Development of regional input-output tables for Northern Ghana: An analysis using location quotient methods

Abstract

We developed regional input–output tables for Northern Ghana using the latest 2018 national input-output table for Ghana and the 2021 Ghana population and housing census data. Four variants of the location quotient concept, based on employment data for 17 industries, were used for the derivation of output impact multipliers. A comparative output analysis of the results was undertaken using the derived regional multipliers compared to those generated from analysis involving national data. The Round location quotient (RLQ) was the best of the four location quotients assessed for the regionalization of the national input–output table. The results of the analysis using the RLQ are consistent with evidence from secondary sources dealing with Northern Ghana based on economic output impacts.

Keywords:

• Ghana
• input–output analysis
• location quotients
• regional economics

Reviewing Editor:

• Michael Hardman, University of Salford, UK

Subjects:

• Rural Development
• Economics and Development
• Statistics for Social Sciences

1. Introduction

1.1. Background

The development of regional input-output (IO) tables from national IO tables has increased over the last four decades (Flegg et al., 2021). The increased development of regional IO tables has arisen due to the development needs of regions within nation states arising from a variety of issues such as endemic poverty and demands for autonomy status and more decentralized types of governance to reduce excessive central government control. The high costs and long delays, linked to the development of IO tables, using the method of surveys, inspired the development of IO tables for regions based on non-survey methods (Flegg & Tohmo, 2018). While non-survey methods are relatively inexpensive, they tend to generate results that may not always conform to field-based evidence. There has been continuing development of non-survey methods with this development featuring simulation and comparative analysis with data generated from surveys. Regionalizing national IO tables has not been extensively incorporated in economic impact studies in African countries for reasons such as data availability. Given the increasing population and decentralization in African countries, the demand for regional input-output tables is high.

1.2. Problem statement

Ghana is administered through 16 regions. Northern Ghana, the area of this study, is made up of five regions: Northern, North-East, Savannah, Upper East and Upper West. Northern Ghana is the poorest part of Ghana and one of the poorest areas in West Africa. Tamale is the area’s largest city and the nodal point of migration and economic activities in the area. These activities have resulted in Tamale growing quickly with population growth rate of 221.3% from 2010 to 2021 (Abubakari et al., 2022). Given its very high level of poverty, including severe food insecurity, Northern Ghana has been the subject of several Ghanaian State development interventions and other programmes undertaken by non-governmental organizations and international development agencies since Ghana gained independence in 1957. These efforts have failed to make major headway in poverty reduction (Salifu & Anaman, 2019). Poverty levels remain very high in the area; poverty rates for many tribes in the area have increased over the last decade.

A common problem of poverty-reducing State interventions is the leakages of benefits away from target areas to non-target areas. These leakages could be due to factors such as unbalanced political power structures within the nation state that negatively impact small groups and tribes, and poor infrastructure in the target areas that does not allow the full absorption of the benefits of societal investments in the areas. An economic tool that could be used to analyse the flow of benefits of societal interventions is input–output (IO) analysis using regional IO tables. Regional IO tables are normally derived from national IO tables due to the high cost of undertaking surveys to collect data for regional IO tables. National IO tables are developed every five or ten years in advanced countries. Ghana had its first IO table developed for 1960 based on ten industries (Birmingham et al., 1966) using the 1960 Census data on demographic characteristics (Ghana Census Office, 1962). New IO tables were developed for Ghana for the years, 2005, 2015 and 2018.

The construction of a regional IO table for the five regions of Northern Ghana is a step to analyse the impacts of Community and State programmes in the area. Farmers’ continuous reliance on rainfed agriculture and the limited environmental capital of the region characterized by only one rainy season are drivers of poverty and outmigration of people from Northern Ghana. Input-output analysis could help in the assessment of the links between agriculture and other industries in Northern Ghana through income impacts and poverty reduction from development projects.

To the best of our knowledge, this study is the first attempt to regionalize the Ghana national IO table for Northern Ghana using location quotients. Given the high poverty rates in Northern Ghana, the regionalization work pursued in this study is justified based on the need to improve the measurement of benefits and impacts of societal projects. The main objective of this study was to generate regional IO tables for Northern Ghana using the 2018 Ghana national IO table and the July 2021 National Census employment data. The remainder of the paper is structured as follows: the summary of the literature review is reported in the next section. The theoretical framework and analytic methods are outlined next. The results of the analysis are reported in the fourth section of the paper and are followed by a discussion of these results. The conclusions and policy recommendations follow.

2. Literature review

2.1. History and political economy of governance of Ghana

2.1.1. Brief history

Ghana is situated in West Africa and has a population of about 32.5 million (March 2024 estimate) growing at a rate of 2.1% (Ghana Statistical Service, 2021). Based on fossil records, Ghana has been settled by humans for at least 30,000 years (Anquandah, 2013; Buah, 1998), possibly through movements from Eastern Africa, the earliest known source of human evolution (Dawkins, 2024). The first surviving settlers in Ghana are called Guans, and they have about 30 small kingdoms established throughout the country. Based on the July 2021 population census, the Guans constituted 3.22% of Ghana’s population, numbering about one million. The other eight broad ethnic groups in Ghana, with shares of the citizen population, shown in brackets, are Akans (45.68%), Dangmes/Gas (7.06%), Ewes (12.80%), Gurmas (6.43%), Grusis (2.71%), Mandes (1.97%), Mole-Dagbanis (18.50%) and All Others (1.63%) (Ghana Statistical Service, 2021). Members of these eight ethnic groups migrated into Ghana from other parts of Africa starting from the 11th century AD. Their 1,000-year settlement on the Ghanaian landmass produced kingdoms through State building processes such as peaceful unions of towns and city states, and also through wars, including those related to the 400-year Trans-Atlantic Slave Trade, from the 15th to 19th century; this trade in slaves aggravated existing intra-ethnic and inter-ethnic conflicts.

Ghana currently has about 180 Traditional Councils (traditional states) belonging to 89 tribes derived from the nine broad ethnic groups. Nine of the 89 tribes are relatively big with shares of the citizen population ranging from 2.9% to 14.8%; the elites of the nine big tribes extensively dominate the State institutional and governance landscape including the control and management of major State economics and financial institutions (Anaman & Bukari, 2021). The remaining 80 tribes are relatively small with shares of the citizen population varying from 0.1% to 1.8%, constituting about 40% of the citizen population. Northern Ghana is largely inhabited by people from the Grusi, Guan, Gurma, Mande and Mole-Dagbani ethnic groups. Thirty two (40%) of the 80 small tribes in Ghana originate from Northern Ghana.

Ghana was known as the Gold Coast as far back as the 15th century due to large deposits of gold discovered by Europeans. From the 15th to 19th century, there was competition among European nation-states, such as Britain, Denmark, France, Netherlands, Portugal and Sweden, along the Gold Coast area, for trade access. European states established alliances with traditional states in the Gold Coast area; for example, the Netherlands signed a friendship treaty with the Ahanta People (in the Western Region) in August 1656 that lasted till April 1872. During the 19th century, British alliances were dominant. On 6 March 1844, the British government signed a defence and security treaty with 17 small traditional states in the Gold Coast area (in the Central, Greater Accra and Western Regions). This treaty led to the formation of a combined Gold Coast army made up of soldiers from the 17 traditional states, other local personnel, army units from Britain, and black specialist field troops from military regiments in the West Indies. This combined army, equipped with more advanced military technologies and systems, was able to defeat invading armies organized by other alliances and states in West Africa during the last leg of the Transatlantic Slave Trade which effectively ended in West Africa at the end of the 19th century.

2.1.2. Political independence struggle and aftermath

These struggle for independence of the Gold Coast was started by King Aggrey of Cape Coast in 1865 and it arose from the modification of the terms of the 1844 treaty into a colonial arrangement by British officials based in Cape Coast. In 1874, the British government unilaterally declared the Gold Coast area as its colony and moved its administrative headquarters from Cape Coast to Accra due to discontent against the British arising from issues such as the introduction of a poll tax. From 1874 to 1902, the British incorporated additional states into the colony through conquest and negotiations. Much of Northern Ghana became part of the colony in 1902 based on negotiations between the British government and principal chiefs of the area.

The struggle for independence was continued by the Aborigines’ Rights Protection Society (ARPS). This organization was founded in Cape Coast in 1897 by several chiefs and intellectuals led by Lawyer John Mensah Sarbah. It was well known for its efforts in defeating the 1896 and 1897 land bills in the British Parliament; these bills would have allowed the British government to take over all “unused” lands in the Gold Coast as the property of the Colonial Government. In its successful submission to the British Government in London, in 1897, the ARPS argued that “unused” lands were a resource preserved for use by future generations of humans and they also had traditional religious values in relation to the preservation of non-human animal species; this argument is similar to the principles of bequest use and existence (non-use) values, two important concepts in natural resource economics, which emerged in the 1960s.

The ARPS was active for 40 years till the late 1930s when it became dormant. It was replaced by the United Gold Coast Convention (UGCC). The principal founder and the first President of UGCC was Mr. George Grant, a latter-day member of the ARPS, who represented the ARPS national council as its member for Sekondi. Mr. Grant was reputed to be the wealthiest businessman in the Gold Coast in the 1940s; he extensively financed UGCC from his personal resources. He registered the organization, together with the Vice President, Lawyer Robert Samuel Blay, in Sekondi, in 1947. The independence struggle was completed by Dr. Kwame Nkrumah and the Convention People’s Party (CPP); the latter party was formed in 1949 in a split from the UGCC. The country achieved independence on 6 March 1957 when the Gold Coast was renamed Ghana. On 1 July 1960, Ghana became a Republic, within the English-speaking Commonwealth, after a referendum, and concurrent presidential election, held on 27 April 1960 won by the CPP.

Ghana experienced the first successful military coup on 24 February 1966. After that coup, the country underwent a 27-year period of political instability, from 1966 to 1992, characterized by four more successful coups. There has not been a successful coup in the Fourth Republican era which started on 7 January 1993. This era has been marked by alternation of power, through quadrennial national presidential elections, by the two main political parties, National Democratic Congress (NDC) and the New Patriotic Party (NPP), after eight years.

2.2. Overview of ghanaian economic growth

During the Fourth Republican era, Ghana achieved high economic growth as measured by changes in its gross domestic product (GDP). In 2022, Ghana’s GDP was 610.222 billion Ghana cedis or 73.769 billion United States dollars (US$) based on exchange rate of 8.22 Ghana cedis per dollar (Ghana Statistical Service, 2023) making it the second largest economy in West Africa, after Nigeria. The country achieved a lower-middle-income status in 2009 when its per capita GDP exceeded 1,000 US dollars. The average growth rate of GDP over the 1993 to 2022 period was 5.4%. This growth was partially produced through severe environmental destruction. Ghanaian rivers are among the most polluted in the world due to mining activities (British Broadcasting Corporation, 2021). The proportion of Ghana’s surface water resources severely affected by pollutants due to mining was indicated as 22% in 2022 by the Minister of Water Resources in a report to Parliament (Ghana Web News, 2022a). The rate of depletion of natural forests in Ghana is among the highest in the world (Atteridge et al., 2022).

The economic growth has been linked to rapid growth of foreign and domestic debts. Total government debts were US$52.4 billion in December 2023, equivalent to 72.5% of GDP (Bank of Ghana, 2024), driven by perennial large government budget deficits. Foreign debts which stood at US$2.1 billion in December 2006 increased to US$30.1 billion in December 2023, an increase of US$28 billion in 17 years. In December 2022, the government announced a default on the payments of its debts owed to local and foreign investors. It suggested debts payment negotiations with investors linked to an agreement with the International Monetary Fund (IMF) completed in May 2023, the 17th agreement that Ghana had signed with the IMF since the 1966 coup. The debt default, the first of its kind in Ghana, was linked to a record loss of 60.82 billion Ghana cedis (10% of GDP) by the Central Bank of Ghana for 2022, and a three-decade inflation high of 54.1% for December 2022.

2.3. Income inequality and marginalization in Ghana

Another feature of the Fourth Republican era has been increasing national income inequality (Oxfam International, 2024). The Gini coefficient of national income inequality rose from 35.3 in 1987 to 43.5 in 2017 (Ghana Statistical Service, 2018). Several factors influence the growing inequality. First, there is widening income gap between Northern Ghana and Southern Ghana due to the poor environmental capital of the former related to its low rainfall which greatly limits rainfed agricultural production; the quality of the environmental capital has worsened due to climate change. Second, due to historical factors and government policy choices, the infrastructure (electricity and water supply, information and communication facilities, and transportation systems (roads)) in Northern Ghana is much less developed than that of Southern Ghana leading to relatively lower economic growth in Northern Ghana.

The third factor is the marginalization of vulnerable groups in the country. Anaman and Anartey (2023) and Adjei-Annan (2023), using Ghana Living Standards Survey (GLSS) data, identified five major vulnerable groups in Ghana characterized by high poverty rates. These are (1) rural people living in Northern Ghana, (2) disabled people, (3) followers of traditional religions, (4) female-headed rural ­agricultural households, and (5) members of small tribes. Rural households also have higher poverty rates than urban households (Ghana Statistical Service, 2018). Anaman and Bukari (2020) established that the average poverty rates of the 80 small tribes were much higher than those of the nine big tribes using GLSS data. The high poverty rates of Northern Ghana are associated with its harsh climate and its ­disproportionate shares of rural areas and small tribes.

As noted by Anaman and Bukari (2021), small tribes in Ghana have been marginalized in the allocation of State resources in the Fourth Republican era. This marginalization is exemplified by the creation of districts. The administrative district is a core entity where State resources are channelled for development. The President of the Republic has the power to create new districts and the power to appoint all Mayors and District Chief Executives (DCE). However, the constitutional requirement for the creation of new districts is a minimum of 75,000 people (Ghana Local Government Service, 2017). About 30% of the existing 261 districts were created without meeting this requirement; many such new districts were created for areas inhabited by large tribes, especially tribes of influential members of NDC and NPP. Using 2021 census data, 73 out of the 261 districts did not have the minimum population of 75,000. These included 27 of the 44 new districts created in 2017/2018 (refer to the district population data from Ghana Statistical Service, 2021, pp. 38–46).

Requests from citizens for new districts, in areas with over 75,000 people, have been continuously rejected by the government, especially so for the Central Region, the foundation region of Ghana, the fourth most populous region, and one of the poorest regions in Ghana, using both income and non-income indicators. Seven of the 80 small tribes in Ghana originate from the Central Region accounting for about 30% of the region’s citizen population. The region’s high poverty rates (outside of Cape Coast, the regional capital, and Kasoa, which is largely part of Accra, the national capital) reflect the marginalization of small tribes in the country. The marginalization includes issues related to the right to use local languages (or closely-related local languages) by members of small tribes in basic schools in their localities. This right is protected by various international human rights conventions signed and ratified by the Government of Ghana. The use of local or native languages in early years of schooling is known to increase overall student performance (refer to many studies on the subject such as Irighweferhe and Kingsley (2023)).

Overall, the 31-year-plus Fourth Republican era in Ghana could be described as a highly-unstable political settlement dominated by the elites of a few big tribes (refer to the various theories of political settlements recently documented by Kelsall et al., 2022). The instability of the political settlement has led to outcomes such as very high levels of public debts which resulted in the financial bankruptcy of the State announced by the Minister of Finance one month to the 30th anniversary of the promulgation of the Fourth Republican Constitution, in December 2022 and six IMF stabilization programmes covering a period of 19 out of 29 years, since 30 June 1995. The lack of elected Mayors and DCEs has reinforced the marginalization of small tribes leading to their reduced governance participation and has contributed to the accelerated destruction of rivers and poor sanitation throughout the country. The large increases in the numbers of skilled workers migrating out of Ghana over the last decade reflect not only the deterioration in the economy but also the perceived inability of the marginalized to change or improve upon the status quo. The current political settlement in Ghana is not sustainable for long-term inclusive development. Like the Biblical Parable of the Sower found in the New Testament Matthew Chapter 13 Verses 3 to 8, the solutions required for inclusive development involve constitutional reforms that would move the biological seed (political settlement) from the rocky surface (current unsustainable situation) to the watered and manure soils (ideal situation) which would ensure bountiful harvests for all including the marginalized and people living in Northern Ghana.¹

2.4. Review of IO analysis

An IO analysis involves the use of national tables that establish the interconnectivity among industries, households and government organizations (Read et al., 2020). The transactions table is a record of the creation and disposal of the goods produced by the economy during a specific period of time. Table 1 is an illustration of the input–output table of Ghana for 2018 where there are 17 domestic industries. Each industry represents the aggregation of activities of thousands of firms in a particular industry at a given period of time. The four quadrants of the economy are (1) the domestic industries or producing quadrant, (2) the payments quadrant, (3) the final demand quadrant, and (4) the payments-to-final demand quadrant. The producing quadrant, represents the transactions among the 17 producing industries. Each industry requires inputs from itself and the other 16 industries to produce one unit of total output. The payments quadrant represents the services of workers, capital owners and government, through indirect taxes, and imports which are required by each of the 17 domestic industries to produce one unit of total output. Indirect taxes are those paid to State entities; these include sales taxes and import duties.

Table 1. Illustration of the 17-industry Ghana National Input–Output Table.

Industry	Outputs				Household consumption	Final demand	Total inputs
Inputs	1	2	3	… 17
Agriculture Mining Manufacturing Electricity, gas and steam Water supply and sewage Construction Wholesale and retail trade Transport and storage Hotels, restaurants and food services Information and communication Finance and insurance Real estate activities Business services Public administration Education services Health and social work All other services	PRODUCING QUADRANT				FINAL DEMAND QUADRANT
P1. Wages and salaries of workers P2. Incomes of land owners and capital owners P3. Indirect taxes paid to government P4. Non-competing imports P5: Competing imports	PAYMENTS QUADRANT				PAYMENTS-TO-FINAL DEMAND QUADRANT
TOTAL OUTPUTS

Imports are included in the payments quadrant and are competing or non-competing. Competing imports are those items imported for which some proportions are produced locally but are not sufficient to meet local demand. Non-competing imports are those which are not produced locally and are imported as inputs to produce outputs required by domestic industries. The final demand quadrant comprises of expenditures by households, government recurrent expenditures, government and private sector investment expenditures, changes in stocks of businesses, and exports. The payment-to-final demand quadrant deals with expenditures of the payment quadrant to final demand quadrant, exemplified by imported items consumed by households.

2.5. Discussion of the concept of input–output impact

Input-output impact analysis deals with changes in output, income and employment arising from changes in the final demand for goods and services (Anaman, 1994). Assume that the transactions (trans) among industries in an economy are a system of equations denoted below. $$\begin{array}{c}\mathbf{t}\mathbf{\text{ran}}{\mathbf{s}}_{\mathbf{11}}+\mathbf{\text{tran}}{\mathbf{s}}_{\mathbf{12}}+\mathbf{\text{tran}}{\mathbf{s}}_{\mathbf{13}}+\mathbf{\text{tran}}{\mathbf{s}}_{\mathbf{14}}+\mathbf{\text{tran}}{\mathbf{s}}_{\mathbf{15}}\dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \mathbf{\text{tran}}{\mathbf{s}}_{\mathbf{1}\mathbf{n}}+{\mathbf{Y}}_{\mathbf{1}}={\mathbf{X}}_{\mathbf{1}}\\ \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \\ \mathbf{t}\mathbf{\text{ran}}{\mathbf{s}}_{\mathbf{n}\mathbf{1}}+\mathbf{\text{tran}}{\mathbf{s}}_{\mathbf{n}\mathbf{2}}+\mathbf{\text{tran}}{\mathbf{s}}_{\mathbf{n}\mathbf{3}}+\mathbf{\text{tran}}{\mathbf{s}}_{\mathbf{n}\mathbf{4}}+\mathbf{\text{tran}}{\mathbf{s}}_{\mathbf{n}\mathbf{5}}\dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \mathbf{\text{tran}}{\mathbf{s}}_{\mathbf{\text{nn}}}+{\mathbf{Y}}_{\mathbf{n}}={\mathbf{X}}_{\mathbf{n}}\end{array}$$ where trans_ij is the output from industry i bought by industry j; Yi is the total final demand for the output from industry i; X_i is the total output from industry i.

A matrix of direct coefficients can be generated if the transactions, trans_ij are divided by the respective output levels of j (X_j.). Calling the direct coefficients, d_ij (trans_ij/X_j), a matrix of direct coefficients is indicated below. $$\begin{array}{c}{\mathbf{d}}_{\mathbf{11}}{\mathbf{X}}_{\mathbf{1}}+{\mathbf{d}}_{\mathbf{12}}{\mathbf{X}}_{\mathbf{1}}+{\mathbf{d}}_{\mathbf{13}}{\mathbf{X}}_{\mathbf{1}}+{\mathbf{d}}_{\mathbf{14}}{\mathbf{X}}_{\mathbf{1}}+{\mathbf{d}}_{\mathbf{15}}{\mathbf{X}}_{\mathbf{1}}\dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots {\mathbf{d}}_{\mathbf{1}\mathbf{n}}{\mathbf{X}}_{\mathbf{1}}+{\mathbf{Y}}_{\mathbf{1}}={\mathbf{X}}_{\mathbf{1}}\\ \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots .\\ {\mathbf{d}}_{\mathbf{n}\mathbf{1}}{\mathbf{X}}_{\mathbf{n}}+{\mathbf{d}}_{\mathbf{n}\mathbf{2}}{\mathbf{X}}_{\mathbf{n}}+{\mathbf{d}}_{\mathbf{n}\mathbf{3}}{\mathbf{X}}_{\mathbf{n}}+{\mathbf{d}}_{\mathbf{n}\mathbf{4}}{\mathbf{X}}_{\mathbf{n}}+{\mathbf{d}}_{\mathbf{n}\mathbf{5}}{\mathbf{X}}_{\mathbf{n}}\dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots \dots {\mathbf{d}}_{\mathbf{\text{nn}}}{\mathbf{X}}_{\mathbf{n}}+{\mathbf{Y}}_{\mathbf{n}}={\mathbf{X}}_{\mathbf{n}}\end{array}$$

Expressing the system of n equations in matrix format yields the identical Equations (1)(1) $$\mathbf{\text{DX }}+\mathbf{Y}=\mathbf{X}$$(1) and (2)(2) $$\mathbf{X}={\left[\mathbf{I}-\mathbf{D}\right]}^{-\mathbf{1}}\mathbf{Y}$$(2) below. (1) $$\mathbf{\text{DX }}+\mathbf{Y}=\mathbf{X}$$(1) (2) $$\mathbf{X}={\left[\mathbf{I}-\mathbf{D}\right]}^{-\mathbf{1}}\mathbf{Y}$$(2) where X represents n × 1 column vector of total output for each industry in the intermediate quadrant of the economy; D is n × n matrix of direct coefficients, which are inputs required from industry i for the production of one unit of output by industry j; I is a n × n identity matrix; Y is a n × 1 column of final demand.

The [I-D]⁻¹ matrix, is called the Leontief inverse matrix, named after the Russian-American concept originator, 1973 Nobel Prize Winner in Economic Sciences, Wassily Leontief (Leontief, 1936). This matrix is the total requirements table and is used for estimating the impacts of one unit change in final demand. These impacts are also called backward linkage effects. The Leontief inverse matrix is estimated by deriving the impact of each round of expenditures incurred by the impacting industry.

2.6. Input–Output (IO) multipliers

For input–output analysis, the degree of interdependence among industries in the larger economy can be measured using the concept of multipliers. Multipliers measure the output changes in one industry in response to changes in the final demand spending originating from other industries. The concept of multipliers is particularly useful for deriving the overall changes in final demand arising from household consumption, government recurrent spending, investments, by both government and private firms, and exports. An IO multiplier for industry j measures the change in an economic indicator, for example, output, value added to economy output, income or employment) for one unit change in final demand. Type 1 multipliers measure the direct and indirect changes in the economic indicator (say total output) in the intermediate quadrant as a result of one unit change in final demand expenditure from an impacting industry. There is double counting involved in output multipliers; hence the value-added multipliers are preferred to total output multipliers (Miller & Blair, 2009). Type 2 multipliers measure the impacts due to recognizing the household sector as an industry in the intermediate quadrant of the economy. This would involve moving the household sector in the payments quadrant and the household expenditure from the final demand quadrant, both into the intermediate quadrant.

2.7. Regionalization of national IO tables

As regional economics and regional science have developed in stature, the direction of IO research work has changed from the national entity to the smaller local entities. A key drawback of analyses involving national IO tables is that they neglect important regional features through the treatment of the entire national economy as one spatial entity (Möbius & Althammer, 2020). Regional characteristics and features are important and need to be recognized as such. It is useful to represent such regional characteristics through the construction of regional IO tables (Muñiz et al., 2011). An essential aspect of regionalization of the IO approach is the adaptation of the IO approach to smaller territorial units, since the analysis is suited to showing local impacts. In many countries, only national IO tables are available. However, using national coefficients in regional studies can provide misleading impacts. The advantage of regional I-O models is their ability to measure policy impacts on the regional economy (Isard et al., 2017).

2.8. Application of location quotients to regionalize national input-output tables

Due to the high costs of constructing regional IO tables from surveys, regional IO tables have been developed from national IO tables based on several methods. These methods include the use of location quotients, constrained balancing methods based on the use of limited data sets subject to linear constraints, and cross entropy methods (Flegg et al., 2021). Location quotients are the commonly-used non-survey methods due to their limited data requirements. An assumption for the use of location quotients is that there is a similar technology in the region and country as a whole. The simple location quotient for industry i (SLQ_i) is defined below in Equation (3). (3) $${\mathit{\text{SLQ}}}_{\mathit{i}}=\frac{\left(\frac{{{\mathit{R}}_{\mathit{i}}}^{\mathit{T}}}{{\mathit{R}}^{\mathit{T}}}\right)}{\left(\frac{{{\mathit{N}}_{\mathit{i}}}^{\mathit{T}}}{{\mathit{N}}^{\mathit{T}}}\right)}\mathrm{ }$$(3)

Where R_i^T is the total output of industry i in the region used for the study, R^T is the total output of all industries in the entire region; N_i^T is the total output of industry i of the whole country, and N^T is the total output of all industries in the whole country. The SLQ indicates the relative specialization of a region vis a vis the nation for a particular industry. If the SLQ is greater than one, then it is suggested that the region is adequately specialised in the production of the particular industry commodity to be able to meet its own local demand for that industry. The national input direct coefficients, indicated in Equation (1)(1) $$\mathbf{\text{DX }}+\mathbf{Y}=\mathbf{X}$$(1) as D matrix, are labelled D_N. The regional industry outputs (R_i^T_/R^T) can be used to develop a matrix of regional input direct coefficients (D_R) by adjusting D_N by SLQ as indicated in Equations (4) and (5) below. (4) $${\mathbf{d}}_{\mathbf{\text{ij}}}\mathbf{R}\mathbf{\text{EGION}}={\mathbf{d}}_{\mathbf{\text{ij}}}\mathbf{N}\mathbf{\text{ATION}}\times {\mathbf{\text{SLQ}}}_{\mathbf{i}}\mathrm{ }\mathbf{\text{if}}\mathrm{ }{\mathbf{\text{SLQ}}}_{\mathbf{i}}\mathrm{ }\mathbf{\text{is}}\mathrm{ }\mathbf{\text{less}}\mathrm{ }\mathbf{\text{than}}\mathrm{ }1$$(4) (5) $${\mathbf{d}}_{\mathbf{\text{ij}}}\mathbf{R}\mathbf{\text{EGION}}={\mathbf{d}}_{\mathbf{\text{ij}}}\mathbf{N}\mathbf{\text{ATION}}\times 1\mathrm{ }\mathbf{\text{if}}\mathrm{ }{\mathbf{\text{SLQ}}}_{\mathbf{i}}\mathrm{ }\mathbf{\text{is}}\mathrm{ }1\mathrm{ }\mathbf{\text{or}}\mathrm{ }\mathbf{\text{greater}}\mathrm{ }\mathbf{\text{than}}\mathrm{ }1$$(5) where d_ijREGION is the input required from industry i by industry j to produce one unit of output for final demand in the region of study. A similar definition holds for d_ijNATION.

An example of the use of SLQ for Ghana was undertaken by Anaman (1985) using the Ashanti Region to regionalize the 1960 IO table. The SLQ does not allow for the situation where a region imports and exports a commodity to the other regions of the country at the same time; the direct input coefficients of the region and the nation are assumed to be the same when the SLQ is greater than 1. The situation where simultaneous imports and exports occur is called cross hauling and this would lead to the underestimation of imports for a region. This criticism led to the concept of cross-industry location quotient (CILQ) (Stevens et al., 1989), defined as shown in Equation (6). (6) $${\mathbf{\text{CILQ}}}_{\mathbf{\text{ij}}}=\frac{{\mathbf{\text{SLQ}}}_{\mathbf{i}}\mathrm{ }}{{\mathbf{\text{SLQ}}}_{\mathbf{j}}\mathrm{ }}=\frac{\left[\frac{\left(\frac{{{\mathit{R}}_{\mathit{i}}}^{\mathit{T}}}{{\mathit{R}}^{\mathit{T}}}\right)}{\left(\frac{{{\mathit{N}}_{\mathit{i}}}^{\mathit{T}}}{{\mathit{N}}^{\mathit{T}}}\right)}\right]}{\left[\frac{\left(\frac{{{\mathit{R}}_{\mathit{J}}}^{\mathit{T}}}{{\mathit{R}}^{\mathit{T}}}\right)}{\left(\frac{{{\mathit{N}}_{\mathit{j}}}^{\mathit{T}}}{{\mathit{N}}^{\mathit{T}}}\right)}\right]}\mathrm{ }$$(6)

A limitation of the CILQ is that it does not incorporate the relative size of the region, defined as R^T_/N^T; this term is cancelled out in Equation (6). The limitations of both SLQ and CILQ have led to the development of enhanced location quotients for regionalization of national IO tables which capture five desirable features. These five features are (1) the relative size of the supplying industry, (2) the relative size of the buying industry, (3) the relative size of the region with respect to the nation as a whole, (4) more precise specification of the relative size of the region to capture the size of imports of relatively small regions, and (5) allowance for the case of regional direct input coefficients being bigger than national direct input coefficients due to highly-intense specialization of the production of a commodity in a region (Flegg et al., 2021). One location quotient developed that sufficiently meets the first three desirable requirements is the Round Location Quotient (RLQ) (Round, 1978). The RLQ is defined in Equation (7). (7) $${\mathbf{\text{RLQ}}}_{\mathbf{\text{ij}}}=\frac{{\mathbf{\text{SLQ}}}_{\mathbf{i}}}{{\mathbf{\text{log}}}_2\left(1+{\mathbf{\text{SLQ}}}_{\mathbf{j}}\right)}$$(7)

Two enhanced location quotients which have the first four ideal attributes are the Flegg Location Quotient (FLQ) and the Augmented Flegg Location Quotient (AFLQ) (Flegg et al., 2021). The FLQ is defined as two equations in Equations (8) and (9). (8) $$\mathit{\text{FL}}{Q}_{\mathit{\text{ij}}}=C\mathit{\text{IL}}{Q}_{\mathit{\text{ij}}}\times \mathit{\text{ALPHA}}\mathit{\text{if}}i\mathit{\text{is}}\mathit{\text{not}}\mathit{\text{equal}}\mathit{\text{to}}j$$(8) (9) $$\mathit{\text{FL}}{Q}_{\mathit{\text{ij}}}=SL{Q}_{\mathit{\text{ij}}}\times \mathit{\text{ALPHA}}\mathit{\text{if}}i\mathit{\text{is}}\mathit{\text{equal}}\mathit{\text{to}}j$$(9) $$\mathit{\text{where ALPHA}}={\left(\mathit{\text{lo}}{g}_2\left(1+\raisebox{1ex}{$R^T$}\!\left/ \!\raisebox{-1ex}{$N^T$}\right.\right)\right)}^{\mathit{\text{DELTA}}}$$

The Delta variable varies from zero to 1 and represents an allowance for extra imports into the region. Flegg et al. (2021), in a study based on South Korean data, suggested the values of Delta from 0.2 to 0.4, similar to those used by Jahn et al. (2020). The FLQ is implemented by modifying the national direct coefficients to regional direct coefficients as shown in Equations (10) and (11). (10) $$d_{\mathit{\text{ij}}}\mathit{\text{REGION}}=d_{\mathit{\text{ij}}}\mathit{\text{NATION}}\times \mathit{\text{FL}}{Q}_{\mathit{\text{ij}}}\mathit{\text{if}}\mathit{\text{FL}}{Q}_{\mathit{\text{ij}}}\mathit{\text{is}}\mathit{\text{less}}\mathit{\text{than}}1$$(10) (11) $$d_{\mathit{\text{ij}}}\mathit{\text{REGION}}=d_{\mathit{\text{ij}}}\mathit{\text{NATION}}\times 1\mathit{\text{if}}\mathit{\text{FL}}{Q}_{\mathit{\text{ij}}}\mathit{\text{is}}1\mathit{\text{or}}\mathit{\text{greater}}\mathit{\text{than}}1$$(11)

The AFLQ is directly derived from FLQ and was developed to allow for very high regional specialization in the production of a commodity in order to incorporate the fifth ideal attribute. The AFLQ is defined as follows in Equations (12) and (13). (12) $$\mathit{\text{AFL}}{Q}_{\mathit{\text{ij}}}=\mathit{\text{FL}}{Q}_{\mathit{\text{ij}}}\times \mathit{\text{lo}}{g}_2\left(1+\mathit{\text{SL}}{Q}_{\mathit{\text{ij}}}\right)\mathit{\text{where}}\text{\hspace{0.33em}S}L{Q}_j\mathit{\text{is}}\mathit{\text{greater}}\mathit{\text{than}}1$$(12) (13) $$\mathit{\text{AFL}}{Q}_{\mathit{\text{ij}}}=\mathit{\text{FL}}{Q}_{\mathit{\text{ij}}}\mathit{\text{if}}\mathit{\text{SL}}{Q}_j\mathit{\text{is}}1\mathit{\text{or}}\mathit{\text{less}}\mathit{\text{than}}1$$(13)

The AFLQ is implemented in the modification of national direct coefficients to regional direct coefficients as follows as indicated in Equations (14) and (15). (14) $$d_{\mathit{\text{ij}}}\mathit{\text{REGION}}=d_{\mathit{\text{ij}}}\mathit{\text{NATION}}\times \mathit{\text{AFL}}{O}_{\mathit{\text{ij}}}\mathit{\text{if}}\mathit{\text{SL}}{Q}_j\mathit{\text{is}}\mathit{\text{greater}}\mathit{\text{than}}1$$(14) (15) $$d_{\mathit{\text{ij}}}\mathit{\text{REGION}}=d_{\mathit{\text{ij}}}\mathit{\text{NATION}}\times \mathit{\text{FL}}{Q}_{\mathit{\text{ij}}}\mathit{\text{if}}\mathit{\text{SL}}{Q}_j\mathit{\text{is}}1\mathit{\text{or}}\mathit{\text{less}}\mathit{\text{than}}1$$(15)

2.9. Summary of location Quotient-based IO analysis

Regionalization of national IO tables for specific areas within nation states has been extensively undertaken and widely reported in the empirical literature, especially for developed countries. More recent comprehensive treatments of the subject of theoretical and empirical applications of regionalization of IO tables are covered by Flegg et al. (2021), Pereira-López et al. (2020), Lamonica et al. (2018), Jahn et al. (2020), Lampiris et al. (2019), Kowalewksi (2015) and Zhao and Choi (2015). The literature has emphasized validation of location quotients using survey data.

3. Methodology utilized for this study

The 2018 Ghana social accounting matrix (SAM) developed by the Ghana Statistical Service and the International Food Policy Research Institute was transformed into a relevant input-output (IO) table. The 76-industry SAM was converted into a 76-industry IO table by the senior author. With the availability of the data from the 2021 Population Census, the 76-industry national IO table was transformed into a 17-industry IO table to fit the data availability from the Census; this table is illustrated in Table 1. The derivation of SLQ was based on the assumption that the technology of production in the five northern regions of Northern Ghana was similar to those in the rest of Ghana. Employment figures were used as proxy for output levels as commonly observed in the literature given the easier availability of employment data compared to output data. Various simulation experiments of the regionalization of the Ghana IO table were undertaken for Northern Ghana based on four types of location quotients: SLQ, CILQ, Round, and Augmented Flegg (AFLQ). These location quotients are discussed in more details by Flegg et al. (2021).

The gross output impacts identified and estimated in this study comprise of two types: (1) backward linkage multipliers and (2) forward linkage multipliers (Miller & Blair, 2009; Isard et al., 2017). Backward linkage multipliers reflect the purchases by one particular industry from all other industries in the domestic economy to produce output in that industry triggered by one unit change in final demand. They represent the demand side of IO analysis. The forward linkage multipliers measure impacts in terms of the supply of inputs from one industry to other industries in the domestic economy arising from one unit change in final demand. The forward linkage multipliers connote the supply side of IO analysis.

4. Results and discussion

4.1. Results

The computed shares of workers in 17 industries as proportions of all workers for Northern Ghana and all the 16 regions of Ghana are reported in Table 2. These shares were used to derive the SLQ for all the 17 industries of the economy. The results of the SLQ indicated that agriculture (SLQ above 1.0) was highly specialized in Northern Ghana and for which the area could meet its demand from local supply. One industry – education – had SLQ slightly above 1.0 indicating that Northern Ghana had the share of educational industry workers corresponding to the national average. For the other 15 industries, the SLQ of Northern Ghana was below 1.0.

The very low SLQs for transport and storage, and information and communication industries reflected the poor-quality infrastructure in Northern Ghana as compared to Ghana as a whole. The SLQs for the other two infrastructure-based industries - electricity supply and water supply - were also less than 1.0. However, their SLQs were higher than those for transport and storage and information and communication industries. This result reflected the bigger societal investments made in electricity and water supply in Northern Ghana as compared to those for transport and storage (roads) and information and communication (computer and internet services).

Table 2. Computed shares of workers in various industries as proportions of all workers for the whole Ghana and for the five regions of Northern Ghana and the derived SLQ for Northern Ghana for 17 industries.

Industry	Share of industry in total employment in Ghana	Share of industry in total employment in Northern Ghana	Northern Ghana SLQ
Agriculture and forestry	0.3300	0.5836	1.768
Mining and quarrying	0.0112	0.0067	0.604
Manufacturing	0.0669	0.0469	0.701
Electricity and gas	0.0054	0.0034	0.636
Water supply	0.0014	0.0012	0.837
Construction	0.0522	0.0351	0.673
Trading and repair	0.1867	0.0985	0.528
Transport and storage	0.0523	0.0187	0.359
Hotels and restaurants	0.0526	0.0283	0.538
Information and communication	0.0049	0.0019	0.388
Finance and insurance	0.0146	0.0062	0.423
Real estates	0.0011	0.0008	0.736
Business services	0.0223	0.0149	0.668
Public administration	0.0166	0.0120	0.722
Education	0.0593	0.0599	1.011
Health	0.0261	0.0254	0.974
All other services	0.0964	0.0563	0.583

Source: The ratios are computed from the 2021 national population census data reported by Ghana Statistical Service (2021).

Table 3(a) provides the computed backward linkage gross output multipliers arising from one unit change in final demand for all the 17 industries for both Northern Ghana and Ghana as a whole based on the use of SLQ to regionalize the national IO table. The results indicate that the most dominant industry at both the national and regional levels was electricity and gas, followed by construction and information and communication industries. For the backward linkage gross output multipliers, those for Northen Ghana were lower than those recorded for Ghana as a whole for all 17 industries Table 3(b) summarizes the results for the forward linkage multipliers. For both Ghana and Northern Ghana, the most important (supplying) industry was manufacturing, followed by the finance and insurance and transport and storage industries. The forward linkage multipliers were virtually the same for both Ghana and Northern Ghana with respect to public services (namely public administration, education and health) and also all other services. However, for the other 13 industries, the multipliers for Ghana were higher than those of Northern Ghana.

Table 3(a). Computed backward linkage gross output multipliers based on one Ghana cedi change in final demand for Ghana and Northern Ghana using SLQ for the regionalization of the 2018 Ghana national input-output table for Northern Ghana.

Industry	Total output multiplier for the whole Ghana based on gross output	Total output multiplier for Northern Ghana based on SLQ	Difference between Ghana and Northern Ghana
Agriculture and forestry	1.120	1.077	0.044
Mining and quarrying	1.378	1.160	0.218
Manufacturing	1.366	1.240	0.126
Electricity and gas	2.191	1.623	0.568
Water supply	1.648	1.354	0.295
Construction	1.774	1.462	0.312
Trading and repair	1.331	1.141	0.191
Transport and storage	1.236	1.133	0.103
Hotels and restaurants	1.530	1.360	0.170
Information and communication	1.746	1.308	0.438
Finance and insurance	1.582	1.220	0.362
Real estates	1.074	1.035	0.039
Business services	1.135	1.060	0.075
Public administration	1.292	1.137	0.155
Education	1.535	1.338	0.197
Health	1.742	1.443	0.300
All other services	1.756	1.356	0.400

Source: The multipliers are computed from data from the 2021 national population census data reported by Ghana Statistical Service (2021) and the 2018 Ghana SAM.

Table 4(a) is a summary of the derived backward linkage output multipliers for Northern Ghana and Ghana as a whole from one unit change in final demand when CILQ is used for regionalizing the national IO table. Similar to the results for SLQ, the national backward linkage output multipliers were bigger than the corresponding output multipliers for Northern Ghana for all 17 industries. However, the differences were very small for the service-based industries, except for the public services (public administration, education and health) and all other services combined. The corresponding forward linkage results are provided in Table 4(b). Similar to the results in the previous section, the computed multipliers were higher for Ghana than those of Northern Ghana for all 17 industries. However, for the majority of the industries, the differences in the magnitude of the multipliers were very small. In fact, for the public services, all other services, water supply and construction, the multipliers were virtually the same for both areas.

Table 3(b). Computed forward linkage gross output multipliers based on one Ghana cedi change in final demand for Ghana and Northern Ghana using SLQ for the regionalization of the 2018 Ghana national input–output table for Northern Ghana.

Industry	Total output multiplier for the whole Ghana based on gross output	Total output multiplier for Northern Ghana based on SLQ	Difference between Ghana and Northern Ghana
Agriculture and forestry	1.440	1.346	0.094
Mining and quarrying	1.492	1.249	0.243
Manufacturing	4.223	2.965	1.258
Electricity and gas	1.720	1.393	0.327
Water supply	1.064	1.048	0.015
Construction	1.089	1.054	0.035
Trading and repair	1.218	1.101	0.117
Transport and storage	1.580	1.181	0.400
Hotels and restaurants	1.147	1.067	0.080
Information and communication	1.338	1.110	0.228
Finance and insurance	2.434	1.462	0.972
Real estates	1.115	1.076	0.039
Business services	1.398	1.223	0.175
Public administration	1.013	1.007	0.006
Education	1.153	1.153	0.000
Health	1.007	1.007	0.000
All other services	1.005	1.003	0.002

Source: The multipliers are computed from data from the 2021 national population census data reported by Ghana Statistical Service (2021) and the 2018 Ghana SAM.

The summary of the results of the analysis of the backward linkage output multipliers is presented in Table 5(a) for both Ghana and Northern Ghana for all 17 industries when the RLQ is used for the regionalization of the national IO table. The electricity and gas industry had the highest multiplier for both Ghana and Northern Ghana. The other important industries in terms of multiplier impacts were construction, all other services combined, and information and communication. In terms of forward linkage multipliers, as shown in Table 5(b) the most impacting industries were manufacturing, finance and insurance and electricity and gas. Overall, unlike the earlier results, the RLQ-based analysis revealed electricity and gas as a very important impacting industry in terms of both backward and forward linkage effects. Furthermore, for forward linkage effects, the multipliers for the three public services and all other services combined had virtually the same magnitude for both Ghana and Northern Ghana. This particular result is also similar to those for other industries such as real estate services, business services, water supply, construction, and trade.

Table 4(a). Computed backward linkage gross output multipliers based on one Ghana cedi change in final demand for Ghana and Northern Ghana using CILQ for the regionalization of the 2018 Ghana national input-output table for Northern Ghana.

Industry	Total output multiplier for the whole Ghana based on gross output	Total output multiplier for Northern Ghana based on CILQ	Difference between Ghana and Northern Ghana
Agriculture and forestry	1.120	1.054	0.067
Mining and quarrying	1.378	1.281	0.096
Manufacturing	1.366	1.333	0.033
Electricity and gas	2.191	2.117	0.074
Water supply	1.648	1.491	0.157
Construction	1.774	1.722	0.052
Trading and repair	1.331	1.263	0.069
Transport and storage	1.236	1.230	0.006
Hotels and restaurants	1.530	1.500	0.030
Information and communication	1.746	1.734	0.012
Finance and insurance	1.582	1.566	0.015
Real estates	1.074	1.055	0.019
Business services	1.135	1.103	0.032
Public administration	1.292	1.207	0.085
Education	1.535	1.359	0.176
Health	1.742	1.493	0.249
All other services	1.756	1.636	0.120

Source: The multipliers are computed from data from the 2021 national population and housing census data reported by Ghana Statistical Service (2021) and the 2018 Ghana SAM.

Table 4(b). Computed forward linkage gross output multipliers based on one Ghana cedi change in final demand for Ghana and Northern Ghana using CILQ for the regionalization of the 2018 Ghana national input-output table for Northern Ghana.

Industry	Total output multiplier for the whole Ghana based on gross output	Total output multiplier for Northern Ghana based on CILQ	Difference between Ghana and Northern Ghana
Agriculture and forestry	1.440	1.412	0.028
Mining and quarrying	1.492	1.422	0.070
Manufacturing	4.223	3.873	0.351
Electricity and gas	1.720	1.654	0.066
Water supply	1.064	1.060	0.004
Construction	1.089	1.075	0.015
Trading and repair	1.218	1.165	0.053
Transport and storage	1.580	1.329	0.251
Hotels and restaurants	1.147	1.107	0.040
Information and communication	1.338	1.217	0.121
Finance and insurance	2.434	2.199	0.235
Real estates	1.115	1.101	0.014
Business services	1.398	1.357	0.041
Public administration	1.013	1.011	0.002
Education	1.153	1.153	0.000
Health	1.007	1.006	0.002
All other services	1.005	1.004	0.001

Source: The multipliers are computed from data from the 2021 national population and housing census data reported by Ghana Statistical Service (2021) and the 2018 Ghana SAM.

Table 5(a). Computed backward linkage gross output multipliers based on one Ghana cedi change in final demand for Ghana and Northern Ghana using RLQ for the regionalization of the 2018 Ghana national input–output table for Northern Ghana.

Industry	Total output multiplier for the whole Ghana based on gross output	Total output multiplier for Northern Ghana based on RLQ	Difference between Ghana and Northern Ghana
Agriculture and forestry	1.120	1.060	0.060
Mining and quarrying	1.378	1.251	0.126
Manufacturing	1.366	1.305	0.061
Electricity and gas	2.191	1.996	0.195
Water supply	1.648	1.451	0.197
Construction	1.774	1.660	0.114
Trading and repair	1.331	1.234	0.098
Transport and storage	1.236	1.221	0.015
Hotels and restaurants	1.530	1.485	0.045
Information and communication	1.746	1.646	0.100
Finance and insurance	1.582	1.464	0.117
Real estates	1.074	1.049	0.025
Business services	1.135	1.090	0.045
Public administration	1.292	1.185	0.107
Education	1.535	1.353	0.182
Health	1.742	1.478	0.264
All other services	1.756	1.650	0.105

Source: The multipliers are computed from data from the 2021 national population and housing census data reported by Ghana Statistical Service (2021) and the 2018 Ghana SAM.

Table 5(b). Computed forward linkage gross output multipliers based on one Ghana cedi change in final demand for Ghana and Northern Ghana using RLQ for the regionalization of the 2018 Ghana national input–output table for Northern Ghana.

Industry	Total output multiplier for the whole Ghana based on gross output	Total output multiplier for Northern Ghana based on RLQ	Difference between Ghana and Northern Ghana
Agriculture and forestry	1.440	1.402	0.039
Mining and quarrying	1.492	1.372	0.120
Manufacturing	4.223	3.729	0.494
Electricity and gas	1.720	1.574	0.146
Water supply	1.064	1.058	0.005
Construction	1.089	1.072	0.018
Trading and repair	1.218	1.151	0.067
Transport and storage	1.580	1.302	0.278
Hotels and restaurants	1.147	1.101	0.045
Information and communication	1.338	1.165	0.173
Finance and insurance	2.434	2.040	0.394
Real estates	1.115	1.099	0.016
Business services	1.398	1.340	0.058
Public administration	1.013	1.011	0.002
Education	1.153	1.153	0.000
Health	1.007	1.006	0.001
All other services	1.005	1.004	0.001

Source: The multipliers are computed from data from the 2021 national population and housing census data reported by Ghana Statistical Service (2021) and the 2018 Ghana SAM.

Table 6(a). Computed backward linkage gross output multipliers based on one Ghana cedi change in final demand for Ghana and Northern Ghana using AFLQ for the regionalization of the 2018 Ghana national input–output table for Northern Ghana with a Delta value of 0.2.

Industry	Total output multiplier for the whole Ghana based on gross output	Total output multiplier for Northern Ghana based on AFLQ	Difference between Ghana and Northern Ghana
Agriculture and forestry	1.120	1.068	0.052
Mining and quarrying	1.378	1.200	0.177
Manufacturing	1.366	1.261	0.105
Electricity and gas	2.191	1.587	0.604
Water supply	1.648	1.302	0.346
Construction	1.774	1.498	0.276
Trading and repair	1.331	1.202	0.129
Transport and storage	1.236	1.268	−0.032
Hotels and restaurants	1.530	1.600	−0.070
Information and communication	1.746	1.576	0.170
Finance and insurance	1.582	1.267	0.314
Real estates	1.074	1.035	0.040
Business services	1.135	1.061	0.074
Public administration	1.292	1.139	0.153
Education	1.535	1.242	0.293
Health	1.742	1.329	0.414
All other services	1.756	1.446	0.310

Source: The multipliers are computed from data from the 2021 national population and housing census data reported by Ghana Statistical Service (2021) and the 2018 Ghana SAM.

Table 6(b). Computed forward linkage gross output multipliers based on one Ghana cedi change in final demand for Ghana and Northern Ghana using AFLQ for the regionalization of the 2018 Ghana national input–output table for Northern Ghana with a Delta value of 0.2.

Industry	Total output multiplier for the whole Ghana based on gross output	Total output multiplier for Northern Ghana based on AFLQ	Difference between Ghana and Northern Ghana
Agriculture and forestry	1.440	1.712	−0.272
Mining and quarrying	1.492	1.259	0.233
Manufacturing	4.223	3.029	1.194
Electricity and gas	1.720	1.308	0.412
Water supply	1.064	1.053	0.011
Construction	1.089	1.063	0.026
Trading and repair	1.218	1.115	0.103
Transport and storage	1.580	1.212	0.368
Hotels and restaurants	1.147	1.074	0.073
Information and communication	1.338	1.132	0.206
Finance and insurance	2.434	1.593	0.842
Real estates	1.115	1.082	0.032
Business services	1.398	1.316	0.082
Public administration	1.013	1.015	−0.001
Education	1.153	1.113	0.040
Health	1.007	1.006	0.002
All other services	1.005	1.002	0.003

Source: The multipliers are computed from data from the 2021 national population and housing census data reported by Ghana Statistical Service (2021) and the 2018 Ghana SAM.

Tables 6(a) and (b) provide the summary of the results of the analysis related to backward linkage and forward linkage multipliers respectively, based on the use of AFLQ for the regionalization of the national IO table. The base analysis was based on the Delta value of 0.2. In terms of backward linkage multipliers, the electricity and gas industry had the highest value for Ghana as a whole but was second in dominance for Northern Ghana; the first position was taken by the hotels and restaurants industry for Northern Ghana. The latter industry and the transport and storage industry were also the only two industries for which the multiplier magnitudes of Northern Ghana had higher values as compared to Ghana. In terms of forward linkage effects, manufacturing was the most important impacting industry for both Ghana and Northern Ghana. For Northern Ghana, agriculture was the second most impacting; however, its multiplier value was bigger than the corresponding magnitude for Ghana as a whole. The bigger multiplier for Northern Ghana agriculture is due to the nature of the AFLQ as it allows for the existence of very high regional specialization for an industry such as agriculture in Northern Ghana.

4.2. Discussion

A challenge of our study is the choice of the best available method for producing regional IO tables for Northern Ghana based on the identified four methods related to the concept of location quotients. Ideally, industry multipliers produced by location quotients should be analytically compared to multipliers which are true, for example, those derived from large scale surveys covering all of Northern Ghana. Given that the true output multipliers for Northern Ghana are unknown, then we have to rely on comparative studies undertaken in the area which have evaluated impacts of investments and expenditures in specific industries.

From our perspective, all four location quotients gave results which had considerable unanimity in terms of the most impacting industries and also for industries whose output multipliers were very similar in values for both Ghana and Northern Ghana. The most impacting industry in terms of backward linkage effects was electricity and gas. For forward linkage effects, the corresponding dominant industry was manufacturing. Further, in terms of forward linkage effects, the three public services (public administration, education and health), plus all other services combined, the construction and real estate industries, had virtually the same multiplier magnitude levels for both Ghana as a whole and Northern Ghana as a region. This particular result for the six industries was obtained for all four location quotients.

Overall, we consider the use of RLQ for regionalizing the national IO table for Northern Ghana as the best of the four methods due to four main reasons. First, the RLQ method is theoretically sound as it incorporates both the relative size of Northern Ghana and the possibility of cross-hauling (exports and imports of commodities from Northern Ghana to the rest of Ghana at the same time). Second, with the use of the closely competitive location quotient, the AFLQ, this particular method produced results which were largely dependent on the value of Delta. With Delta values below 0.05 and over 0.35, we established increasing differences in the magnitudes of the multipliers, for both backward and forward linkage effects, for both Ghana and Northern Ghana. The results using the AFLQ method were generally not stable. Third, the preferred RLQ method, in terms of the backward linkage impact, produced relatively high multipliers for the construction industry in Northern Ghana as compared to the other methods related to SLQ and AFLQ. This result supported the common view of labour-intensive construction related projects providing higher levels of impacts for Northern Ghana. This view is supported by State-provided labour-intensive construction projects often undertaken during the dry season in Northern Ghana.

Fourth, for the empirical analysis results reported earlier, while the RLQ method produced several similar results to those generated by the other three methods, the RLQ method was the only one that was able to establish the electricity and gas industry as very important in terms of output impacts, at the regional level, and also for both backward and forward linkage impacts. This result highlighting electricity and gas a very important industry for Northern Ghana, in terms of its output impacting nature, is supported by other studies of Northern Ghana, such as Salifu and Anaman (2019) and Addo (2022); these two studies indicated direct link to improved economic welfare from the use of electricity.

Specifically, Addo, in her study of 2,920 households in rural Northern Ghana, based on data from the Ghana Living Standards Survey (GLSS) Round Seven, collected by the Ghana Statistical Service in 2016 and 2017, established that the economic welfare of a householder was positively influenced by the use of electricity. Addo (2022) also showed that a household with multiple streams of income had increased likelihood of higher economic welfare. These streams of income were more easily acquired when the householder used electricity in his/her home.

Salifu and Anaman (2019), using GLSS6 data (2012/2013) covering 1,194 rural households in the Northern, Savannah and North East regions, established that the number of income activities was positively influenced by electricity use. The high value of output multiplier from electricity generation is also revealed through qualitative-data-based information gathered in March 2022 from opinion leaders and businessmen from several villages in the Greater Tamale Area hit by severe power cuts for several weeks (Ghana Web News, 2022b).

5. Conclusions and recommendations

5.1. Conclusions

Using various types of location quotients, we constructed regional IO tables for Northern Ghana based on the latest 2018 national IO table for Ghana and the 2021 national population and housing census data. We suggest that the RLQ is the best location quotient in regionalizing the Ghanaian national IO for Northern Ghana table based on the supporting evidence from other studies in Northern Ghana. The results of our analysis are corroborated by evidence from secondary sources linked to the output impacts of electricity generation and gas, manufacturing, and off-farm wage-based employment in non-agricultural industries, such as construction. We have demonstrated that a Northern Ghana regional IO table can be developed from national data based on location quotients. The enhancement of the regional IO table for Northern Ghana is a continuous exercise that is undertaken with increased availability of data, for example, through national living standards surveys undertaken by the Ghana Statistical Service.

The results of this study indicate another important factor for the observed increasing poverty rates of groups and tribes in Northern Ghana. This factor is the limited availability of infrastructural investment, especially in the area of electricity and gas supply. Like thirsty living things, whose appetite of available water is very high, the high output multipliers of the electricity and gas industry in Northern Ghana would reflect the relatively low level of infrastructural development (very high thirst for water) in Northern Ghana.

A theoretical contribution of our research study is that regional IO tables produce results which are different from those generated by national IO tables. Hence there is a need to encourage the development of regional IO tables through the use of various methods rather than one single method. Further, the results from both regional and national IO tables need to be compared and the method producing the most credible results chosen for regionalization of national IO tables. The societal benefits of our research study include the improvement of policy analysis to reduce poverty arising from the use of regionalized national IO tables with focus on the measurement of welfare indicators focused on particular economically distressed regions and areas.

5.2. Recommendations

We conclude that the Northern Ghana regional IO table offers a useful tool to analyse the impacts of development projects in the area, for example, by the Northern Development Authority, and allied State organizations. Given the high impacts of electricity and construction industries, increased provision of infrastructure related to these items in Northern Ghana could reduce the poverty levels in the area. This infrastructural provision could also boost the output of the manufacturing industry; the latter industry was shown in our study to be most impacting in terms of forward linkage (supply) effects. Government and Community organizations working in Northern Ghana could put greater emphasis on the provision of infrastructural services as this emphasis would boost development in the areas through large output multiplier impacts.

Acknowledgments

This article was produced from the Ghana National Input-Output Analysis Project located at the Department of Agricultural Economics and Agribusiness, University of Ghana, Legon, Accra. The project has provided data for the development of four Doctor of Philosophy and one Master of Philosophy theses in the Department since August 2020. The research assistance support of Messrs. Kinsley Delanyo Adjei, Samuel Ampomah and Gideon Ayettey is gratefully acknowledged.

Disclosure statement

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

Additional information

Notes on contributors

Kwabena Asomanin Anaman

Kwabena Asomanin Anaman is a Professor of Applied Economics at the University of Ghana, Legon, Accra, Ghana. He has teaching, research, community service and extension interests in resource and environmental economics, political economy, international trade, economic growth (firm, industry and macroeconomy), business and managerial economics, applied econometrics and statistics, and operations research and economic optimization.

Abdul-Fatawu Shaibu

Abdul-Fatawu Shaibu is a Lecturer in Agribusiness at the University for Development Studies, Tamale, Ghana. He has teaching and research interests in agribusiness, agricultural economics and finance.

Notes

1 An important constitutional reform required in Ghana is the direct election of Mayors and DCEs. The inability of Ghanaian citizens to elect their Mayors and DCEs has acted as a barrier to the emergence of proven and tested local leaders who could take the mantle of national leadership as has happened in many countries such as Australia, Britain, France, India, Philippines and the United States where people directly elect their Mayors and DCEs. Many people aspire to be President in Ghana and some have contested national elections with very little impact. This very limited impact is due to the voters’ lack of knowledge of the potential effectiveness of leaders outside the two main parties. Such knowledge could be obtained if potential national leaders had a chance to work first as elected Mayors at the district level, similar to the situation of the outgoing Indonesian President Mr. Joko Widodo. Indonesia made a transition to democratic rule in 1998, after a 32-year military rule starting from the coup in 1966, the same year of Ghana’s first coup. Mr. Widodo, currently 62 years old, was elected as Mayor of his home district of Solo at the age of 43 and reelected with 90% of the votes, before being elected as Governor of Jakarta, the capital of Indonesia. He has since governed Indonesia for almost 10 years as the twice-elected President overseeing the transition of the country, with a similar chequered history like Ghana, to being the seventh largest economy in the world based on purchasing power parity through an expanded programme of infrastructural development and more inclusive economic growth.