Money, Inflation and the Arab Spring in Bahrain

Abstract

1. Introduction This paper investigates money and inflation in the Kingdom of Bahrain with a focus on the period of the Arab Spring (AS), 2011-13, to determine the effects of social unrest on general and relative prices. The amount of research on the economies of the Middle East and the Arab world has expanded lately, especially in light of the AS. In addition to our own work on the effect of the AS on employment and output in Bahrain (Miller & Epstein, 2015), researchers have recently investigated the effects of unrest in the Middle East in general on stocks (e.g. Abdelbaki, 2013; Chau, Deesomak & Wung, 2014; Bouri, 2013), banking (e.g. Jackowicz & Kowalewski, 2013; Kim, 2013; Thalassinos et al., 2013), unemployment (Campante & Chor, 2012), and output (Khandelwal & Roitman, 2013). The issues of money and monetary policy in Bahrain have been addressed by Jusof, Al Wosabi & Majid (2009), Darrat and Al-Sowaidi (2009), Espinoza and Prasad (2012), Hojjat & Barnhorst (2013). Narrowing the focus to the connection between unrest and prices, Dube & Vargas (2013) is representative of analyzes that consider the unrest provoked by price movements. In this paper we reverse the analysis and consider what changes in general or relative price might have occurred as a result of the unrest. To do so we have organized the paper as follows. We begin by investigating the measurement and co-movement of two inflation indicators--the GDP Deflator and the Consumer Price Index (CPI)--over the period 1990-2011, followed by an analysis of the specific movements in the two indexes to determine what nonmonetary factors may play a role in the movement of general prices both over time and in short periods of uncertainty in Bahrain. Third, we look at disaggregated monthly CPI data just before and then during the AS to determine the effect on prices in various consumer sectors of Bahrain, if any, of the onset of social upheaval. The last part sets out our conclusions. 2. The Nature and Movement of Inflation Indicators over Time To understand the movement of prices during the Arab Spring we must first determine exactly what the two dominant inflation indexes are capturing. Once done we can then track the movement of the two before and during the AS to discover any causal connection between prices and social unrest. In 2007 Tahir and Samad applied Vector Autoregression to five macroeconomic variables in the Bahraini economy, including the CPI. However, since then changes have been made in the reporting of inflation figures, which provide analysts with advantages and disadvantages. The monthly reporting of disaggregated data does allow us to look more deeply into the relation of money and prices over shorter periods of time, yet we are constrained in what we can do from a sophisticated statistical standpoint because of data limitations. While Annual Consumer Price Index figures using a consistent base year (2006) go back to 1990, the database of consistent monthly Headline Rates of consumer inflation is quite limited, covering the period since August 2007 only. Disaggregated CPI data by consumer sector is even more limited, dating back to January 2009. Data on the GDP deflator on an annual basis go back to 1989, but the base year was changed in 2011 from 2001 to 2010. The transition to a new base year was coupled with a redefinition of nominal GDP by the Central Informatics Office (CIO), but recalculated time-series data were limited to the period since 2006 only. Thus, to create a consistent time series of deflator values for the period 1990-2013 we had to couple the actual deflator data reported under the 2001 base year for 1990-2006 with an imputed value of that deflator for the period 2007-13. We did so by applying each additional year's GDP inflation in the new 2010 series to the previous year's deflator based on 2001. The resultant time series of annual CPI and Deflator data appear in the appendix. …