81999 ASSESSING PRIVATE SECTOR CONTRIBUTIONS TO JOB CREATION: IFC OPEN SOURCE STUDY THEME: ACCESS TO INFRASTRUCTURE LINKING POWER SUPPLY TO JOBS: ESTIMATING EMPLOYMENT EFFECTS OF POWERLINKS TRANSMISSION LIMITED PROJECT IN INDIA & BHUTAN1 This study estimates the impact on jobs from power transmission lines constructed by Powerlinks Transmission Limited (PTL), a joint venture company supported by IFC, which helped bring power from a Bhutanese hydropower plant to India. The project had a significant development impact in both India and Bhutan. Unlike most studies that focus on jobs creat- ed through construction and operations and maintenance (O&M) (Category 1 jobs) of power projects, this study looks at employment effects more comprehensively by including estimates for second-order growth effects (Category 2 jobs) - i.e. those jobs that are created as the increase in power supply brought by PTL helps firms expand output and hence create more jobs. The study uses a mix of methodologies including Input Output models, econometric time series models and step-by-step estimation to determine the different types of employment effects. It finds that indirect and induced jobs created through an investment in power are much larger than the number of direct jobs created. But the most significant employment effect is from second-order growth (category 2 jobs that were created because the increased power supply relieved a key constraint for firms in India.) The project also had a broader poverty-reduction impact because the trans- mission lines were constructed through some of the poorest states in India. In addition, a large proportion of induced jobs were in the agricultural sector, creating employment and income for the rural low-skilled population. THE PROJECT Note on Methodology In 2003 IFC committed a loan of US$75 million to Power- The study uses a mix of methodologies for estimating links Transmission Limited (PTL), a joint venture between the job effects. private utility Tata Power Company and state-owned Power Grid Corporation Limited. The project involved 1. Category 1 job effects (which are of three building, owning and operating power transmission lines types - direct, indirect and induced jobs) are in India that enabled transmission of power from the Tala estimated using an Input-Output model. Hydropower Plant in Bhutan to India’s power grid. The 2. Category 2 jobs: second order growth effects Tala project in Bhutan was built and financed by India of increase in power supply are estimated on an understanding that most of the power generated using a time series econometric approach would be bought by India to meet its needs while also (a Vector Error Correction Model). The job increasing revenue for the Bhutanese government. Con- effects due to increased reliability were struction of the transmission lines started in 2003 and estimated using a step-by step estimation they were operational by 2007. model with World Bank Enterprise Survey data supplemented with data on CONCEPTUAL FRAMEWORK power outages from primary interviews. By helping finance the transmission lines from Bhutan to India IFC helped generate two categories of jobs: KEY FINDINGS 1. Category 1: Jobs created during construc- tion, operation and maintenance (O&M) of Finding 1: Indirect and induced job effects are much the transmission lines. These include direct jobs larger than direct jobs created (created by PTL or its contractors), indirect jobs (jobs Category 1 jobs are created first during construction and created in firms supplying material etc to PTL) and later during O&M of the transmission lines. Using a 2006 induced jobs (jobs created because new direct and Input-Output model for India we estimated that the con- indirect jobs create additional demand for house- struction and O&M of the power lines would create a total hold consumption goods etc, which has a spillover 243,000 person-years employment over the 25 years of effect in the economy). the life of the project2. The number of indirect and induced 2. Category 2: Second-order growth effects. jobs created is much larger than the direct jobs created by These are jobs created because the increased PTL, demonstrating that the project has large economy- supply and reliability of power helped firms wide spillover effects. In fact, induced jobs are the largest increase output, and hence employment. contributor to category 1 employment at 144,000 person- 1 Figure 1: Chain of employment effects of the Powerlinks Transmission Project cause the new income generated was spent largely on Figure 2: Total employment effects of PTL food and that in turn created more jobs in the labor-inten- (Person years) sive agricultural sector. 140,000 Finding 2: Significant impact of the project on poverty 120,000 There were two main channels through which poverty 100,000 was reduced. First, jobs created during construction 80,000 added Rs 47,000 lakh (US$ 94 million) to household in- 60,000 come, which was particularly significant since the trans- 40,000 mission lines were constructed through some of the 20,000 poorest states in India. The project used local labor so 0 Construction Operation the construction activity added to household incomes of unskilled and skilled workers who were employed. Sec- Induced Indirect Direct ondly, a large proportion of induced jobs were created in the agricultural sector where poverty is widespread. Since the agricultural sector has the highest employment- to-output ratio and one quarter of total household ex- years. Interestingly, sectors like agriculture and forestry penditure goes to buy primary food the total number of that initially appear to have no direct relationship to power induced jobs created is highest in agriculture and thus transmission in fact were the most strongly affected be- the poverty-reduction impact for rural areas is significant. 2 Finding 3: Largest job effects come from second- Finding 4: Reliability of power supply is a key con- order growth effects (Category 2) i.e. jobs created straint for firms. An improvement in power reliability due to increased power supply, which confirms other (reduction of power outages) helps firms reduce pro- studies that show that power is a major binding con- duction losses, save the cost of running generators, straint for output, growth and employment in India. and increase their output, hence create more jobs. Unlike most studies that only estimate jobs created The reliability of power supply (using power outages as through construction and O&M, this study also estimated a proxy for measuring reliability), and not just quantity how increased power supply helps firms expand output of power supply, is a key constraint for firms. Using data and hence create more jobs. We used an econometric time from the 2005 World Bank Enterprise Surveys in India we series model - Vector Error Correction Model (VECM), to constructed a step-by-step estimation model to establish establish the relationship between electricity consump- an approximate relationship between employment and tion and employment. VECM is a dynamic time series power outages.3 model that allows a dynamic interrelationship between Power outages impose two costs on firms that, if avoid- variables. This time series model allows us to measure ed, can be reinvested in productive activities that increase the long-term relationship between the endogenous output and employment. We estimate two main types of variables (electricity consumption, employment and real costs: a) value of production loss, and b) the additional GDP) and also to estimate the short-term dynamics of cost of using generators. the adjustment to the long-term equilibrium relationship. a. Value of production loss. This is the amount The analysis found a causal relationship between electricity of output that firms lose due to lost production consumption and employment for India, and also for West time, time needed to reset machines, and losses Bengal and Bihar, the two states that received a majority and damages due to interruption of production share of the power transmitted through PTL lines. The processes. The Enterprise Surveys ask firms to esti- model gives us a long-term equilibrium relationship mate these losses, which we apply in our model as between electricity consumption and employment and potential additional sales that would occur if there establishes that the elasticity of employment with respect to were no power outages. Assuming constant sales- electricity consumption is 0.53 for India and 0.24 for West to-worker productivity, it can be estimated how Bengal. In other words, a one percent growth in electricity many workers would be needed to produce these consumption in India will result in 0.53% growth in em- additional sales. ployment for India because additional power consumption will increase out- put and result in employment growth. Steps in the estimation model Discounting the power generated at Tala by the amount of power inter- A) nally consumed in Bhutan, and further discounting for internal transmission and distribution losses in India, we convert this additional power supply brought to India from Tala into an amount actually available for con- sumption in India. Then, this can be used in the long-run equilibrium elas- ticity established by the model to arrive at the final figure for second- order employment effects. B) An increase in power supply from Bhutan’s Tala hydropower plant trans- mitted through PTL’s transmission lines created about 75,000 new formal jobs in India over the period 2006-2012, including about 4,600 new formal jobs in West Bengal. The impact on informal job creation would be larger than on formal jobs but our analysis focused only on jobs in the formal sec- tor due to data limitations. 3 b. Cost of using generators. Firms try to mini- increase in government revenues further enables the Bhu- mize production losses due to power outages by tanese government to spend more on social sectors like investing in backup generators. However, running health and education, thus improving quality of life and generators imposes an additional cost on firms employment prospects for the people of Bhutan. because using generators is more expensive than Using a simple National Accounting model we find that using power from the grid. This additional cost of as a result of Tala’s contribution to GDP over the period running generators, as opposed to using power 2007- 2010, on average, each year 973 students are en- from grid, could be saved and reinvested by firms rolled and 1,215 patient days are supported. That repre- more productively if there were no power outages. sents more than half of total beneficiaries that the elec- Such additional investment would create additional tricity sector reaches as a whole. For example in 2010, jobs if we assume a constant capital or asset-to- 933 students were reached by the government with the worker ratio. We use the additional incremental increased revenue from Tala as compared to 575 students cost of using power from a generator as compared that were reached as a result of increased revenue from to power from the grid, and ignore the capital cost the rest of the power sector. In fact, the total number of of installing power generators as sunk cost. students and patients reached more than doubles with By combining the impact on jobs from both a) and b), the the contribution of Tala (figures 3 and 4). Improved educa- study estimates the number of jobs created by improving tional access increases employment opportunities and in- the reliability of power. Primary data collected from the come levels. Similarly, healthy people are expected to get government in West Bengal and corroborated by the pri- and keep jobs more than unhealthy ones. Therefore, Tala vate sector shows the number of hours of outages per hydroelectric power plant not only contributes to Bhutan’s year decreased from 440 hours to 180 hours from 2005 GDP and growth, but also supports employment indirectly to 2010. Combining this information with the estimated by making it possible for the government to spend more jobs per hour of outage ratio from the estimation model on providing crucial health and education services. outlined above, the study finds that a total of 10,636 jobs were created in West Bengal by this reduction in power Conclusion outages. Of that total, about 1,600 jobs can be attributed This study shows the significant impact of an increase in to power brought by PTL from Tala. For Bihar, information power supply on jobs in an economy where power sup collected in the region indicated that there had not been a ply is a key constraint for firms. Improved power supply significant improvement in reliability.4 can play a significant role by helping developing coun tries Finding 5: By enabling the evacuation of power from increase their output and create jobs. While assessing the Tala, the PTL project helped bring in significant rev- employment effects of power projects, it is very impor- enue for the Bhutanese government, contributing tant for policy makers to focus more on the second-order almost 9% to its GDP, and enabling it to spend more growth effects (category 2 jobs) that are created as a result on education and health. of increased power supply that enables increased output The PTL transmission lines help evacuate power from Bhu- and hence jobs, rather than focusing on short term direct tan’s Tala hydropower plant to India. The Tala plant was jobs that are created through construction or even through built with funding from the Indian government on the un- O&M of the project. The significant impact of investment derstanding that power from it would largely be exported in the power sector on jobs, through second order growth to India. This project has a significant development impact effects, can be transformational for an economy. in Bhutan contributing almost 9% to Bhutan’s GDP. The Figure 3: Number of students enrolled Figure 4: Number of patient-days 2500 2000 2000 1500 1377 1016 970 933 1500 1000 1129 1140 1000 500 706 670 575 500 784 788 848 0 0 2008 2009 2010 2008 2009 2010 Tala Electricity (other) Tala Electricity (other) Source: Bhutan National Account Statistics (2010) & Bhutan Statistical Yearbook 2011 4 TYPES OF JOB EFFECTS METHODOLOGY USED Category 1 jobs (created during construction and O&M) 1 Direct 44,000 person years Input-Output model 2 Indirect 55,000 person years 3 Induced 144,000 person years Total 243,000 person years (about 9,700 jobs over 25 years) Category 2 jobs (created as an increase in power supply relieves a constraint for firms) 1 Impact of power 75,000 jobs in 6 years Econometric time series (VECM model) supply 2 Impact of reliability 1,600 jobs in 6 years in Step-by-step estimation model using Enterprise of power supply West Bengal Surveys and primary data from interviews. (subset of 1) Endnotes 1 This study was led by Namita Datta (CDI) with a team consisting of Govinda Timilsina (DECRG), Mahima Khanna (CSA) and Ferran Casadevall Massuet (CDI), under the overall guidance of Roland Michelitsch, (Manager, CDI). This case study is part of a broader study on ‘Private sector contributions to Job Creation’ led by CDI, IFC, which is aligned with World Development Report 2013 on Jobs. More information can be found at www.ifc.org/jobcreation. Acknowledgements: The study benefitted from insightful comments, suggestions and support from Clive Armstrong (CNGSF), Tom Davenport, (CSA), Shalabh Tandon (CSA), Soumya Banerjee (CSA), Priyanka Sood (CSA), Fan Zhang (ECSSD), Robert Bacon, Ulrich Bartsch (PREM), Luis Andres (SASSD), Alexis Diamond (CDI), Gabriela Armenta (CDI ), Anqing Shi (CDI) and Om Bhandari (CSA). A more detailed version of this study can be found at www.ifc.org/jobcreation. 2 While Input–Output models have been used in a large number of studies that estimate employment effects of infrastructure projects, the models are built on a number of assumptions. For example, the model assumes a constant production function, which excludes the possibility of efficiency and productivity improvements or other dynamic effects. Given that a new power transmission line might increase productivity, the job estimates provided by the input-output model will be conservative. 3 This step by step model is only meant to provide an approximate estimation of the magnitude of the importance of reliability of power - highlighted in various studies, in Enterprise Surveys, and also in our primary interviews as a very important issue for firms in India. The model involves a number of assumptions at various steps. In the absence of a more recent Enterprise Survey data set, we used the power –employment relationship in India in 2005 to give us an estimate. 4 These results are conservative estimates as it only considers the effects on formal employment. In 2004-05 the informal sector employment was six times that of the formal sector in India. The informal sector would also have benefitted from the improved power supply. Power theft for example, would have led to more production in the informal sector, even though it is a drain on the revenues of the power utility or the government. Contact: ifcjobstudy@ifc.org www.ifc.org/jobcreation 5