Can credibility offset electricity price effect on business confidence? An empirical investigation from a large emerging economy

ABSTRACT

We analyse the ‘offsetting effect’ that the central bank credibility can have in mitigating the harmful effects of the increase in electricity price on business confidence. Based on data from the Brazilian economy from 2010 to 2019, we perform several regressions with different specifications and methods, taking into account aggregate level and the eleven industrial sectors with higher electricity consumption. In general, our findings support the view that an increase in electricity price harms business confidence, but there is an ‘offsetting effect’ when central bank credibility increases.

KEYWORDS:

• Electricity price
• central bank credibility
• business confidence
• emerging economy

JEL CLASSIFICATION:

• E31
• E52
• Q43.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ Throughout this paper, we use the term ‘business confidence’ to refer to ‘entrepreneurs’ confidence’.

² For an analysis regarding the electricity price as the cost of production, see JOHANSSON and THOLLANDER (2018) and SARWAR et al. (2018).

³ Concerning the effect of the entrepreneur’s sentiment on the economy, see KHAN and UPADHAYAYA (2020), SALHIN, SHERIF, and JONES (2016), HOLMES and SILVERSTONE (2010), and TAYLOR and MCNABB (2007).

⁴ Because there are central bank credibility indexes scored as zero over the period under analysis, and we use the variable in logs in our models, we changed the variable as log(1+ CRED).

⁵ For an analysis concerning exchange rate pass-through in emerging markets, see de MENDONÇA and TIBERTO (2017) and Tunç (2017).

⁶ Table A1 (appendix) shows the sources of data, and description of the variables, and descriptive statistics.:

⁷ It is important to note that the business confidence index is a ‘leading indicator’, so it anticipates the economy’s performance. Therefore, business confidence is a forward-looking variable, and thereby it is natural that our models consider only the current value of the dependent variables.

⁸ Concerning the instruments, we use economic growth rate, next calendar year growth rate expectations, next calendar year exchange rate expectations, next year gross public debt expectations, economic activity index, government’s evaluation (disapprove), and dummy equal to ‘1’ for periods with a technical recession and ‘0’ otherwise (sources: Central Bank of Brazil and Brazilian National Confederation of Industry). The list of instruments for each model is available upon request to the authors.

⁹ In order to eliminate the proliferation of the instruments, the maximum of lags applied for each instrument was 4. Moreover, the number of instruments used for all models is less than 15% in relation to the total of observations).