Inflation expectations of households: do they influence wage-price dynamics in India?

ABSTRACT

This paper examines the usefulness of survey-based measures of inflation expectations to predict inflation using hybrid versions of New Keynesian Phillips Curve (NKPC). While both 3 months ahead and 1-year ahead inflation expectations of households emerge statistically significant in explaining and predicting inflation in India, effectively they work as substitutes of backward looking expectations given that household expectations are found to be largely adaptive. Unlike in other countries, this paper does not find much evidence on flattening of the Phillips curve. Also, no robust evidence is found on expectations induced wage pressures influencing CPI inflation.

KEYWORDS:

• Inflation expectations
• Phillips curve
• wage-price dynamics

JEL classification:

• E52
• E24
• E31
• P24

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemented data of this article can be accessed here.

Notes

1. Expectations induced wage-price spiral requires a tight labour market condition. Along with high inflation expectations as per the household survey, one should also look at household survey results on the outlook for income and employment and actual wage/compensation growth data to assess risks to inflation from high inflation expectations (Meyer 2011).

2. In empirical estimates when the output gap coefficient is insignificant, that could reflect either output gap is a poor proxy of marginal cost (because only under certain restrictions on technology and labour market structure that output gap could be a proxy of marginal cost) or incorrect measurement of output gap. In pure versions of NKPC, with no role for backward looking expectations or intrinsic inflation inertia, current inflation is essentially discounted future marginal costs, i.e. prices are set by firms on assessment of future demand and cost conditions, and if monetary policy can credibly commit to keep output gap zero in future, disinflation without sacrifice of output is possible. In real life, however, disinflation involves sacrifice of output.

3. Sharma and Bicchal (2018) used data on wholesale price index (WPI) inflation (for the sample period Q4:2006 to Q2:2015) and Das (2014) used data on WPI and CPI-C (for the sample period September 2008 to December 2013) to test the rationality of household inflation expectations in India.

4. According to Das, Lahiri, and Zhao (2016), the RBI’s first survey started in September 2005, and only qualitative information was the focus in the first two rounds, collected from four major cities. From the third round in 2006, quantitative information (3 months ahead and 1-year ahead) from 12 cities started being collected. Since the 30^th round in December 2012, data are being collected from 16 cities. Since the data would have taken some time to stabilize, it may be appropriate to use these data after 2008 for drawing relevant empirical inferences.

5. Monetary Policy Report, RBI, April 2019.

6. In India, unlike household inflation expectations (assumed as expectations of employees) which have remained persistently higher than actual inflation, inflation expectations of professional forecasters (whose analysis may matter to firms for their investment and pricing decisions) are closer to the inflation trajectory projected by the RBI and importantly, inflation expectations of firms (as per the Business Inflation Expectations Survey of IIM, Ahmedabad) are closer to actual inflation. This experience is similar to that in Poland where anchoring of forward looking expectations of financial analysts and enterprises is found to be much higher than backward-looking inflation expectations of consumers (Lyziak 2016).

7. These data relate to non-government non-financial companies (growth in per employee staff costs) and are sourced from https://www.rbi.org.in/scripts/Pr_DataRelease.aspx?SectionID=360&DateFilter=Year and capital line database. For caveats while using this data for analysis, please see RBI’s Monetary Policy Report of October 2018. Moreover, like any nominal data on wages, these data are not adjusted for changes in productivity.

8. CPI-C back-casted data are taken from the Report of the Expert Committee to Revise and Strengthen the Monetary Policy Framework (Chairman: Dr Urjit R Patel). Such back-casted data are not available for CPI-C (excluding food and fuel). Therefore, wherever required, CPI-IW (excluding food and fuel) has been used as a proxy measure of underlying inflation.

Additional information

Notes on contributors

Sitikantha Pattanaik

Sitikantha Pattanaik is an Adviser in the Department of Economic and Policy Research (DEPR), Reserve Bank of India, Fort, Mumbai, 400001 (E mail: [email protected]).

Silu Muduli

Silu Muduli is a Manager in the Department of Economic and Policy Research (DEPR), Reserve Bank of India, Fort, Mumbai, 400001.

Soumyajit Ray

Soumyajit Ray is a Research Fellow in the Ministry of Finance, Government of India. He worked earlier as a Research Intern in the Reserve Bank of India.