Bayes factors. Journal of the American Statistical

Association, 90, 773–795.

Kumhof, M., Nunes, R., & Yakadina, I. (2010). Simple monetary rules under fiscal

dominance. Journal of Money, Credit and Banking, 42(1), 63–92.

Lima, L. R., Godeiro, L., & Mohsin, M. (2019). Time-varying dictionary and the predictive

power of fed minutes. (Available at SSRN 3312483).

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3312483

Lim, G., & McNelis, P. (2008). Econometric methods. The MIT Press.

Lin, C., & He, Y. (2009). Joint sentiment/topic model for sentiment analysis. In Proceedings

of the 18th ACM conference on Information and Knowledge Management (pp. 375–384).

Association for Computing Machinery.

Loughran, T., & McDonald, B. (2011). When is a liability not a liability? textual analysis,

dictionaries, and 10-ks. The Journal of Finance, 66(1), 35–65.

Lubik, T. A., & Schorfheide, F. (2007). Do central banks respond to exchange rate

movements? a structural investigation. Journal of Monetary Economics, 54(4), 1069–1087.

Machado, M. A. V. (2019). Índice de sentimento textual: uma análise empírica do impacto

das notícias sobre risco sistemático. Revista Contemporânea de Contabilidade, 16(40), 24–42.

Minella, A., Freitas, P. S. de, Goldfajn, I., & Muinhos, M. K. (2003). Inflation targeting in

Brazil: Constructing credibility under exchange rate volatility. Journal of International

Money and Finance, 22(7), 1015–1040.

Nóbrega, W. C. L., Maia, S. F., & Besarria, C. da N. (2020). Interação entre as políticas fiscal

e monetária: uma análise sobre o regime de dominância vigente na economia brasileira.

Análise Econômica, 38(75).

Ornellas, R. (2011). Fiscal and monetary interaction in Brazil. In 33º Meeting of the Brazilian

Econometric Society.

Ornellas, R., & Portugal, M. S. (2011). Fiscal and monetary policy interaction in Brazil. In

Anais do XXXIII Encontro Brasileiro de Econometria (pp. 1-28). Sociedade Brasileira de

Econometria (SBE).

Sánchez, F. V., & Maldonado, M. T. (2024). Definición y efectividad de la política monetaria

mexicana, estudio econométrico y de causalidad inductiva. Análisis Económico, 39(102),

153–165.

Santana, P., Cavalcanti, T., & Paes, N. (2012). Impactos de longo prazo de reformas fiscais

sobre a economia brasileira. Revista Brasileira de Economia, 66(2), 247–269.

http://www.scielo.br/pdf/rbe/v66n2/a06v66n2.pdf

Sargent, T. J., & Wallace, N. (1981). Some unpleasant monetarist arithmetic. Federal Reserve

Bank of Minneapolis Quarterly Review, 5(3), 1–17.

Schymura, L. G. (2015). A sombra da dominância fiscal e a reação do sistema político.

Revista Conjuntura Econômica, 69(11), 8–10.

Shapiro, A. H., Sudhof, M., & Wilson, D. J. (2020). Measuring news sentiment. Journal of

Econometrics, 218(2), 528-545.

Silva, M. E. A. d., & Besarria, C. d. N. (2018). Política monetária e preços dos imóveis no

Brasil: Uma análise a partir de um modelo dsge. Revista Brasileira de Economia, 72(1), 117–

143.

Silva, W., Paes, N., & Ospina, R. (2014). A substituição da contribuição patronal para o

faturamento: Efeitos macroeonômicos, sobre a progressividade e distribuição de renda no

Brasil. Revista Brasileira de Economia, 68(4), 517–545.

http://bibliotecadigital.fgv.br/ojs/index.php/rbe/article/view/14269

Smets, F., & Wouters, R. (2003). An estimated dynamic stochastic general equilibrium model

of the euro area. Journal of the European Economic Association, 5(1), 1123–1175.

Smets, F., & Wouters, R. (2007). Shocks and frictions in US business cycles: a Bayesian

DSGE approach. American Economic Review, 97(3), 586–606.

https://www.aeaweb.org/articles?id=10.1257/aer.97.3.586

Tanner, E., & Ramos, A. M. (2003). Fiscal sustainability and monetary versus fiscal

dominance: Evidence from Brazil, 1991–2000. Applied Economics, 35(7), 859–873.

Taylor, J. (1993). Discretion versus policy rules in practice. Carnegie-Rochester Conference

Series on Public Policy, 39, 195–214. http://www.nviegi.net/teaching/taylor2.pdf

Venugopalan, M., & Gupta, D. (2022). An enhanced guided LDA model augmented with

BERT based semantic strength for aspect term extraction in sentiment analysis. Knowledge-

based systems, 246, 108668.

Wesselbaum, D. (2017). Expectation shocks and fiscal rules. International Economics and

Economic Policy, 14, 1–21. https://link.springer.com/article/10.1007/s10368-017-0389-z

Appendix A – Parameters of Calibration

Table 4 - Calibration of Parameters

Parameters

Description

Value



Price rigidity factor

0.85



Depreciation rate of physical capital

0.02



Physical capital adjustment

2.00



Elasticity of substitution between intermediate goods

6.00



Proportion of wage used as collateral

0.90



Proportion of property value used as collateral

0.85



Tax rate on domestic consumption

0.2313



Tax rate on labor income

0.1713



Tax rate on capital income

0.1441



Discount factor of patient households

0.99



Discount factor of impatient households

0.94

Source: Own elaboration.

Appendix B – Impulse-Response Functions

Figure 1 - Impulse-Response Functions to a One Standard Deviation Shock in the Nominal

Interest Rate

Note: The dotted lines represent a 68% credibility interval for the case with fiscal policy sentiment.

Appendix C – Bayesian Estimation

Table 5 - Bayesian Estimation Results

Prior

Model without Fiscal Policy

Sentiment

Model with Fiscal Policy

Sentiment

Parameter

Prior Dist.

Prior

Mean

Std.

Dev.

Posterior

Mean

Lower

MDP

Upper

MDP

Posterior

Mean

Lower

MDP

Upper

MDP



Beta

0,95

0,02

0,9918

0,9874

0,9964

0,9913

0,9866

0,9958



Beta

0,95

0,02

0,9528

0,9229

0,9799

0,9942

0,9907

0,9979



Beta

0,95

0,02

0,9494

0,9212

0,9822

0,9364

0,9027

0,9643



Beta

0,95

0,02

0,9985

0,9978

0,9994

0,9985

0,9977

0,9995



Beta

0,95

0,02

0,9470

0,9147

0,9732

0,9475

0,9139

0,9836



Beta

0,95

0,02

0,9578

0,9362

0,9850

0,9598

0,9432

0,9769



Gama Reversa

0,01

0,0118

0,0079

0,0153

0,0233

0,0185

0,0279



Gama Reversa

0,01

0,0091

0,0023

0,0169

0,0384

0,0316

0,0449



Gama Reversa

0,01

0,0221

0,0173

0,0266

0,0097

0,0022

0,0188



Gama Reversa

0,01

0,0089

0,0022

0,0168

0,0120

0,0081

0,0159



Gama Reversa

0,01

0,0084

0,0023

0,0157

0,0085

0,0023

0,0157



Gama Reversa

0,01

0,0093

0,0023

0,0173

0,0108

0,0022

0,0210



Gama Reversa

0,01

0,0089

0,0023

0,0166



Beta

0,50

0,02

0,3058

0,2768

0,3360

0,3245

0,2946

0,3490



Beta

0,50

0,02

0,3021

0,2697

0,3401

0,3027

0,2788

0,3266



Beta

0,80

0,10

0,8436

0,7591

0,9459

0,8281

0,7183

0,9335



Normal

1,50

0,50

1,4539

1,0432

1,9279

0,7083

0,1919

1,2302



Normal

0,125

0,05

0,0729

0,0293

0,1153

0,0719

0,0160

0,1268



Normal

0,300

0,05

0,3118

0,2569

0,3759

Normal

0,300

0,05

0,1526

0,0947

0,2040

0,1846

0,1299

0,2388

Source: Own elaboration.