THIS DISCUSSION PAPER CONTRIBUTES TOWARDS THE FOLLOWING SUSTAINABLE DEVELOPMENT GOALS:

important note: you can click on the graphs to expand them. This will also be mentioned for each graph within the main text

EXECUTIVE SUMMARY:

COVID-19 had a profound impact on global economic output, resulting in a contraction of approximately 3% in world GDP in 2020. Furthermore, the impact was not only significant but also unevenly distributed, with certain nations being more severely affected than others. While the effects of COVID-19 on economic performance are well-documented on both global and national levels, one may ask: what effect did COVID-19 have on a regional scale?

In this analysis, we develop a simple model to understand the impact of the COVID-19 pandemic on the private economic output of different municipalities in Bahrain. To the authors' knowledge, this is not only the first study to explore the municipal-level effects of the COVID-19 pandemic on economic growth but also the first to estimate private economic activity at the municipal level for the Kingdom of Bahrain. As of 2022, our results indicate that most municipalities had returned to their pre-pandemic levels of economic activity. However, when compared to pre-pandemic economic trends, the results show that the Muharraq and Northern Municipalities were the most affected compared to the rest of the economy.

Introduction | Why look should we care about the Municipality level impacts

The COVID-19 pandemic had a profound impact on global economic output, resulting in world GDP contracting by approximately 3% in 2020. To further illustrate the severity of the pandemic, we can compare the effect of the 2020 COVID-19 crisis to that of previous recessions, using the World Bank's definition of a recession (Kose, Sugawara, & Terrones, 2020).

The implications of the COVID-19 pandemic on global growth are highly evident. Furthermore, if we examine the changes in economic activity across different nations, we see that some countries "suffered" significantly more than others. By using the third quartile of real GDP per capita in 2019 as a measure of "high-income" nations, we can analyze how real GDP (in absolute terms, not per capita) changed between 2019 and 2020 among these high-income nations.

The decline in global economic activity during the COVID-19 pandemic was the most severe on record when measuring the change in GDP from 2019 to 2020. Prior to this, the global financial crisis had caused the most significant recession, whereas earlier recessions typically resulted in a "slowdown" in growth rather than an outright contraction in real GDP.

From the graph, we observe a wide range of GDP growth rates. In the unusually low GDP growth category, nations such as Macao, the Northern Mariana Islands, the Turks and Caicos Islands, Aruba, and the Bahamas experienced GDP declines of over 20%, with Macao seeing a decline of 54%. Upon closer inspection, we find that a significant proportion of these nations' GDP is derived from the services sector. When we plot the change in GDP from 2019 to 2020 against the percentage of GDP from services for each nation, a relationship emerges between the two, particularly among high-income nations.

Page 2

The comparison of high-income nations reveals that different countries were affected differently by the impacts of COVID-19. The graph above shows a modest correlation, with nations whose economies are more service-oriented appearing to experience higher contraction rates. How does this relationship look for the Gulf Cooperation Council (GCC) nations? Below is a graph that plots GDP growth between 2019 and 2020 for each GCC nation:

Even within the GCC region, there is notable variation in the economic outcomes of different states. Kuwait "suffered" the most in terms of changes in economic activity, while Oman fared the best among the GCC nations. However, all GCC countries experienced a contraction in GDP growth rates.

While the above analysis shows that differences exist among nations, even within similar income ranges and regions, one limitation is that each country implemented different policies to handle the COVID-19 pandemic. This variation makes it difficult to compare how different economic structures would have been impacted under similar conditions. Therefore, analyzing the economic effects within a particular country or region, where a consistent set of COVID-19 policies and restrictions were applied, would allow for a clearer understanding of how such areas (or even cities) were affected by the pandemic.

Hill et al. 2012 studied the economic resilience of various regions through case studies of U.S. metropolitan areas, finding that the effects of shocks can vary significantly from region to region. Fayrer, Sacerdote, and Stern 2007 found that the decline in steel and auto manufacturing had large post-shock effects on employment in certain U.S. regions. More recently, in the context of COVID-19, Wilson et al. 2020 found that the pandemic had varying employment impacts across London boroughs. Overall, the literature suggests that a shock can have different impacts on different regions.

Our aim in this study is to examine the impact of COVID-19 on different municipalities in the Kingdom of Bahrain (hereinafter referred to as Bahrain). We hypothesize that even within our nation, which is smaller than the Greater London area, different municipalities have varying concentrations of industries, leading to different effects from COVID-19. It is also reasonable to assume that COVID-19 policies across Bahrain were the same or highly similar, with comparable levels of enforcement.

Page 3

Our study will accomplish the following: 1) Create a pre-COVID-19 estimation of the economic structure and its evolution over time for each municipality; 2) Examine the impact of the COVID-19 shock on each municipality; and 3) Explore potential policy recommendations and outline further research to provide a more comprehensive policy framework.

To our knowledge, this is the first set of estimates established for municipality-level private economic activity in the Kingdom of Bahrain. With our simple estimation procedure—which can be refined into more sophisticated methods in the future—such datasets will enable us to conduct a significant number of econometric and economic simulation studies, unlocking valuable insights into the behavior of the Bahraini economy.

Data and Estimation Procedure | Data

We use two datasets to conduct this study: the first is business registration data, and the second is national accounts data. The business registration data was sourced from the older version of Sijilat’s “information request” system, which allows us to pull records of all commercial registrations (CRs). While this dataset is not an “official” published dataset, it offers a platform that: A) shows the opening and status of every business CR since the 1960s, and B) can be filtered by municipality and economic activity type.

The advantage of using this dataset is that it provides records at the branch level, allowing us to account for one business with multiple branches in different municipalities. However, a disadvantage is that the dataset may contain missing or incorrect entries, which we attempt to address. After cleaning our dataset, we remove observations with missing or incorrect entries, resulting in 292,253 branch observations. This enables us to calculate the number of active businesses during any given period in this study.

Our second dataset is the national accounts dataset, which enables us to use GDP statistics from a value-added perspective. Since the national accounts statistics loosely follow the ISIC industry classification—which the Sijilat dataset is based on—we can match the business dataset to the relevant sectors found in the national accounts data.

Our dataset spans from Q1 of 2010 to Q2 of 2022. In our estimations, we focus on quarterly private output rather than aggregating it to yearly private output for each quarter (i.e., the sum of the past four quarters during that time).

Data and Estimation Procedure | Estimation Method of Municipality Private GDP

Our method of calculating GDP is relatively straightforward. Fist, we measure what we refer to as “Output per branch”, which can be expressed as:

The output per active branch in each sector (oa) is defined as the total output of a sector (O) divided by the total number of active branches in that sector (A). We define a branch as "active" as long as the company continues to operate and exist during the given period and is not classified under other statuses such as "deleted by law" or "under mortgage."

Then, using our output per active branch, we can measure the output for each municipality and sector within that municipality via the following expression:

Which states that the output by municipality and a given sector is the total number of active branches within a given municipality & given sector, multiplied by the output per sector as earlier defined in 1.1. Then, we can measure the total output by municipality via:

Where 1.2 shows that the total output in each municipality is simply the sum of all the given sectors in each municipality and their respective output. Our simple method allows us to give us an approximation of Private GDP by each municipality.

Page 4

Pre-COVID | Municipality Economic Output and its Evolution

To start our analysis, first let us look at a snapshot of the overall Private Economy output per quarter. Below is a breakdown by municipality at Q4 2010.

In Q4 2010, private quarterly output was 2,279 million BHD. Approximately 58% of the overall output came from the Capital Municipality, while the remaining municipalities accounted for nearly equal shares of the total private economic output, with Muharraq having slightly lower shares than the Southern and Northern municipalities. However, how has quarterly output grown over the past nine years (prior to the COVID-19 shock) across the different municipalities? Below is a graph that plots the growth rates of the various municipalities.

Between Q1 2010 and Q4 2019, the overall private economic output per quarter grew at an accumulated growth rate of 37%. In comparison, the Muharraq and Northern municipalities experienced the highest growth rates of 71% and 62%, respectively, while the Capital and Southern municipalities grew by 28% and 26%, respectively. The next question we should consider is: how did the shares of each municipality change by Q4 2019? Below is a breakdown of the overall private quarterly GDP by municipality in Bahrain.

Overall, private GDP output per quarter increased to 2,763 million Bahraini Dinars, representing approximately a 21% increase in private economic activity over the past nine years compared to Q4 2010. This translates to about a 2% increase per year in quarterly private economic activity. From the chart above, we see that the Capital Municipality experienced a 3% decline in its share of overall private GDP activity, while the Southern Municipality saw a 2% decline. In contrast, the Northern Municipality experienced approximately a 3% increase in its overall private GDP per quarter, with Muharraq showing about a 2% increase.

While the above information illustrates how overall economic activity has changed across the years for each municipality, it is more important to understand how each municipality differs in terms of its economic structure. First, let us examine the “broad structure” by aggregating the economic activity of each municipality in terms of the share of their overall output. On the next page is the breakdown for Q4 2010:

Page 5

In Q4 2010, when examining overall private economic activity, we find that the economy is slightly more specialized in industrial activities (54%) compared to services (46%). However, significant differences exist across the various municipalities. For instance, the Southern Municipality is highly concentrated in industrial activities, with a distribution of 73% for industry versus 27% for services. This is expected given the presence of large industrial operations, such as Alba in Askar and Bapco around Awali and other areas of the Southern Municipality.

Additionally, the Northern Municipality, though at a smaller scale, shows a notable concentration in agricultural-related activities compared to the other municipalities. Lastly, the Capital Municipality is more focused on service-related output activities, which is also expected due to the high concentration of financial services in this area.

The question then becomes: how does the broad concentration appear as of Q4 2019? On the right is the ultra-broad classification breakdown for the overall private economy and each municipality.

In the overall private economy, we observe that the economy is now roughly equally split between services and industrial activities, with the share of services increasing by 4% and the share of industrial activities decreasing by 4%. When examining the different municipalities, we find that the Southern Municipality experienced the most significant change, with a 12% increase in its share of services and a corresponding 12% decrease in industrial activities. Additionally, the Capital Municipality is the only area that shows relatively similar shares to those in Q4 2010.

While this illustrates the overall ultra-broad categorization, we are also interested in a more detailed classification of industries. Instead of merely showing the shares of each municipality, it would be particularly useful to examine how relatively “specialized” a municipality is in a particular sector. Below, we introduce the “location quotient” measure:

Page 6

The location quotient indicates the relative concentration of output for a given sector in each municipality. It considers how concentrated the output of a specific sector is within each municipality (the Numerator) compared to the overall concentration of that sector's output in the economy (The denominator). A figure of one indicates an exact concentration relative to the economy, while a value greater than one signifies a higher concentration, and a value lower than one indicates less concentration. Below is a table showing the location quotient for each municipality and sector as of Q4 2010:

From the location quotient in Q4 2010, we observe that the Capital Municipality is highly concentrated in financial corporation activities (1.41), which aligns with its higher share of service-related output seen in the ultra-broad classification. Additionally, we note relatively higher levels of concentration in real estate and trade activities (1.16).

Moving to the Muharraq Municipality, we find that it has a significantly higher concentration in manufacturing-related activities (1.38). Additionally, the municipality shows higher levels of concentration in construction (1.20), transport & communication (1.15), trade (1.18), and hotel & restaurant activities (1.10).

In the Northern Municipality, there is a notably higher concentration in manufacturing (1.50), construction (1.65), electricity & water activities (1.26), and agriculture & fishing activities (2.52). The high concentration in agriculture & fishing is expected, given that the Northern and Northwestern areas of the Kingdom feature a fertile strip (Supreme Council for the Environment, 2012).

Lastly, in the Southern Municipality, we observe a significantly higher concentration in mining & quarrying (1.51), manufacturing (1.26), and electricity & water-related activities (1.92). This is largely due to the significant presence of BAPCO and ALBA in the Southern Municipality district.

While this provides an understanding of the specialization of various economic activities as of Q4 2010, let us examine how these activities evolved over time by Q4 2019. We will compare the data from Q4 2010, Q4 2015, and Q4 2019.

First, examining the Capital Municipality, we see that it has experienced a relative increase in its concentration of mining & quarrying activities, with an increase in the location quotient of 0.10 in 2015 and a total increase of 0.19 by 2019. This shift has brought the overall mining & quarrying activities of the municipality from a slightly less concentrated level to having some relative concentration in this area (0.95 in 2010, 1.04 in 2015, and 1.14 in 2019).

It is important to note that an increase in this concentration does not necessarily indicate a rise in the number of mines and other physical drilling activities; it could also reflect support-related activities, such as headquarters, offices, and other technical support services. Furthermore, we observe that financial corporations have seen a relative decrease in their concentration (0.06 by 2015 and 0.12 by 2019). However, the financial corporation industries remained highly concentrated as of 2019, with a location quotient value of 1.29. Lastly, there has been a relative decline in the location quotient for hotels & restaurants, decreasing by 0.09 and 0.12 by 2015 and 2019, respectively, bringing this sector into the “less concentrated” range for the Capital Municipality.

Looking at the Muharraq Municipality, we see that financial corporations have experienced an increase in their location quotient by 0.19 and 0.31 by 2015 and 2019, respectively. This brings the overall concentration from “significantly less concentrated” (0.55) to “slightly less concentrated” (0.86), indicating a significant increase in financial corporation-related activities in the municipality.

Page 7

Conversely, manufacturing facilities have seen a decline in their location quotient figures of 0.20 and 0.26 by 2015 and 2019, respectively; however, this sector continues to remain concentrated in the municipality. Lastly, we observe that electricity & water activities have experienced a substantial decline in their concentration, decreasing by 0.26 and 0.45 by 2015 and 2019, respectively, which brings the overall figure from being “slightly concentrated” (1.02) to “significantly less concentrated.”

Next, examining the Northern Municipality, we see that mining and quarrying activities experienced a decrease of 0.22 and 0.19 by 2015 and 2019, respectively. This brings the overall location quotient level from “less concentrated” (0.76) in 2010 to “significantly less concentrated” (0.57) in 2019. Additionally, construction activities have seen a relative decline of 0.01 and 0.23 by 2015 and 2019. However, this sector remains “highly concentrated” in this municipality, with a location quotient of 1.65 in 2010 decreasing to 1.42 in 2019. Conversely, hotels and restaurants have experienced an increase in concentration of 0.07 and 0.14 by 2015 and 2019, bringing the overall concentration level from “slightly less concentrated” (0.93) in 2010 to “slightly concentrated” (1.07) in 2019.

Lastly, we examine the Southern Municipality. We see that the Southern Municipality has experienced a decrease in its location quotient level for mining & quarrying activities, with declines of 0.25 and 0.47 in 2015 and 2019, respectively. This shifts the Southern Municipality from being “highly concentrated” (1.51) in 2010 to “slightly concentrated” (1.04) in 2019. Overall, this decline in concentration may be attributed to a combination of an increased presence of branches in other municipalities and a decrease in overall output per branch for mining and quarrying activities due to the oil glut of 2014/2015, with little change in the number of branches in the Southern Municipality.

In contrast, when examining hotels & restaurants, we find that the location quotient has increased by 0.22 and 0.35 by 2015 and 2019, respectively. This brings the overall sector from being nearly as concentrated as the overall economy (0.98) in 2010 to “highly concentrated” (1.33) in 2019. Finally, in agriculture & fishing activities, we observe an increase of 0.58 and 0.34 by 2015 and 2019. This elevates the overall concentration level from “significantly less concentrated” (0.73) in 2010 to “slightly concentrated” (1.08) in 2019.

We observe that the overall share of private GDP output has changed during the pre-COVID measurements over the past nine years, primarily due to the varying accumulative growth rates of different municipalities. The Northern and Muharraq municipalities experienced growth rates higher than the overall economy, while the Southern and Capital municipalities saw significantly lower accumulative growth rates.

Consequently, we observe that the shares of industrial and service-related activities have also shifted. This indicates that growth is not solely attributable to changes in overall economic activity levels across various sectors; it also reflects a change in the composition of each municipality, influenced by a potential combination of new branches opening and the economic output of specific sectors. Furthermore, the relative concentration of each detailed sector has evolved over time, highlighting a shift in the relevance of each economic activity. While this analysis illustrates the economy's state before COVID-19, what does the economy look like during and after the COVID-19 shock?

Page 8

COVID-19 Shock

To begin our analysis, we examine the following items: first, we analyze how GDP changed between Q1 2010 and Q2 2022, and then we specifically look at how GDP evolved between Q4 2019 and Q2 2022. Next, we create a "counterfactual" economy to estimate a pre-COVID trend in the number of active branches and productivity levels for each sector before the COVID shock. This allows us to establish a "Pre-COVID" GDP trend for each municipality (and by sector). Finally, we compare the overall economy and each municipality's GDP to this "Pre-COVID" GDP trend from Q4 2019 to Q2 2022.

First, let's examine the accumulated growth rates. Below are the accumulated growth rates for the different municipalities and the overall economy from Q1 2010 to Q2 2022:

We observe that the overall private economy experienced a 2% reduction in its accumulated growth rate compared to its Q4 2019 level. This is closely followed by the Southern and Capital municipalities, with reductions of 3% and 2%, respectively. The municipality with the largest decline in growth rate is Muharraq, which saw a 7% drop in its accumulated growth rate from Q4 2019 when comparing the period from Q1 2010 to Q2 2022. The only municipality that did not experience a change in its accumulated growth rate is the Northern municipality, where the accumulated growth rate from Q1 2010 to Q2 2022 remained the same as the accumulated growth rate from Q1 2010 to Q4 2019.

However, how do the overall growth rates compare between Q4 2019 and Q2 2022? Below is a chart that illustrates the changes in the overall private economy and the growth rates of individual municipalities:

When comparing to Q4 2019, we observe that most municipalities (and the overall economy) had nearly recovered to their initial starting point by Q2 2022. The overall economy, however, was still 1.69% below its pre-pandemic level. By municipality, the Northern Municipality had almost fully recovered, with a minor GDP decline of 0.32%, followed by the Capital Municipality (1.33% decline) and the Southern Municipality (2.37% decline). Muharraq Municipality experienced the largest decline, with a 4.14% drop in its accumulated GDP growth rate as of Q2 2022 compared to Q4 2019. Overall, as of Q2 2022, the private economy has largely recovered to its pre-pandemic position.

However, what is more important is to examine how each municipality compares to its previous "trend path." First, let's look at the output gap for the overall economy on the next page:

Page 9

The output gap for the COVID-19 pandemic is highlighted in orange. Prior to the pandemic, the overall output gap was nearly 0% (at 0.05% in Q4 2019). However, during the pandemic, the output gap increased to 11.14% by Q2 2020. As of Q2 2022, the output gap has decreased to 4.56%, indicating that the economy is still lagging behind its trend path. When examining how each municipality compares to their respective trend paths, the graph below illustrates these differences.

When considering each municipality's trend, we observe that the Northern and Muharraq municipalities have "suffered" the most, with output gaps of 10.65% and 11.33%, respectively, as of Q2 2022. In contrast, the Capital municipality has shown greater resilience, with its GDP now 0.55% above the overall trend. The differences in the private economy output gap across municipalities are due to a combination of the concentration of specific industries and the growth of active business units in each municipality compared to their trends. Since the productivity rate for each sector is generally consistent across all municipalities, it does not independently account for the differences in the economic trajectories and current activity of each municipality.

Overall, we can take a closer look at which industries are driving the output gap for the Northern and Muharraq municipalities. First, let's examine the Muharraq municipality. Below is a graph showing the three top sectors contributing to the output gap:

In the Muharraq Municipality, the mining and quarrying sector has contributed the most to the output gap. Interestingly, the location quotient for this industry experienced the second-highest change after the finance industry. As of 2019, the mining and quarrying industry had a location quotient of 0.96, indicating it was almost in line with the overall economic concentration. By Q2 2022, the transport and communication sector contributed 2.32% to the total 11.33% output gap. Notably, this sector was slightly concentrated in Muharraq by Q4 2019, and while its location quotient declined, it remained an important part of the local economy compared to sectors like financial corporations, electricity and water, or agriculture and fishing. Lastly, the manufacturing sector contributed 1.60% to the overall output gap. As discussed earlier, the manufacturing sector is considered "concentrated" in Muharraq when analyzing its location quotient.

Page 10

Now, let’s examine the Northern Municipality to identify which sectors contributed to the overall output gap:

By Q2 2022, the financial corporations, transport and communication, and social and personal services sectors accounted for the majority of the output gap of 10.65%. Notably, financial corporations contributed 5.01% to the overall output gap, which is surprising given their relative concentration compared to the overall economy is significantly lower. Further analysis is needed to explain why financial corporations constitute such a large portion of the output gap. The transport and communication sector contributed 1.56% to the overall output gap, which is more understandable, as this sector has a significant concentration in the Northern Municipality. Lastly, the manufacturing sector accounted for 2.27% of the overall output gap, which aligns with the fact that manufacturing is highly concentrated in the Northern Municipality.

How do the Northern and Muharraq municipalities compare to the rest of the economy (ROE)? on the right is the decomposition of the output gap for the "rest of the economy," which consists of the Southern and Capital municipalities:

When examining the rest of the economy (ROE), we see that "all other sectors" contribute to the ROE being above the trend path by 3.74%. However, three familiar sectors—transport and communications, manufacturing, and social and personal services—contribute to the ROE falling below the trend path by 4.94%. The overall net effect is that the ROE is below the trend path by 1.20%. Notably, all municipalities appear to be affected by the transport and communications sector, while the social and personal services sector is common across most municipalities, with the exception of the Muharraq Municipality.

Policy Recommendations

Before we begin our policy recommendations, let’s recap the overall results, starting with the pre-COVID findings. Over the past nine years, the overall share of private GDP output has changed due to varying accumulative growth rates among the different municipalities. The Northern and Muharraq municipalities experienced higher growth rates compared to the overall private economy, while the Southern and Capital municipalities saw substantially lower accumulative growth rates. Additionally, we observed changes in the relative concentration of industries across the various municipalities.

Page 11

Looking at the COVID-19 shock, we saw that from an overall accumulated growth level, different municipalities were impacted in significantly different ways, where the private economy had seen a 2% reduction in it’s overall accumulated growth rate, and that the Muharraq municipality saw a 7% decline, with the Northern Municipality recovering back towards their initial accumulated growth rates when comparing the accumulated growth rates of Q1 2010 to Q4 of 2019 to that of the Q1 2010 to the Q2 of 2022. We also saw that the private economy, as of Q4 2022 is still below it’s trend level, and that when comparing to the output gap for each municipality, we see that the capital municipality is slightly above their trend trajectory, with the Muharraq and Northern municipalities being significantly lower than their trend. We also saw that the drivers behind this are due to a common set of industries, being the transport & communications industry, the social & personal services industry, and the manufacturing industry affecting each municipality, but in varying degrees.

The implication of this is that, consistent with the literature, different municipalities are indeed impacted by economic shocks in varying ways due to their distinct industrial compositions. Additionally, our basic model indicates that the levels of active businesses grow at different rates across municipalities. Therefore, it is crucial to implement policies that aim to better insulate municipalities from economic shocks.

In reviewing the literature, we reference two key papers to explore potential policy recommendations. Kim, Lim, and Colletta (2022) examine how COVID-19 affected two groups of industries: those classified as “essential,” which have low levels of interpersonal interaction, and those deemed “non-essential,” characterized by high levels of interpersonal interaction. Their findings indicate that states specializing in essential industries with limited interpersonal interactions were more successful in maintaining higher resistance levels throughout most of the pandemic. Conversely, states that specialized in non-essential industries with more intensive interpersonal interactions experienced a decline in their resistance to economic shocks during the pandemic.

Martin and Gardiner (2019) examine the resilience of British cities to major economic shocks. The authors identify a distinct shift in the relationship between resistance and recovery in response to different types of shocks. Additionally, they find that the divide between Northern and Southern UK cities has increased over time. Their research suggests that variations in resilience to major shocks can significantly influence the long-term growth trajectories of cities.

Overall, the literature suggests that to enhance resilience against economic shocks, municipalities should aim to specialize in more “essential” industries rather than “non-essential” ones. However, this specialization may come at the cost of lower overall growth, as different major UK cities have experienced varying recovery outcomes from different shocks. This observation partially explains the results seen above, where the Capital Municipality's strong specialization in financial corporation industries may account for its lower growth rates, while simultaneously contributing to its apparent resilience during the COVID-19 shock.

The overall policy recommendations from this paper suggest that efforts should focus on fostering recovery in the Muharraq and Northern municipalities. This recovery should aim to increase the share of “essential” industries rather than “non-essential” industries, thereby enhancing resilience against future economic shocks. However, this approach may present a trade-off, as boosting resilience could potentially slow down economic growth in these municipalities, which may, in turn, affect the overall growth of the private economy. To address this complexity, further research—particularly literature reviews and the development of economic models such as Computable General Equilibrium (CGE), Dynamic Stochastic General Equilibrium (DSGE), and Input-Output (IO) models—could provide deeper insights into how to enhance the resilience of each municipality and whether a trade-off exists between “growth” and “resilience.”

Page 12

Limitations of the Study

While the above study offers a solid starting point for expanding our understanding of the economies of different municipalities, several limitations should be acknowledged. First, the data used is not seasonally adjusted, as the initial work on this piece did not account for the seasonality of the data regarding both GDP and business units. Future iterations of this research could implement seasonal adjustments, allowing for a comparison to determine the extent to which seasonally adjusting the data affects the results.

The second limitation is the quality of our dataset concerning business units. Since we utilized a manually collected dataset from Sijilat, there is a potential for incomplete or inaccurate data within the dataset. Although every effort was made to clean the data and make necessary adjustments, the accuracy of the study could be enhanced by using an official and thoroughly cleaned dataset.

The third limitation pertains to the assumption of productivity. In our analysis, we assume that each business unit shares the same level of productivity. However, firms are not perfectly competitive, particularly in industries that exhibit more monopolistic characteristics. For instance, a small cold store selling SIM cards would not have the same productivity as branches of larger companies like Batelco, STC, or Zain, which are likely to achieve higher levels of output per branch.

The fourth limitation is the exclusion of government economic activity. Since we lack a mechanism to capture government activities and allocate them to specific municipalities, the study focuses solely on the "private economy." Additionally, this study assumes that businesses and overall economic activity occur in isolation within their respective municipalities. For instance, a construction company in the Northern Municipality could deploy its equipment and personnel to the Southern Municipality for a specific project. Consequently, we do not have a method to account for what could be termed the “isolated municipality” assumption.

Conclusion

This study examines the effects of COVID-19 on the municipalities of the Bahraini economy. We introduce a simple yet effective method for allocating private GDP across the Capital, Muharraq, Northern, and Southern municipalities. To our knowledge, we have created the first estimates of private GDP by municipality and measured how it has evolved over time. Our findings indicate that the Muharraq and Northern municipalities were the fastest-growing prior to COVID-19, and that post-COVID-19, these municipalities were arguably hit harder than others due to their rapid growth trend before the pandemic.

Additionally, we briefly explored the literature for policy recommendations, suggesting that the Bahraini government should consider integrating more “essential industries” in all municipalities. However, this may come at the cost of higher growth rates in the future. More studies are needed to identify which industries to target for creating more resilient municipalities and to determine whether a trade-off between resilience and growth exists, as indicated by our literature review.

Page 13

Featured

Baland Al Rabayah- Director

Baland Al Rabayah is the co-founder and Director of BRG. He currently holds a Diploma in Accounting and Finance from Bangor University, a Bachelor of Science (BSc) in Accounting and Finance from Bangor University, and a Masters degree in Applied Economics from the University of Strathclyde. Baland will begin his PhD studies at the University of Strathclyde in 2022 looking at the economic effects of renewables on the GCC economies (energy transition). Baland’s interests are in Environmental, Energy, Public Policy, and Health Economics.

email: b.rabayah@bhres.group

Feyrer, J. et al. (2007) Did the Rust Belt Become Shiny? A Study of Cities and Counties That Lost Steel and Auto Jobs in the 1980s, JSTOR. Available at: https://www.jstor.org/stable/25067440.

Hill, E. et al. (2012) ‘Economic shocks and regional economic resilience’, in. Brookings Institution Press, pp. 193–274.

Kim, A., Lim, J. and Colletta, A. (2022) ‘How regional economic structure matters in the era of covid-19: Resilience capacity of U.S. states’, The Annals of Regional Science, 70(1), pp. 159–185. doi:10.1007/s00168-022-01134-w.

Kose, M.A., Sugawara, N. and Terrones, M.E. (2020) ‘Global recessions’, SSRN Electronic Journal [Preprint]. doi:10.2139/ssrn.3535972.

Martin, R. and Gardiner, B. (2019) ‘The resilience of cities to economic shocks: A tale of four recessions (and the challenge of Brexit)’, Papers in Regional Science, 98(4), pp. 1801–1833. doi:10.1111/pirs.12430.

Supreme Council for the Environment, S.C. for the E. (2012) Bahrain’s Second National Communication, UNFCC. Available at: https://unfccc.int/resource/docs/natc/bhrnc2.pdf.

Wilson, T. et al. (2020) Impact of the COVID-19 lockdown on Central London: A presentation to Central London Forward, Central London Forward. Available at: https://centrallondonforward.gov.uk/wp-content/uploads/2020/07/CLF-Downturn-Presentation-IES-14-07-20-FINAL.pdf.

world bank, world bank (2024) World Bank Open Data. Available at: https://data.worldbank.org/ (Accessed: 07 October 2024).