Interdependence Between the Tourist Regions of Sergipe, Brazil

. We constructed an interregional input-output system for the tourist regions of Sergipe and identified the contribution of Tourism Characteristic Activities (TCAs) to the state’s economy. It is the first system built for tourist regions in Brazil that disaggregates tourism activities by sector and region, representing a novel approach in Brazilian literature. By measuring the weight of tourist activities, we avoid overestimating tourism in the regional economy. Researchers can use this method for countries and regions that do not have a Tourism Satellite Account. The main results estimate that TCAs in Sergipe accounted for 1.53% of the state’s gross value added (GVA) in 2015, 3.7 times lower when we do not properly disaggregate the tourism activities. The Polo Costa dos Coqueirais stands out among the tourist regions, particularly regarding the distribution of TCAs’ GVA within the state. Tourist road transportation is considered a key sector in all tourist regions.


Introduction
Recent years have begun to show a scenario of world recovery in tourist activity, which had been severely affected by the COVID-19 pandemic.In 2019, according to the World Travel & Tourism Council's annual Economic Impact Report (EIR) data, tourism accounted for 10.3% of the world's Gross Domestic Product (GDP).In 2021 and 2022, however, this share dropped to 6.1% and 7.6%, respectively, which is still below the prepandemic levels.Brazil, an important tourist destination in South America, was also severely impacted by the pandemic.Ribeiro et al. (2021) estimated a 31% drop in the GDP of Brazilian tourist activities in 2020.
In this present scenario of recovery in the sector, the existence of based on concrete planning instruments is fundamental.Tourism is an important development alternative for poorer countries or regions.In Brazil, tourism has already been used explicitly as a regional development policy through PRODETUR Nacional, specifically in the Northeast region -PRODETUR NE I and II.Studies by Haddad et al. (2013) and Ribeiro et al. (2017Ribeiro et al. ( , 2022) ) show that tourism reduces regional inequalities in the country.
The scarcity of resources in poorer states, often located in peripheral regions, contributes to the fact that tourism policy is not a priority in state public management, as is the case of Sergipe, located in the Brazilian Northeast.Sergipe is the smallest state in the country in territorial terms and accounts for 4% of the Northeast GDP and 0.6% of the national GDP, respectively.Although the state has tourism potential in several segments (sun and beach, adventure, and historical-cultural), they are not fully exploited.
Constructing tools that can aid tourism planning is fundamental for tourism development.Thus, this paper aims to build an inter-regional input-output (IO) system for Sergipe's tourist regions and identify the contribution of Tourism Characteristic Activities (TCAs) to the state economy.TCAs brings together tourism-related sectors, such as transportation, accommodation and food services, travel agencies and entertainment and leisure services.
Although many studies have used the IO model to estimate the intra and interregional economic effects of activities associated with tourism (Polo, Valle 2008, Lee et al. 2020, Lee, Hlee 2021, Kumara et al. 2021), the evaluation of the productive and regional interdependence of TCAs has been little explored in the literature.For the Brazilian case, some studies measure the economic contribution of tourism in specific regions, but they do not consider the inter-regional interdependence of TCAs.
Casimiro Filho, Guilhoto (2003) built an IO model for the tourist economy in Brazil in 1999 and measured intersectoral linkages and the ability to induce investments in economic growth.Takasago, Mollo (2011) examined the potential for stimulating production growth, income generation, and employment in the Federal District.To do so, they use the IO matrix to calculate the linkage effects and the potential generators of production, employment, and tourism income, enabling a more accurate sectorial view.Souza (2015) use an interregional IO matrix for Brazil to analyze the economic contribution of tourism in the Northeast region.Based on this, the authors seek to measure the influence of tourism on job and income generation, as well as its impact on reducing labor income inequality.Gonçalves et al. (2020) propose a method to measure the size of TCAs and their evolution in the Brazilian economy and its states between 2010 and 2015.The method consists of measuring TCAs on the supply side, using the same techniques employed to measure activities within the scope of the System of National and Regional Accounts of Brazil.They estimate a structure of weights applied to the value-added of economic activity groups.
The input-output method has been employed in various contexts in Europe.Mikulić et al. (2023), for instance, estimate the regional economic impact of tourism in Croatia.According to these authors, the reduction in economic activity due to the pandemic had a significant negative effect on GDP in all regions of Croatia.The direct effects are more pronounced in the Adriatic region, while the indirect effects are higher in the continental region.Rokicki et al. (2021) analyze different approaches to the construction of multiregional IO tables for Austria and Pérez et al. (2009) use an inter-regional IO model to estimate the economic impact of European Union structural funds on the regions of Spain from 1995 to 1999.Ivanova et al. (2019) and Araújo Junior et al. (2020) also adopt the IO model to assess specific issues in European countries.
By employing an inter-regional IO system, the study avoids potential overestimation of the economic contribution of tourism by disaggregating TCAs.The absence of a Tourism Satellite Account (TSA) implies a failure to differentiate between expenditures by residents and non-residents within tourism-related sectors.For instance, expenditures such as meals consumed by residents while away from their usual residence are amalgamated within the accounting framework of the food sector.Consequently, this amalgamation leads to an overestimation effect on the perceived impact of tourism within the given locale.
This methodological insight not only enhances the precision of the findings for Sergipe but also provides a valuable approach for regions and countries globally facing similar challenges in accurately assessing the economic impact of tourism.The study's focus on trade flow dynamics, value-added concentration, and employment multipliers within  Moreover, identifying key sectors provides tangible insights for policymakers and researchers promoting sustainable tourism development.While the research refrains from making explicit cross-regional or cross-national comparisons, its emphasis on precise regional data offers a rich foundation for future comparative studies.Additionally, the research highlights the need for tailored policies in Sergipe, leveraging regional production chains.This focus on practical applications adds depth to the broader international discourse on effective tourism planning and development, making it pertinent to a global audience of researchers, policymakers, and practitioners navigating the complexities of regional economic recovery and growth.
Despite efforts in the national literature to assess the productive interdependence of TCAs in Brazil, no study simultaneously deals with the regional and sectoral specification of tourist activities.Our main contribution, therefore, is: i) to regionally disaggregate the weight of trade flows from tourist activities and ii) to provide a tourism planning tool for Sergipe to encourage tourism development.In other words, this paper offers an unprecedented database for Brazil and Sergipe by sectorally and regionally disaggregating tourist activities.The method for estimating tourism can be replicated in countries and regions without a TSA, such as Brazil.The disaggregation of the TCAs avoids overestimating the effects of tourism on the state economy.By building this system, regional heterogeneity is addressed.Thus, issues such as productive linkages, employment and income multipliers and spillovers can be discussed by tourist regions.
The remainder of this paper is organized as follows.Next section (Section 2) describes the tourist regions of Sergipe, based on socioeconomic data.Section 3 presents the stepby-step construction of the inter-regional IO system by tourist region and describes the databases used.Section 4 contains the results and discussion, followed by the main conclusions and policy recommendations.

Tourist Regions of Sergipe
Sergipe possesses considerable untapped potential in tourism.The state is subdivided into five distinct regions, each contributing to a diverse landscape encompassing sunlit beaches, rugged mountains, and historically rich towns, as we can see in Figure 1.However, realizing this potential is contingent upon addressing existing challenges.
The city of Aracaju, the dynamic capital, anchors the bustling Polo Costa dos Coqueirais.This region represents Sergipe's economic and tourism epicenter, characterized by pristine beaches.It serves as the primary gateway to the various wonders scattered The Brazilian Tourism Map defined the tourist regions of Sergipe (Brasil Ministério do Turismo 2022).In general, this document guides the preparation and implementation of public policies by the Ministry of Tourism.In Brazil 2021, 338 tourist regions were defined, of which five belong to the state of Sergipe, as shown in Table 1.Only some municipalities are part of a tourist region since they must meet criteria jointly established by state agencies and the Ministry of Tourism.Municipalities are categorized (A, B, C, D, or E) due to the performance of their tourism economy, with A being the best classification and E being the worst.
Of the 75 Sergipe municipalities, 45 were included in the Tourism Map and constituted the formation of five tourist regions in the state, as we mentioned: Polo Costa dos Coqueirais, Polo dos Tabuleiros, Polo Serras Sergipanas, Polo Sertão das Águas, and Polo Velho Chico.Most municipalities in Sergipe were classified in Categories "D" and "E", which indicates that tourist activity is incipient in most of Sergipe.This is not a particularity of Sergipe since, according to Santos et al. (2018), the supply structure of labor in the tourism sector is incipient in 90.6% of Brazilian municipalities.Only the capital, Aracaju, was classified as Category "A".Table 2 shows some indicators of Sergipe's tourist regions for 2020 to understand this regionalization better.
Polo Costa dos Coqueirais is home to almost 50% of the state population and accounts for approximately 55% of Sergipe's GDP.This tourist region aggregates all the municipalities that form the Metropolitan Region of Aracaju (Aracaju, Barra dos Coqueiros, Nossa Senhora do Socorro and São Cristóvão).On the other hand, Polo dos Tabuleiros accounts for 5.19% of the state's GDP.Although Polo Velho Chico has one of the main tourist destinations in the state, the Xingó Canyons, its GDP per capita, the highest among tourist regions, is justified by the presence of the São Francisco Hydroelectric Company, as pointed out by Ribeiro, Jorge (2019).Source: (1) Most recent year available for GDP data.
Despite these unique offerings, several municipalities in Sergipe remain relatively unnoticed, categorized as "developing" or "emerging" concerning tourism infrastructure.This pattern reflects a broader nationwide trend, emphasizing the necessity for strategic investments to unlock the latent potential of these inland regions.An additional impediment lies in accessibility challenges.While Aracaju boasts an international airport, venturing into the interior entails navigating winding roads and limiting public transportation options.Notwithstanding these obstacles, Sergipe finds itself at a pivotal juncture.Its landscapes, culture, and authentic experiences hold substantial allure for discerning travelers.Prioritizing accessibility, endorsing responsible development practices, and adeptly showcasing its hidden treasures could herald a transformative chapter in Sergipe's tourism narrative.
3 Inter-Regional Input-Output System for Tourist Regions, Databases, and Indicators The construction of the interregional system used the Interregional Input-Output Adjustment System -IIOAS method, widely employed in the international literature for several countries worldwide, such as Brazil (Haddad et al. 2017), Egypt (Haddad et al. 2016), Greece (Haddad et al. 2020a), Indonesia (Hulu, Hewings 1993), and Mexico (Haddad et al. 2020b).The IIOAS is a hybrid method that blends data provided by official agencies, such as the Brazilian Institute of Geography and Statistics (IBGE, Portuguese acronym), with non-census techniques for estimating unavailable information.The key advantages of IIOAS lie in its alignment with the national IO matrix data and the flexibility of its regionalization process, which can be applied to any country that publishes its national Supply and Use Tables (SUTs) and provides a system of sectoral regionalized information (Haddad et al. 2015).
The IIOAS method is recommended in contexts where statistical information is limited.Nonetheless, the method demonstrates adherence, consistency, and robust results.In the absence of an official IO matrix for the state of Sergipe, we use the latest official Brazilian IO matrix for the base year 2015, which comprises 67 sectors (or industries) and 127 commodities (IBGE 2018), to generate an interregional system that includes the Sergipe's tourist regions.
The Brazilian IO matrix is disaggregated according to sectoral production in Sergipe and the rest of Brazil.In other words, input usage, consumption of final goods, and value-added payments in Sergipe are generated as components parts of the national economy.This approach allows the construction of a Sergipe-specific matrix with unique characteristics regarding technical coefficients and production multipliers.Furthermore, due to the scarcity of regional information for all activities, it was necessary to reduce the number of sectors to 59, as shown in Appendix, Table A.1.Figure 2 summarizes the stages of the IIOAS method.
The first column presents the data required for constructing the interregional system, i.e., the SUTs of Brazil and the regional shares of production and final demand vectors.From the second column onward, the estimation process stages are depicted.After constructing the regional vectors, the regional trade matrices are estimated based on the following steps: Number 2, 2024 Source: Author's own.Note: The subscripts s, c, r, and f mean sectors, commodities, regions, and final demand components, respectively.

Figure 2: Steps of the IIOAS method
Step 1 Organizing regional shares of production and final demand components using municipal data from the state of Sergipe and the rest of Brazil (see Appendix, Tables A.2 and A.3).
Step 2 Estimating the domestic sales of each industry by region (DOM Sales), which can be done by excluding respective exports (x) and stock variation (sv) from the corresponding gross production vector (go), that is: where i refers to a given industry, and R represents a certain region of the state of Sergipe or the rest of the country.
Step 3 Estimating the total demand for each domestic (dom dem) and imported goods (imp dem) in each region if the demand structure of respective users follows the preference patterns of national demand.
Here j refers to a given input, IC refers to intermediate consumption flows, gfcf refers to gross fixed capital formation, hc household (consumption expenditure), and ge refers to government expenditure.
Step 4 Estimation of trade matrices representing the transactions of each commodity between origin and destination for each industry (intrasectoral flows), the so-called SHIN matrices.
In equation ( 4), d represents any of the 6 regions of the state of Sergipe or the rest of the country, where sales and consumption of each good occur.The term F (i) defines the pattern of international trade for goods (sectors), with values closest to 1 indicating non-tradable sectors or "local goods".Table 3 shows the values used.As expected, most non-tradable sectors are service activities.
This means that for each sector there is a proportional matrix (SHIN table) to distribute the total value of trade (sales and purchases) among all regions.Table 4 shows the distance matrix (dist j,d ), which refers to the road distance in kilometers between origin and destination, where the reference point for each region is the municipality with the highest GDP in 2015.
Step 5 The calculation of intraregional and interregional flow matrices ("initial values") between any combination of o and d is expressed in equations ( 6) and ( 7): REGION : Volume 11, Number 2, 2024 Step 6 Balancing the trade matrices to equate the supply and demand of each commodity using the bi-proportional adjustment method.
Given that the construction of the inter-regional system requires data from different statistical sources, a system balancing procedure is performed, which was carried out using the bi-proportional adjustment method (RAS1 ), ensuring consistency and balance between supply and demand.
Step 7 Finally, the combination of transactions within and between the different regions of the sample enables the generation of an inter-regional system related to the trade of intermediate goods.

Databases
We obtained information on sectoral production from different municipal data sources.For the agriculture sector, the value of production from temporary and permanent crops is aggregated directly from the Municipal Agricultural Production Survey (PAM, Portuguese acronym) (IBGE 2015a) for the year 2015.For the livestock sector, the value of animal production is considered from the Municipal Livestock Survey (PPM, Portuguese acronym) of 2015 (IBGE 2015c).For the forestry production sector, the values of production in silviculture and plant extraction are combined from the Plant Extraction Production Survey in 2015 (IBGE 2015b).For the remaining 56 productive sectors, regional shares are measured using the following proxy variables: (i) wages paid to formal workers and (ii) wages paid to formal and informal workers.The choice of proxy variable is made based on the characteristic of each sector, with information from the Annual Employment Information Report (RAIS, Portuguese acronym) for industrial activities and microdata from the 2010 Demographic Census (IBGE 2011) for the service sectors (see Appendix, Table A

.2).
A new regional distribution is organized to represent workforce employment in each sector.This new approach allows sectoral employment per production unit to be flexible and not bound to a fixed rate, as established in the 2015 Brazilian IO matrix.In other words, a given sector employs proportionally depending on the peculiarities of each region.Due to the scarcity of municipal data in primary sector surveys, the employment distribution follows the regional share of production in corresponding activities.We use the number of active employment contracts as of December 31, 2015, available in RAIS (MTE 2023) for industrial activities.For service sectors, the total number of employees, both with and without formal contracts, is estimated using microdata from the 2010 Demographic Census.
For government consumption expenditures, we use the participation of each tourist region in the value-added of the public administration in Sergipe.To do this, we aggregate the municipal values provided at the municipality level by IBGE (2017) (see Appendix, Table A.3).
Given the unavailability of other proxy variables at the municipal level, the regional share of the remaining macroeconomic aggregates follows the regional distribution of sectoral production.To do this, we adopted new assumptions that gross capital formation, household consumption, and consumption of nonprofit institutions serving households (NPISH) are proportional to regional production in monetary terms.
The data on foreign trade for tourist regions is obtained from the Federal Government's Comex Stat.In this case, it is necessary to reconcile the commodities classified under the Harmonized System Code (HS4) with the 127 commodities in the 2015 IO matrix.The sectoral regionalization is prepared by applying the proportions of each commodity's exports to the weighted values of exports in the matrix.For commodities in the IO matrix for which data is not available in Comex Stat, the total export value of each commodity is multiplied by the share of each region in total output.
However, the values provided in Comex Stat consider the municipality of the exporting company rather than the municipality of origin of the commodities.For the state of Sergipe, where commodity distribution varies among regions, the use of such data leads to significant distortions in the interregional system.For example, the Polo Costa dos Coqueirais accounted for 89.3% of agricultural exports in 2015 while being responsible for only 15.5% of the state's total output.Considering these issues, Haddad et al. (2016Haddad et al. ( , 2017Haddad et al. ( , 2020a,b) ,b) suggest relying on the regional distribution of sectoral production.
For the identification of tourism-related activities (TCAs) in Brazil, we use the study "Economia do Turismo -Uma Perspectiva Macroeconômica 2003-2009" (IBGE 2012).According to this study, TCAs accounted for 3.6% of the country's gross value added in 2003.Moreover, tourism comprises the following activities: i) restaurants and accommodation services, ii) passenger transportation, iii) travel agencies and tour operators, and iv) recreational and entertainment services.Matching this information with the IO matrix, we identify six TCAs: S34 -Land transportation; S35 -Water transportation; S36 -Air transportation; S38 -Accommodation; S39 -Food services; S57 -Artistic, creative, and entertainment activities; and S50 -Other administrative and support services.The last activity includes Travel Agencies.
However, given the absence of a TSA in Brazil and Sergipe, using these sectors directly without any statistical treatment would overestimate the weight of tourism activities in the economy.Thus, it is necessary to disaggregate these sectors' TCAs.Based on information from the wage mass of RAIS, Gonçalves et al. (2020) constructed weights for the disaggregation of TCAs in Brazil.According to these authors, the weights had low variability between 2010 and 2015.For the state of Sergipe, the Institute of Applied Economic Research (IPEA, Portuguese acronym) provided monthly sectoral weights for 2015, as shown in Table 5.In other words, these weights represent the size of tourism in each TCA.
Due to minor weight variations throughout 2015, we consider the weight for December.It can be observed that Air transportation, Accommodation, and Travel agencies have the highest weights, with 95%, 86%, and 83%, respectively, of these sectors corresponding to tourism activities.On the other hand, sectors such as Culture and leisure, Food Services, and Non-metropolitan land transportation have the lowest weights.The Water transportation sector will not be considered as its weight in Sergipe was zero.Based on these weights, the trade flows of the corresponding sectors in the IO matrix were disaggregated.Thus, the matrix now recognizes six additional tourism sectors: Tourist land transportation, Tourist air transportation, Tourist accommodation, Tourist food services, Professional tourist services (travel agencies), and Artistic, creative, and entertainment tourist activities (culture and leisure).The analyzes in the results section (Section 4) will refer to these activities.

Indicators
To structurally evaluate the TCAs in the tourist regions of Sergipe, we calculate the simple production and employment multipliers and the backward-forward indexes.To define these indices, the starting point is the solution of the IO model, formally expressed as: where x is the output vector, L = (I − A) −1 is the Leontief Inverse matrix, in which I is an Identity matrix, A is the Technological matrix and y is the final demand vector.The simple production and employment multiplier of sector j can be defined, respectively, as m(o) j = n i=1 l ij and m(h) j = n i=1 a n+1,i l ij , in which l ij is the Leontief inverse elements, and a n+1,i is the employment coefficient, that is, the ratio between employment and the output in sector i.
The indices of Rasmussen (1958) and Hirschman (1958) measure the degree of backward and forward linkages of a given productive structure.The indices are expressed by a ratio between the average of the impacts of the sector and the total average of the economy, that is: where U oj is the backward linkage (BL), and U io is the forward linkage (FL), n is the number of sectors.The sector is considered a key sector when it presents both indices above one and, therefore, when it has intermediate purchases and sales above the economy average.

Results and Discussion
The first three tables (Tables 6, 7, 8) provide an exploratory analysis of the inter-regional system to assess the generation of value-added and the regional composition of trade flows between the tourist regions of Sergipe.We estimated that tourist activities accounted for only 1.53% of the state GVA in 2015.IPEA estimated the weight of tourism in the Northeast region and Brazil at 2.1% and 2.2%, respectively, considering occupation data in December 2014.When considering wages in the formal labor market, Gonçalves et al. (2020) estimated at 3.02% the weight of TCAs in the total GVA of Sergipe in 2015.Without the disaggregation of TCAs from the coefficients shown in Table 5, the weight of the "tourism sector" in the total GVA of Sergipe would be overestimated by 3.7 times, that is, 5.6%.
Table 6 presents the GVA distribution of the TCAs in Sergipe's tourist regions.We can see an intense concentration in the Polo Costa dos Coqueirais, which accounts, on average, for 83.2% of the value-added generation of TCAs within the state of Sergipe.Except for the Polo Costa dos Coqueirais, the GVA for Tourist land transportation has a more homogeneous distribution among the other tourist regions.The only airport in Sergipe is in the capital, Aracaju, which explains the generation of 100% of GVA for Tourist air transportation in the Polo Costa dos Coqueirais.
Tables 7 and 8 show the share of trade flows among the tourist regions of Sergipe, the rest of Sergipe, and the rest of Brazil based on the origin of purchases and destination of intermediary sales, respectively.We highlighted the intra-regional flows on the main diagonal: purchases and sales made within the region.
The Polo dos Tabuleiros and the Polo Costa dos Coqueirais have the highest degree of self-sufficiency among the tourist regions of Sergipe since 14.7% and 12.7% of their  purchases and 17% and 12.3% of their sales, respectively, have region itself as origin and destination (see Table 7).Sergipe is the smallest state in Brazil, so we can see the substantial importance that the rest of Brazil has in the composition of trade flows for all the tourist regions in the state.
The origin of purchases from the other tourist regions (R2 to R5), except for those originating in the region itself, is greater in the Polo Costa dos Coqueirais than the sum of purchases originating in the other regions of Sergipe.The Polo Velho Chico (92.5%) and the Polo Sertão das Águas (90%) have the greatest dependence on the rest of Brazil about the origin of their purchases.
The relative importance within the state of Sergipe of the Polo Costa dos Coqueirais also appears in the sales' destination, as shown in Table 8.The Polo Tabuleiros is the tourist region that proportionally sells fewer inputs and goods to the rest of Brazil, whose region accounts for 59.6% of the destination of its intermediary sales.
Table 9 presents the simple production multipliers by TCA and tourist regions in Sergipe.As it is an inter-regional system, these multipliers are broken down into intra (region itself), inter (spillover effect), and total (sum of the two previous ones).The last row of the table shows the regional multipliers, which consider all economic sectors per region.A significant advantage of these multipliers is the possibility of explicitly measuring the spillover effect to other regions, which can help in elaborating and implementing tourism policies in Sergipe with a focus on regional production chains.Moreover, Fleischer, Freeman (1997) warn about the importance of considering the interactions of multiregional models not to underestimate the multiplier effects of tourism.The highest regional production multipliers are from the Polo dos Tabuleiros and the Polo Costa dos Coqueirais.For the first one, for every variation of $1 in its final  demand, the entire economy would generate $1.68, with $1.06 in the region itself, and $0.63 would leak to other regions.The lowest regional leakage effect is from the Polo Serras Sergipanas (0.57) and the highest from the Polo Sertão das Águas (0.64).
From the sectoral point of view, the interregional multipliers differ more among Sergipe's tourist regions when compared to the total multipliers, which are more similar across regions.For instance, the simple production multiplier of Tourist land transportation varies between 1.94 and 1.99 between tourist regions.It means that for each variation of $ 1 in its final demand, the economy would produce between $ 1.94 and $ 1.99 depending on the region considered.However, the spillover effect (inter multiplier) varies between 0.88 and 0.97.For each variation of $ 1 in the final demand of Tourist land transportation in Polo dos Tabuleiros, for instance, the entire economy would have to produce $ 1.98 to meet this variation, with $ 1.09 being produced in the region itself and $ 0.90 would be leaked to other regions.
The highest total production multiplier in all tourist regions, including the above regional multiplier, is the Tourist air transportation, with values ranging between 2.07 and 2.15.However, in all tourist regions, this sector has a strong spillover effect (inter).For Polo Velho Chico, for example, for every $ 1 variation in the final demand of this sector, the entire economy would need to produce $ 2.15, but only $ 1.01 would be in the locality itself, and $ 1.14 would leak to other regions.According to Souza (2015), the tourism sector had a production multiplier 1.31 in the Brazilian Northeast.
Table 10 presents the simple employment multipliers by TCA and tourist region in Sergipe.The last row of the table shows the regional multipliers, which consider all economic sectors.Generally, there is greater regional variability in the total multiplier and a smaller one in the inter-regional employment multiplier.Furthermore, the spillover effect (inter) is low in all ACTs in all tourist regions since the activity is developed locally.Ribeiro et al. (2017) pointed out a similar result when estimating the impact of tourist spending in the Brazilian Northeast.These authors observed a low effect of job leakage outside the region.These results highlight the comparative advantage of tourism in the Brazilian Northeast, driven by the region's natural resources and development potential, as corroborated by Ribeiro et al. (2022).
The highest employment multiplier among the TCAs is that of Artistic, creative, and touristic entertainment activities, varying between 40 and 89 among the tourist regions, even well above the regional multipliers.This means that, for every $ 1 million variation in the final demand of this sector, between 40 and 89 jobs would be created directly and indirectly depending on the region.For each variation of R$ 1 million in the final demand of this sector in Polo Velho Chico, for instance, 89 jobs would be created throughout the economy, 84 of which would be in the region itself, and 4 jobs would spillover to other regions.
Except for the Polo Costa dos Coqueirais, the employment multiplier of Tourist air transportation is zero in all tourist regions.This means that all jobs generated due to variations in the final demand of this sector would be generated outside the respective regions.This result is consistent with what has already been shown in Table 6.The employment multiplier of Tourist food services is also relevant across regions.Its spillover effect of Polo Velho Chico is slightly higher than the economy average.According to Souza (2015), the main activities that contributed to the generation of employment in the Brazilian Northeast were Accommodation, road transportation of passengers, and food services.
Only Tourist air transportation and Professional tourist services have shown employment multipliers smaller than the regional multiplier (considering all economic sectors) in all the tourist regions, except for the last sector in Polo Sertão das Águas.
Table 11 presents the results of the Hirschman-Rasmussen (HR) indices by TCA and Sergipe's tourist regions.We have highlighted in red and blue, respectively, the aboveaverage forward and backward linkages.Tourist land transportation is the only TCA ranked as a key sector across all tourist regions, i.e., both indices above one.Prado (1981) a nd Guilhoto et al. (2005) state that key sectors should be considered strategic to stimulate economic growth.A similar result for the Brazilian capital (Federal District) was found by Takasago, Mollo (2011).They identified that the road transportation and intercity tourism sector was also considered a key sector along with the recreational and cultural activities sector.
In general, backward linkages are greater than forward linkages, which means that tourism activities buy more inputs from other sectors than they sell.This result is expected and consistent with previous studies carried out for Bermuda (Archer 1995), Seychelles (Archer, Fletcher 1996), China (Oosterhaven, Fan 2006), East Asia (Blake 2008), Brazil (Takasago et al. 2010), South Korea (Lee, Hlee 2021) and Indonesia (Kumara et al. 2021).This occurs because tourist activities mostly meet final demand.
The Tourist accommodation sector, according to Miller, Blair (2022) can be classified as dependent on inter-industry supply as it only presents purchases above the average for the economy (BL > 1) in all tourist regions, except Polo dos Tabuleiros.Most tourist activities are not strongly connected with other sectors since their intermediate purchases and sales are below average (BL and F L < 1).Gabriel et al. (2020) state that industrial segments are more expected to be classified as key sectors since they purchase and sell a greater diversity of activities.An example of this for the state of Sergipe is that the four key sectors, according to Ribeiro, Leite (2012), are industrial: Food and beverages, Textiles, Paper and cellulose, and Rubber and plastic.
For Brazil, Casimiro Filho, Guilhoto (2003) identified six key sectors of the tourism segment: air transportation, travel agencies, auxiliary activities to air transportation, Accommodation, restaurants, and other food service establishments.It is noteworthy, however, that these authors did not perform any statistical treatment regarding the weight of the TCAs.
Our findings offer significant socioeconomic insights for Sergipe, and it can serve as a case study for all Brazilian states and similar regions worldwide.Identifying key sectors within the tourism industry, characterized by high economic multipliers, presents an opportunity to bolster economic development.In addition, we highlight the regional dynamics of tourism activities.By acknowledging and leveraging the distinct economic contributions of different regions, regional policies can work towards reducing disparities and promoting more inclusive development.
Understanding employment multipliers across various tourism activities provides a valuable tool for crafting labor market policies.Policymakers can prioritize sectors with higher job creation potential, contributing to local and regional employment opportunities.Furthermore, the study's insights into income generation and inequality highlight the potential of tourism to play a role in addressing socioeconomic disparities.Crafting targeted policies that harness the economic benefits of tourism can contribute to reducing REGION : Volume 11, Number 2, 2024 income inequality and enhancing overall economic well-being across the state.

Conclusions
This research advances the estimation of an inter-regional IO system specified for tourist regions in Sergipe and disaggregates the tourism activities.By building this system, regional heterogeneity is addressed.Thus, issues such as productive linkages, employment and income multipliers and spillovers can be discussed by tourist regions.However, the ideal scenario is the Brazilian statistical officers' availability of the Tourism Satellite Accounts.Thus, the impacts of these activities can be estimated more precisely since, in tourist activities, only what is consumed by tourists will be considered.
The exploratory analysis revealed that TCAs accounted for only 1.53% of Sergipe's state GVA in 2015.Comparatively, the weight of tourism in the Northeast and Brazil was estimated at 2.1% and 2.2%, respectively.Furthermore, without disaggregating the TCAs from the presented coefficients, the weight of the tourism sector would be overestimated 3.7 times.
The Polo Costa dos Coqueirais was identified as the region concentrating the largest generation of GVA from the TCAs in Sergipe, corresponding to 83.2% of the state's total.Furthermore, Tourist air transportation had the highest production multiplier, varying between 2.07 and 2.15 in all regions.However, tourist regions also showed a strong spillover effect, indicating that part of the generated production is destined for other regions.As for employment multipliers, artistic and creative activities and tourist shows had the highest values.The spillover effect of jobs to other regions was low in all TCAs and tourist regions, indicating that the activity is predominantly developed locally.
Tourist land transportation was a key sector in all tourist regions of Sergipe.In general, backward linkage indices were higher than forward linkage indices, indicating that TCAs purchase more inputs from other sectors than they sell.Tourist activities would mainly meet the final demand.The Tourist accommodation sector depended on inter-industry supply, with purchases above the average in all regions except for the Polo dos Tabuleiros.Most tourist activities are not strongly connected to other sectors, as their intermediate purchases and sales are below average for the economy.
We used an unprecedented method in the Brazilian literature that disaggregates tourist activities by sector and region.In addition, with the identification of TCAs in Sergipe, it was possible to measure the spillover effect to other regions explicitly.It can be useful for elaborating and implementing tourism policies focused on regional production chains.Furthermore, researchers can replicate this method for countries and regions that, like Brazil, do not have a Tourism Satellite Account.
The main limitation of the research, however, is that the technical coefficients of disaggregated tourist activities, for example, tourist accommodation and non-tourist Accommodation, are the same.Ideally, we would have specific coefficients for each tourism sub-activity, which is only possible with the Satellite Account.
The utilization of disaggregated data from this study offers a concrete foundation for crafting targeted policies and interventions in Sergipe's tourism sector.With a detailed understanding of various TCAs' specific contributions and regional distribution, stakeholders can tailor strategies to each tourist region's unique needs and potential.This granularity in data analysis becomes instrumental for optimizing resource allocation, fostering economic growth, and a practical approach to tourism planning.
Addressing regional development disparities is imperative, as the study reveals the concentration of tourism-related economic activities.Actionable measures such as strategic infrastructure improvements, direct support for local entrepreneurial ventures, and targeted promotion of distinctive attractions must be implemented to stimulate economic growth in less-developed tourism regions.These interventions should spur economic development, reduce regional inequalities, and cultivate a more diversified and equitable tourism landscape in Sergipe.Moreover, it is essential to recognize the significance of place identity in shaping residents' perceptions and aspirations for their region's future.Residents' deep-rooted connections to Sergipe's cultural heritage and economic potential REGION : Volume 11, Number 2, 2024 can inform decision-making processes and foster inclusive, community-driven approaches to development and planning.Sustainable practices and specific measures should involve implementing stringent environmental regulations, fostering community engagement initiatives, and promoting eco-friendly tourism practices that resonate with residents' sense of place identity.

Table 9 :
Production multiplier by tourist activity and tourist region of Sergipe, 2015 Source: Author's own based on IO matrix.REGION : Volume 11, Number 2, 2024

Table 3 :
Value of the term F (i) for the sectors of the IO matrix, 2015

Table 4 :
Distance matrix (in km)between the regions of the interregional system, 2015 Source: Author's own based on information from Google Maps.

Table 5 :
Weight for the disaggregation of TCAs in Sergipe, December 2015

Table 6 :
GVA regional distribution of TCAs in Sergipe, 2015

Table 7 :
Share of the origin of trade flows by tourist region of Sergipe, 2015

Table 8 :
Share of the trade flows' destination by tourist region of Sergipe, 2015

Table 10 :
Employment multiplier by tourist activity and tourist region of Sergipe, 2015Source: Author's own based on IO matrix.

Table 11 :
HR indexes of tourist activity by tourist region of Sergipe, 2015Source: Author's own based on IO matrix.