Subregional Economic Cooperation - Corridor Road Investment project (RRP BAN 53382)

ECONOMIC AND FINANCIAL ANALYSIS

A. Introduction

1. Economic rationale. Road transport is ’s dominant mode of transport, and the road network serves as the backbone for domestic connections and regional and international trade. Bangladesh is experiencing positive economic growth, rising incomes, increased domestic consumption, and additional trade activities resulting from deeper engagement in the global value chain, causing road traffic to increase at an average rate of 6.6% per year between 2016 and 2020.1 However, the limited capacity and deteriorating condition of roads—about 25% are in poor condition—have served as major obstacles to stable economic growth and caused serious road safety issues. To resolve these issues, the Government of Bangladesh is upgrading its major roads, especially those that constitute subregional trade corridors, to international standards, with the goal of facilitating international trade.

2. Project road. The Dhaka–Sylhet–Tamabil corridor is part of National Highway No. 2 and Asian Highway No. 2 and is considered a critical international trade corridor. As a result of customs policy reform and trade and transport infrastructure development, the importance of Sheola and Tamabil land ports in regional trade is expected to increase, along with the traffic on the corridor connecting the border with the national capital city of Dhaka. However, the existing road has only two lanes with no provision for slow-moving and nonmotorized vehicles; is of poor quality because of a multiyear maintenance backlog; and poses a significant safety risk. The proposed investment project will finance construction works to widen about 210 kilometers (km) of the road to a four-lane carriageway, with design features to assist the elderly, women, children, and differently abled persons.

3. Economic analysis. An economic evaluation of the proposed project components was undertaken using the Highway Development Model 4 (HDM-4), and following Asian Development Bank (ADB) guidelines. 2 The model requires input data on traffic, road geometry, condition, pavement structure, and material characteristics of the existing road; maintenance and road improvement costs; and vehicle operating cost (VOC) and value of time parameters. The improvement of road corridors will result in savings to road users and the society as a whole in the form of reduced VOCs and travel time costs for passengers and freight traffic. The costs to the road agency and road users in the “without” and “with” project cases were estimated and used to derive the project’s net costs and benefits and to calculate the economic viability of the project road sections.

B. Demand Analysis

4. Traffic demand. The project road was divided into six homogeneous sections. The length of each section and the base year (2020) average annual daily traffic (estimated based on the traffic counts conducted on each section of the project roads) are in Table 1.

5. Traffic growth. Traffic growth forecasts were prepared for the analysis period by vehicle class based on the economic growth outlook and the elasticity calculated by comparing historical registered vehicle growth trends and observed gross domestic product, population, and per capita income during 2011–2019. The impact of the proposed economic corridor development was also considered in forecasting the freight demand.3 The derived traffic growth rates are summarized in Table 2.

1 ADB. 2021. Country Assistance Program Evaluation for Bangladesh. Manila. 2 ADB. 2017. Guidelines for the Economic Analysis of Projects. Manila. 3 ADB. Manila:. Bangladesh Economic Corridor (Draft Report). Unpublished. 2

Table 1: Base Year Traffic on Project Road Sections AADT Length AADT Non- (km) Motorized Traffic Motorized Traffic SN Road Name Total Urban Rural Total PV Bus FV 1 Dhaka–Narsingdi 33.78 15.77 18.01 20,142 5,960 5,227 8,955 4,848 2 Narsingdi–Bhairab 32.85 17.58 15.27 14,624 4,417 3,602 6,605 1,247 3 Sarail–Jagdispur 34.80 7.66 27.14 17,449 5,067 2,138 10,244 1,420 4 Jagdispur–Sayestaganj 34.60 10.14 24.46 10,890 4,593 2,663 3,634 313 5 Sayestaganj–Sherpur 35.70 10.80 24.90 10,610 4,391 2,244 3,975 228 6 Sherpur–Sylhet 35.90 9.94 25.96 17,640 3,767 3,245 10,628 135 AADT = average annual daily traffic, FV = freight vehicle, km = kilometer, PV = passenger vehicle. Source: Asian Development Bank’s estimates.

Table 2: Traffic Growth Rates by Vehicle Class 2020–2024 2025–2029 2030–2034 2035–2039 2040–2044 Parameter Growth Rates (%) GDP 7.00 6.80 6.60 6.40 6.20 PCI 5.53 5.38 5.23 5.09 4.94 Population 1.39 1.35 1.30 1.25 1.20 Traffic Growth Rates (%) Motorcycles 15.1 13.9 12.8 11.8 10.8 Auto-rickshaw 7.7 7.1 6.6 6.1 5.6 Car, jeep, or van 6.8 6.3 5.8 5.3 4.9 Bus 4.9 4.5 4.1 3.7 3.4 Goods vehicles 8.9 9.9 7.4 7.4 4.3 PCI = per capita income, GDP = gross domestic product Source: Asian Development Bank’s estimates.

6. Diverted and generated traffic. There are no routes in direct competition with the project road, and therefore local diverted traffic was not considered in this analysis. However, the project road is part of Asian Highway No. 2, and diversion to the project road is anticipated for freight moving between (i) other areas of and its northeastern states, and (ii) Chattogram port and the northeastern states of Bhutan and India. Such diverted traffic is conservatively estimated at 10% of the freight demand based on the travel time reduction resulting from the road condition improvement, which does not include likely future increase that may occur with further improvement of bilateral and multilateral trade among Asian countries. In addition, an estimated 10% generated traffic is also applied for both privately owned passenger vehicles and freight vehicles, as the improvement will (i) encourage communities to travel for jobs that are currently difficult to access, (ii) open new markets for communities to sell agricultural products at competitive prices and in turn increase production levels, (iii) motivate inhabitants to attend better health and educational facilities, and (iv) facilitate cross-border trade activities.

C. Economic Analysis

7. Key assumptions. The economic analysis was conducted in the domestic price numeraire (Bangladesh taka). All costs and benefits were valued in monetary terms as of February 2021 and expressed in economic prices. A discount rate of 9% was used for calculating the net present value. Key economic analysis assumptions are summarized in Table 3.

Table 3: Input Parameters Used for the Analysis Parameter Value Analysis period from opening year 20 years Discount rate 9% Construction period 5 years for the first two sections; 6 years for the rest Construction start year 2022 Opening year for road traffic 2027 for the first two sections; 2028 for the rest Source: Asian Development Bank estimates. 3

8. Construction cost. The economic cost of constructing the project road was estimated based on the financial estimate provided by the Roads and Highway Department (RHD) of Bangladesh. 4 The economic cost of construction includes the costs of civil works, construction supervision, project management, land acquisition and resettlement, and physical contingencies, but excludes price contingencies and removes taxes and duties. A shadow exchange rate factor of 1.028 was applied to traded goods in civil works cost, VOCs, and emission costs. A shadow wage rate factor of 0.79 was applied to unskilled labor components, based on wages in the construction and agriculture sectors. The economic cost of relocation and resettlement is based on the economic value of resettlement allowances given to displaced families and of replacement of structures. The cost of land acquisition is calculated based on the loss of agriculture crops during the entire analysis period and discounted to the present value. In addition, the cost of the institutional strengthening of the RHD is included as part of the construction cost in this analysis. Table 4 presents the estimated economic cost of construction per km for the homogeneous sections by urban and rural subsections.

Table 4: Economic Cost of Construction (Tk million per km) Economic Cost Number Road Name Urban Rural 1 Dhaka–Narsingdi 905.42 591.20 2 Narsingdi–Bhairab 838.10 545.19 3 Sarail–Jagdispur 866.91 564.39 4 Jagdispur–Sayestaganj 840.34 546.67 5 Sayestaganj–Sherpur 963.86 629.02 6 Sherpur–Sylhet 806.66 524.22 km = kilometer, Tk = taka. Source: Asian Development Bank estimates.

9. Maintenance costs. The periodic and routine maintenance costs are based on the estimates provided by the RHD. The economic cost of periodic overlays for existing roads was estimated at Tk1,177 per square meter and for improved roads at Tk1,471 per square meter. The periodic overlay was applied following the prevailing maintenance strategy and guidance received from the RHD. The economic cost of annual routine maintenance for existing roads was estimated at Tk82,400 per km, and for improved roads at Tk164,800 per km.

10. Residual value. The project includes newly constructed elements that have a usable life much longer than the analysis period of 20 years. Thus, a residual value of 20% at the end of the analysis period was estimated by using straight-line depreciation method assuming a usable life span of 40 years for structures; 25 years for embankments, base, and sub-base; and 20 years for all other elements, including the bituminous course.

11. Project benefits. The quantified benefits of the project include travel time savings and VOC savings for motorized vehicles resulting from higher vehicle speeds and improved riding quality with the project (Table 5). Reductions in carbon dioxide (CO2) emissions were also quantified as a potential benefit. Under the without-project scenario, speeds are significantly reduced in the analysis period as road capacities are reached earlier than under the with-project scenario. The rate of road degradation under the without-project scenario is also higher compared to the with-project scenario as more volume is traveling on one single lane.

4 Roads and Highway Department. Dhaka. Consulting Services for Reviewing and Updating Designs (including new designs as necessary) of Highway, Bridge, Flyover, Interchange and other Road Infrastructure, Updating Feasibility Study Report, Resettlement Action Plan, Preparation of Tender Documents and Other Pre-construction Activities for the Improvement of the Existing Dhaka (Katchpur)–Sylhet National Highway () into a 4-Lane Highway with Service Lane on Both Sides. Unpublished. 4

Table 5: Operating Characteristics in the Analysis Period Average Average Average Roughness Vehicle Speed Vehicle Operating Cost Scenario (m/km) (km/h) (Tk/vehicle-km) Without project 5.93 40.13 26.82 (car), 39.19 (medium truck) With project 3.99 63.10 22.62 (car), 31.10 (medium truck) h = hour, km = kilometer, m = meter, Tk = taka. Source: Asian Development Bank estimates.

12. Vehicle characteristics and costs. Data required for the HDM-4 analysis were obtained from the detailed project report, including inventory and condition surveys, and the material and pavement investigations. The HDM-4 model predicts VOCs as a function of operating conditions, based on inputs that include the vehicle technical and operational characteristics, vehicle prices, tire prices, fuel prices, and vehicle maintenance and operation costs. The study adopted (i) vehicle and tire prices (excluding taxes and labor cost for vehicle maintenance and operation) based on the data presented in the draft final report (footnote 3) and updated to 2021 prices; and (ii) fuel prices (excluding taxes and duties) based on market statistics.

13. Passenger time value. For passenger vehicles, passenger time values were calculated based on per capita income, population, estimated working hours, and the employment rate. The time value for non-work trips is estimated at 30% of work-time value.5 The estimated unit time values for passengers are given in Table 6.

Table 6: Value of Travel Time for Passengers (Tk per hour) Vehicle Type Value of Passengers’ Work Time Value of Passengers’ Non-work Time Two- and three-wheeler 130.1 39.0 Passenger car, van, and jeep 212.6 63.8 Bus 99.6 29.9 Tk = taka. Source: Asian Development Bank estimates.

14. Cargo delay. For freight vehicles, the value of cargo delay is based on its opportunity cost calculated as the value of goods carried times the commercial interest rate paid by the owners as an inventory cost. These time values were estimated per hour at Tk4.01 for light trucks, Tk8.09 for medium trucks, Tk10.79 for heavy trucks, and Tk16.18 for multi-axle trucks, considering a reference value of Tk65,000–Tk75,000 per ton of cargo and an interest rate of 9%.

15. Carbon dioxide emissions. CO2 emissions were monetized using a unit value of Tk3,334.0 per ton of CO2 in 2020 prices, adjusted from $36.30 per ton in 2016 prices, and projected to increase by 2% annually in real terms (footnote 1). The project will result in an increase in CO2 emissions, mainly as a result of the generated traffic resulting from the road improvement.

16. Nonmotorized benefits. Benefits for nonmotorized transport were estimated for bicycles and rickshaws in terms of operating cost and travel time savings. HDM-4 predicted a speed increase of about 2–3 km per hour after the improvement of the project road sections, which was translated to time savings by using the unit time value of Tk35 per hour for bicycles and Tk50 per hour for rickshaws. The model also estimated operating cost savings based on the local vehicle characteristics and operational conditions.

5 ADB. 2013. Cost-Benefit Analysis for Development: A Practical Guide. Manila. 5

17. Unquantified benefits. In addition to quantified benefits, the project is expected to deliver other significant economic benefits, which include (i) an expected reduction in traffic accidents and fatalities resulting from improved road geometry and safety facilitates, notably for vulnerable road users; and (ii) improved access to markets; employment, social and educational opportunities; and tourist destinations.6 Furthermore, the project will contribute to the economic growth of Bangladesh through additional trade resulting from the improved Dhaka–Sylhet corridor. The corridor will eventually connect to Tamabil land port and provide connectivity to Akhaura and Sheola land ports via other major roads. Currently, most of the country’s imports are via Benapole (bordering West , India), while most exports are shipped from Chattogram Port (para. 6). An upgraded Dhaka–Sylhet corridor would allow more goods, especially raw materials for various industrial goods, to be imported from northeast India via Tamabil and transported on the Dhaka– Sylhet corridor to the rest of Bangladesh. This would decrease the logistics costs by reducing the congestion at Benapole and Chattogram resulting from limited road connections. In the long term, the project could increase export sector diversification by allowing more imports from various border points, and induce industrialization along the corridor, because of increased accessibility for businesses and workers. These unquantified economic benefits could be significant.

18. Economic analysis. The project’s overall economic internal rate of return (EIRR) is estimated at 14.2%, with an EIRR for individual sections of 10.5%–18.1%. Cost and benefit streams are presented in Table 7. Periodic maintenance is classified as a capital cost due to the large investment amount, while recurrent cost mainly refers to the cost of routine maintenance.

Table 7. Economic Costs and Benefits Streams (Tk million) Increase in Costsa Increase in Benefitsa,b Non- Present Value Year Capital Recurrent MT VOC MT Time Emission Motorized Net Benefits (2021 prices) 2022 12,874.0 0.0 0.0 0.0 0.0 0.0 (12,874.0) (11,811.0) 2023 27,908.0 0.0 0.0 0.0 0.0 0.0 (27,908.0) (23,489.6) 2024 27,287.7 0.0 0.0 0.0 0.0 0.0 (27,287.7) (21,071.1) 2025 29,096.0 0.0 (2,455.7) (104.1) (17.6) (98.1) (31,771.6) (22,507.8) 2026 27,908.0 0.0 (6,209.9) (226.8) (46.9) (112.8) (34,504.3) (22,425.4) 2027 12,634.5 5.5 (1,649.2) 1,005.0 (60.3) 53.7 (13,290.7) (7,924.8) 2028 0.0 17.1 6,320.1 2,951.7 (276.1) 118.3 9,097.0 4,976.3 2029 0.0 17.1 9,532.2 3,902.5 (281.8) 151.6 13,287.4 6,668.5 2030 0.0 17.1 13,518.2 5,111.9 (272.6) 206.6 18,547.1 8,539.6 2031 (620.3) 17.1 20,343.3 7,281.7 (232.3) 397.3 28,393.2 11,993.6 2032 (1,211.5) 17.1 13,141.2 7,585.4 (340.7) 60.4 21,640.7 8,386.5 2033 3,100.7 17.1 2,717.1 8,788.3 (480.5) 1.3 7,908.3 2,811.7 2034 6,056.4 17.1 8,675.3 10,935.2 (443.0) 93.6 13,187.7 4,301.5 2035 0.0 17.1 21,034.5 14,266.8 (312.7) 152.2 35,123.6 10,510.6 2036 0.0 17.1 28,139.5 16,940.4 (275.1) 192.6 44,980.3 12,348.8 2037 0.0 17.1 36,557.6 20,025.6 (276.1) 247.3 56,537.3 14,240.0 2038 (620.3) 17.1 47,606.0 23,448.6 (278.5) 334.0 71,713.3 16,571.0 2039 (1,211.5) 17.1 52,623.4 26,552.3 (369.1) 172.0 80,173.0 16,996.2 2040 3,100.7 17.1 23,713.7 26,684.7 (964.0) 25.9 46,342.4 9,013.1 2041 6,056.4 17.1 31,426.7 28,423.6 (1,072.2) 107.8 52,812.4 9,423.4 2042 0.0 17.1 43,674.2 29,824.2 (1,172.6) 161.3 72,470.0 11,863.2 2043 0.0 17.1 50,777.5 30,915.1 (1,454.0) 200.7 80,422.2 12,077.9 2044 0.0 17.1 57,803.2 31,877.0 (1,804.7) 249.6 88,108.1 12,139.7 2045 (620.3) 17.1 68,304.3 33,152.5 (2,173.2) 468.8 100,355.6 12,685.4 2046 (10,809.7) 17.1 54,155.6 32,971.5 (2,795.6) 122.9 95,247.0 11,045.6 2047 (18,309.8) 11.6 17,204.3 20,027.3 (2,084.2) 5.4 53,451.0 5,686.8 NPV 93,046.7 EIRR 14.2%

6 Sylhet has several tourist destinations well known in Bangladesh such as hill station, , Ratargul Swamp Forest, and Bichnakandi Village. 6

Increase in Costsa Increase in Benefitsa,b Non- Present Value Year Capital Recurrent MT VOC MT Time Emission Motorized Net Benefits (2021 prices) Discount Rate 9.0% ( ) = negative, EIRR = economic internal rate of return, Tk = Bangladeshi taka, EIRR = economic internal rate of return, MT = motorized vehicles, NPV = net present value, VOC = vehicle operating cost. a Compared with without-project scenario. b The share of increased benefits for generated and diverted traffic is estimated at 13% of the total. Source: Asian Development Bank estimates.

D. Sensitivity Analysis

19. Sensitivity tests were carried out to investigate the robustness of the project to the following adverse changes: (i) increase in capital cost, (ii) delay in construction completion, and (iii) decrease in benefits. Despite targeting an accuracy level of 10% in estimating project cost, a 15% increase in capital cost was tested in this analysis, adding a further 5% to cover adverse cost changes that may incur because of the impact of the prevailing coronavirus pandemic. For the same reason, a construction completion delay of 1.5 years was tested, adding0.5 years to the expected construction delay of 1 year. The test of a 15% decrease in benefits was to cover the likely reduction of the predicted traffic growth and time value due to the coronavirus pandemic. Finally, a worst-case scenario combining all the above cases was also tested.

20. The results of the sensitivity analysis (Table 8) confirm that the project is economically viable in the face of all independent scenarios, with EIRR values above the threshold value of 9%. Even under the worst-case scenario, the project remains viable with an EIRR value of 10.5%. The switching values further indicate the viability of the project, requiring an unlikely increase of 89.5% in the estimated capital cost and a decrease of 47.4% in the predicted benefit. Overall, the sensitivity analysis demonstrates that the project is economically robust even under adverse scenarios, and hence the project is recommended for implementation.

Table 8: Economic Analysis Results EIRR NPV Switching Value Scenario (%) (Tk million) (%) Base Case 14.2 93,046.7 1 Increase in capital costs by 15% 13.0 77,450.8 89.5 2 Decrease in benefits by 15% 12.8 63,627.0 (47.4) 3 One-and-half year delay in construction completion 12.6 64,631.7 4 Worst-case (combination of 1, 2, and 3) 10.5 26,146.5 ( ) = negative, EIRR = economic internal rate of return, NPV = net present value, Tk = taka. Note: The switching value indicates the percentage by which cost increases and benefits decrease to result in a net present value of 0. Source: Asian Development Bank estimates.

E. Distribution Analysis

21. Distribution and poverty analysis was carried out following ADB guidelines.7 The share of the population living below the poverty line in the two project-impacted divisions of Dhaka and Sylhet was estimated at 14.2% in 2021.8 Assuming a similar portion of the population is vulnerable to poverty, 28.4% of the population was considered to be poor or vulnerable in the project influenced area in 2021. In estimating the benefits accruing to poor and vulnerable project

7 Footnote 1 and ADB. 2001. Handbook for Integrating Poverty Impact Assessment in the Economic Analysis of Projects. Manila. 8 World Bank Group. 2019. Bangladesh Poverty Assessment. Washington DC. The poverty head count ratio (PHCR) of 0.142 was estimated projecting the weighted average of PHCRs of (0.135), and (0.167) in 2016. 7

beneficiaries, it was assumed that (i) car owners and/or users and bus owners are not poor or vulnerable, (ii) motorcycle and auto-rickshaw owners are equally as poor as the general population, (iii) motorcycle and auto-rickshaw users are less likely (factor of 0.9) than the general population to be poor or vulnerable, (iv) bus users are more likely (factor of 1.5) than the general population to be poor or vulnerable, and (v) almost all laborers involved in the project are likely to be poor or vulnerable. Table 9 presents the distribution of the project-generated benefit of Tk93,046.7 million among the project beneficiaries. The estimated poverty impact ratio of 0.322 exceeds the estimated poverty and vulnerable headcount ratio (poor and vulnerable) of 0.284 in the project influenced area, indicating all poor and vulnerable people are served by the project.

Table 9: Poverty Impact Assessment Vehicle Vehicle Vehicle Item Labor Government Total Owners Passengers Freight NPV (Tk million) 77,529.6 68,916.3 52,987.4 4,015.1 (110,398.8) 93,049.7 Share accruing to the poor & vulnerable (%) 3.44% 39.14% 14.19% 95.0% 10.0% Benefits accruing to the poor & vulnerable 2,666.3 26,976.3 7,516.8 3,814.3 (11,039.9) 29,933.9 (Tk million) Poverty Impact Ratio 0.322 ( ) = negative, ENPV = net present value, NPV = net present value,Tk = taka. Source: Asian Development Bank estimates

F. Sustainability Analysis

22. The project will not generate revenue. Following ADB guidelines, the financial sustainability of the project was assessed using the sustainability of recurrent costs to be assumed by the RHD. 9 The RHD has no financial burden for maintenance during the first 6 years as 6-year performance-based maintenance periods are included in all civil works contracts. Once the 6-year period is over, the RHD will assume responsibility for maintenance. The first 6 years of this tenure will be limited to routine interventions without major periodic interventions. Periodic interventions, including applying an overlay, will be required on an interval of 7 years.

23. The sustainability of the project was also evaluated by assessing historical budget data for road operation and maintenance at the national level. Table 10 summarizes the RHD’s budgetary allocation and expenditure for road maintenance since fiscal year 2016. The maintenance budget increased by 55% from FY2016–FY2020. Annual expenditure varies based on actual maintenance requirements during a year; however, on an overall basis, there has been a significant increase in the maintenance budget during this period.

Table 10: Expenditure on National Highways Maintenance (Tk million) FY2016 FY2017 FY2018 FY2019 FY2020 Actual expenditure 6,894.89 6,983.26 10,953.90 10,487.23 10,669.00 Expenditure per km 1.81 1.83 2.86 2.68 2.73 FY = fiscal year, km = kilometer. Source: Roads and Highways Department.

24. The average annual recurrent costs are estimated at Tk1.04 million per km (at an estimated per km cost of Tk183,850 for routine maintenance and Tk855,927 per km for annualized periodic overlays). While this may be lower than the historical expenditure on the

9 Following the expiration of the performance-based maintenance contracts, the RHD is exploring adoption of an operate, maintain, toll model to ensure asset maintenance over a longer period, and introduce private sector participation with an affordable toll structure on the project road. However, currently no projections or estimates are available regarding generation of tolling revenue from the project roads, and therefore a financial evaluation of the project was not possible. 8

national highways, the project will use higher performance pavement material (polymer-modified bitumen) in construction, which is expected to increase resilience of the project road and reduce the frequency of periodic maintenance. The performance-based maintenance included in the construction contracts will support increased efficiency in routine maintenance through the engagement of the private sector in the initial years following construction. The annual recurrent costs for the entire project road are Tk217 million for about 210 km. This equals about 1.8% of the FY2020 maintenance budget for all national highways managed by the RHD. Although this is a relatively small amount compared to the current annual expenditure, as the number of roads maintained by the RHD increases, adequate road maintenance funding must be in place and maintenance must be planned and implemented in a systematic manner, based on a robust maintenance budget programming framework.