Embracing the melody of dissent: A symphony in policymaking

In recent times, members of the Monetary Policy Committee of the Reserve Bank of India, have expressed views that are contrary to the MPC’s official stance. This article investigates the influence of dissent within the Committee on forecasts of professional forecasters regarding growth and inflation. It finds that such discordance enhances forecast accuracy, suggesting that fostering an environment conducive to nuanced opinions could improve policy outcomes.

As the former Reserve Bank of India (RBI) Governor once quipped, “Setting monetary policy was like juggling six balls – webs within webs – balls connected with elastic; you throw one up a little bit more and the other balls go out of string.” The International Monetary Fund’s World Economic Outlook for April 2024 highlights the global economy's gradual approach to a soft landing, emphasising the importance for central banks to ensure a smooth descent in inflation by avoiding premature easing or prolonged delay in policy adjustments. In India, policymakers face a similar challenge. In the past, traditional monetary policy tools alone have often faced limitations in effectively steering the economy. During times of economic uncertainty, relying solely on interest rate adjustments or quantitative easing measures might not yield the desired outcomes. Against this backdrop, one of the prominent views that emerged to alleviate this conundrum is the strategic use of communication.

The importance of central bank communication

Empirical evidence supports the idea that effective communication by a central bank enhances the impact of monetary policy. Macroeconomists have emphasised that when everything else fails, communications that influence expectations are able to affect inflation and boost economic activity (Eggertson and Woodford 2003). Blinder and Morgan (2005) discuss the two main functions of central bank communication, which include broadcasting news and reducing ‘noise’. However, when many voices speak together, it is interesting to see whether it leads to a cacophony or symphony for market watchers. Recently, in the case of India, we have observed central bankers dissenting in their communication, diverging from the voices of the majority to put forth a contrarian view. A closer examination of recent monetary policy statements by the RBI reveals a divergence of voices within the sync of symphony. Jayanth Varma, an external member of India’s Monetary Policy Committee (MPC), in a recent interview advocates against unwarrantedly high real rates, citing significant costs to the economy – a departure from the popular choice. Further, he underscores that “words must match actions”. It is therefore intriguing to examine this aspect of central bank communication in detail. To this end, we seek to explore the impact of dissent and examine the channels through which it contributes to improved policymaking.

In an earlier I4I article, Mahambre and Pathak (2022) argue, based on data from 30 RBI-MPC meetings, that the statement tone and other communication characteristics of external and internal members have differed over the period of time. They further suggest that the change of the RBI governor appears to have a greater influence on the statements of external members. In a previous study (Sil et al. 2023), we examined another noteworthy aspect of communication: the coherence of the actions and stances of MPC members. Analysing minutes of all RBI-MPC meetings, we developed two innovative measures of implicit dissent, both at the individual and group levels. The proposed measures sought to detect latent discord present in the minutes of the meetings, which otherwise seem to portray explicit consensus in voting. For instance, a particular meeting may have the consensus of all members in favour of a specific stance, however once contrasted against their respective minutes, divergence of views within the committee becomes apparent. The presence of such discordance indicates, and is a testimony to, the existence of dissenting opinions amongst committee members. Through these measures, we investigate the influence of dissent or discord on the forecasts of professional forecasters regarding economic growth and inflation. Our empirical analysis revealed that discordance among committee members enhanced forecast accuracy, suggesting that fostering an environment conducive to nuanced opinions could lead to improved policy outcomes. Expanding this perspective globally, Banerjee et al. (2024) confirm that disagreement among policymakers across various economic aspects provides valuable signals to forecasters, aiding in the refinement of their forecasts. However, despite these insights, the precise channels through which dissent impacts monetary policymaking have remained elusive.

Analysing influence of discordance among RBI-MPC members, on economic forecasts

We construct an empirical measure of market sentiment and consumer perception in the Indian context. We track the adjustment process of both the forecasters’ as well as the consumers’ reception of monetary policy communication by central bankers. This is built into a multi-layered model, namely the policymaker’s level, forecaster’s level, and consumer’s level, and further used to estimate the transmission mechanism of how discordance of views amongst policymakers would influence the forecast accuracy of policy variables.

We initiate our analysis by constructing measures of dissent using a natural language processing tool, which captures sentiment intensity by assigning positive, negative, or neutral scores to each word extracted from the minutes of the MPC meeting. These dissent measures assess the consistency of each MPC member's stance and action and are further disaggregated to compute group-level metrics. This forms the foundation of the first layer, representing the policymakers. To assess the impact of market perception, we utilise RBI's Survey of Professional Forecasters. Across consecutive survey rounds, we compute the 'forecast dispersion', which quantifies the level of disagreement among forecasters. Additionally, we construct the 'market perception gap' to gauge surveyors' perceptions of growth and inflation. This robust measure is derived from the probabilities assigned to various outcomes of real GDP (gross domestic product) growth and CPI (consumer price index), as compared with their actual values. Together, these variables contribute to the second layer of our model.

Finally, we propose a measure that captures the inherent dynamism embedded in the perception of Indian consumers. This measure is derived from the RBI’s Consumer Confidence Survey data and it is calculated as the deviations between consumers' perceptions of inflation/growth and the actual values of inflation/growth. These deviations are tracked over quarters and mapped with the bimonthly meetings of the MPC. Additionally, we incorporate the index of economic policy uncertainty developed by Baker et al. (2016), which is based on the frequency of coverage in major newspapers. This inclusion allows us to investigate whether the introduction of an intermediate layer, based on newspaper articles, has an immediate impact on consumers' perceptions.

Dissent reduces economic uncertainty

Bringing together all layers within our model, our empirical investigation affirms that dissent among policymakers initiates a 'domino effect' of revisions. This sequence of adjustments first aligns market perceptions and subsequently enhances forecast accuracy. Moreover, our findings indicate that dissent contributes to a reduction in economic uncertainty. These findings not only emphasise the significance of nuanced perspectives among policymakers but also establish a transmission channel, highlighting the responsiveness of existing market perceptions to policymakers' decisions in India. Interestingly, we observe that professional forecasters adjust the probabilities associated with various ranges of growth and inflation forecasts. This adjustment serves to minimise the risk of opportunistic reading of data, resulting in a tighter confidence interval¹ and facilitating a more cohesive estimation process.

However, we also find that consumer perception does not appear to react to the alignment of market perception at the forecaster level. This observation may be attributed to the lack of sufficient knowledge or skill of the average Indian to comprehend the technicalities presented in reports documenting the views and stances of forecasters, or is simply lack of awareness of the existence of such materials. Consequently, they may not receive the signals embedded in the forecasters' views, resulting in an insignificant relationship. This finding challenges the perspective put forth by other scholars, such as Ehrmann and Fratzscher (2013), who argued in favour of the negative impact of divergent views of policymakers on public perception. Our study contributes significantly to the literature on the 'disagreement-uncertainty' nexus. To the best of our knowledge, it is one of the few empirical investigations into the relationship between disagreement and uncertainty in economic policymaking within the Indian context.

In conclusion, the presence of dissenting voices amongst policymakers often strikes a symphony of divergent yet pertinent views, harmonising to navigate the intricate melodies of economic policy. When there is a consensus in voting, such tacit disagreements often tend to escape plain sight. Once such latent discordance becomes evident to forecasters, it casts doubt upon their prior judgement that was based on the apparent consensus in voting. Hence, it warrants a closer re-examination of their proposed estimates of policy variables, thereby reduces the possibility of lapses in future estimation. If dissent does enhance forecast accuracy, creating an environment free from byzantine regulations and institutional goading would foster the expression of diverse, nuanced opinions that could explicitly help in correcting forecast errors. Our analyses demonstrate that a judicious presence of dissenting voices fosters a richer understanding among policymakers, leading to enhanced forecast accuracy and diminished economic uncertainty.

Views expressed are personal and do not necessarily reflect that of the authors’ employers or other associated parties, including the I4I Editorial Board.

The views expressed in this post are solely those of the author, and do not necessarily reflect those of the I4I Editorial Board.

I4I is now on Substack. Please click here (@Ideas for India) to subscribe to our channel for quick updates on our content

Note:

A confidence interval is a way of expressing uncertainty about estimated effects – specifically, it indicates the share of times that the true effect would be contained within a certain range.