India: are the years of easy economic growth over?
- The current economic crisis in India is fuelling the debate about India’s structural economic growth
- Based on a fully endogenous economic growth model, we estimate India’s annual average economic potential at between 5% and 6% over the course of the decade
- Investment in human capital and innovation is necessary to foster Indian future economic growth
- In a scenario where the government implements (and intensifies) its education agenda, average annual economic growth could end up being 1.8ppts to 2.8ppts higher than under our baseline scenario (over the period 2023-2030)
- In a scenario where, on top of the education agenda, the government intensifies its innovation agenda and is able to secure a significant pick-up in FDI, average annual growth would end up a staggering 5ppts higher than in our baseline
Is the current crisis cyclical or structural?
Between 2014 and 2018, India outpaced its peers as fastest growing major economy (Figure 1). Recently, however, the economy has been running out of steam. After five consecutive quarters of lower economic growth and the very weak GDP print for fiscal q2 of 4.5%, it is safe to say the Indian economy is going through its most serious economic crisis since taper tantrum in 2013. The Reserve Bank of India (RBI) has lowered its growth forecast for fiscal 2019/20 to a meagre 5%. Moreover, Gita Gopinath, Chief Economist of the IMF, recently stated that they have been not making the right call on India’s growth trajectory, projecting 6.1% for fiscal 2019/20 and 7% for 2020/21 and that a downward revision will be published in January 2020.
The current slowdown is fuelling the debate about India’s structural economic growth potential. In this debate, former Chief Economic Advisor Arvind Subramanian has been particularly vocal: he claims that the current economic woes are structural in nature and due to unresolved balance sheet problems in four segments of the economy: infrastructural companies, the banking sector, non-banking financial companies and the real estate sector. Subramanian says that: “India is now trapped in an adverse interest-growth dynamic, in which risk aversion is leading to high interest rates, depressing growth, and generating more risk aversion.” In contrast, Arvind Panagariya, Economics Professor at Columbia University, argues that Subramanian has been wrong in the past about India’s gloomy economic growth perspective and concludes that there is no reason to panic. He says that India’s history of past economic ups and downs “should give us some pause as we assess the prospects of the Indian economy in the medium to long run.”
In this report we give our view on India’s structural economic growth potential. We first present a simple framework for analysing structural economic growth. Next, we break down our structural growth expectations into three components, which we discuss separately: the labor market, capital contribution and total factor productivity (TFP). In addition, we use our structural growth model to calculate the economic growth trajectory in two scenarios. In the first scenario, Indian policymakers implement and intensify an education/human capital agenda. In the second scenario, besides the human capital agenda, policymakers focus on stimulating innovation and foreign direct investment (FDI).
A framework for analysing structural economic growth
To shed more light on India’s structural economic growth we have developed a structural growth model. Structural growth is determined by labor inputs and labor productivity per hour (Figure 2). The contribution of labor to GDP growth is determined by structural employment, expressed as total hours worked. This is equal to the number of persons working multiplied by the amount of hours per worker. The number of people working, in turn, depends on the working-age population, participation rates and unemployment. In Appendix A we discuss the technicalities to derive all the underlying components and conduct forecasts for these components.
Labor productivity per hour is determined by capital per hour, which encompasses IT capital (software, computers, mobile phones, etc.) and non-IT capital (buildings, infrastructure, machines), and a residual factor: total factor productivity (TFP). TFP is arguably the purest indicator of technology, as it indicates how productive both capital and labor are in generating value added. Some important proven key factors of TFP are investment in knowledge capital, human capital, foreign knowledge spillovers and technological catching-up and hours worked and participation. In Appendix B, we discuss all technicalities on gauging and forecasting TFP.
The labor market
Due to data quality issues, assessing and forecasting India’s labor market is no easy task. In Appendix A we elaborate on all the underlying parameters and extrapolations. Ultimately, we arrive at the overall picture illustrated in Figure 3. Over the last five years, India’s structural employment growth (in hours) contributed roughly 1.4ppts on average to economic growth. For the upcoming years, we expect this contribution to end up a bit lower at 1.2ppts annually, on the back of ongoing downward-trending participation rates for different age groups. From 2022 onwards, we expect an upward-trending trajectory of the growth contribution of structural employment, mainly due to recovering participation rates in combination with lower unemployment levels. In the long run, the contribution of labor to GDP growth is expected to trend lower again due to a declining growth of the working-age population. While this population dividend easily exceeded 2ppts in the previous decade, it has already declined to less than 1.5ppts in 2019 and is expected to drop to less than 1ppts by the end of the decade.
Productivity: capital and TFP
Next, we turn to labor productivity per hour, which consists of the growth contribution of capital and total factor productivity growth.
To extrapolate the contribution of capital deepening, we use ARIMA modelling. This is a relatively agnostic method which assumes that the economic structure of the Indian economy does not change markedly. In our baseline (see Table A.4 in the Annex), we use a relatively optimistic assumption that capital deepening can contribute to GDP growth by between 2ppts and 2.5ppts on an annual basis.
We have developed a model in order to forecast TFP for India. The technicalities of this model are discussed in Appendix B and the assumptions for the baseline are discussed in Table A.4 in the Annex. We expect TFP growth to be relatively subdued in the next couple of years, partly caused an anticipated growth slowdown of inward FDI. Between 2000 and 2018, FDI as a percentage of GDP (i.e. the FDI ratio) grew from 3.6% to 14.6%. This was achieved in the slipstream of rapid globalisation of economic activities. Against the backdrop of a reversed globalising trend and India’s increasingly protectionist stance (e.g. the November withdrawal from the Regional Comprehensive Economic Partnership, or RCEP), it is unrealistic to assume that the FDI ratio will continue to grow at the same pace as in recent years. In combination with weaker investment in domestic innovation, we expect an annual contribution of TFP of 1.9ppts on average in the coming five years, rather than the 3.4ppts that we saw in the past five.
Decomposition of India’s structural growth potential
Combining the labor market forecasts with our productivity forecasts, we are able to calculate a baseline for India’s structural growth (Figure 4). We argue that India’s structural economic growth potential lies between 5% and 6%. Keep in mind that Figure 1 does not illustrate Rabobank’s official economic growth forecast for India, but shows our assessment of India’s growth potential, which is stripped of cyclical movements. This baseline is, of course, subject to much uncertainty. Labor market data is shaky and, as said, we adopt a relative positive assumption that capital deepening can contribute 2ppts to 2.5ppts annually. But if Subramanian is correct that the unresolved ‘Four Balance Sheet’ issues will continuously weigh on investment (see Subramanian and Felman, 2019), growth is be expected to end up somewhere near 5%.
Despite the uncertainty, we feel that our forecasts are more accurate than simply using a filter technique on GDP figures to gauge structural growth, an approach that we used before. Moreover, the fully endogenous model that is now in place enables us to calculate structural growth paths for India under certain scenarios.
Scenario analyses: how to foster structural growth?
Now that we have our fully endogenized structural growth model in place, we can run scenarios of the potential impact of different policy options on the Indian economy. Paul Romer, founding father of the endogenous growth theory (for which he received the Nobel Prize for Economics in 2018) has shown that growth of potential output is not an exogenous process, but requires investment in knowledge and human capital (Romer, 1990). Eichengreen et al. (2013) also argues that developing high-quality human capital and having a high share of high-tech products in exports reduces the probability of getting stuck in the so-called ‘middle income trap’.  Hence, Indian structural growth is not a completely autonomous process, but can be influenced by policy choices. Let’s start with the human capital agenda.
Scenario 1: fostering human capital
With its draft National Education Policy (NEP), the Modi government launched its plans to foster human capital and raise the quality of India’s education system. The plans cover a broad set of measures, e.g. the universalization of pre-primary education by 2025, increase of autonomy of higher education institutes, creation of a new independent State School Regulatory Authority (SSRA), the aim for 100% gross enrollment by 2030 and an increase of the engagement of the private sector. With these plans, NEP certainly addresses policy to tackle some thorny issues in India’s education system, but underwhelms in other areas. For instance, the draft NEP does not address ways to solve poor school and teacher accountability, prepare teachers how to teach and ignores the importance of learning English (for some critical comments: see this assessment). Moreover, in order to implement the plans, the draft NEP recommends raising public expenditure in education from the current 10% to 20%, over a ten year period. At the current juncture, however, it is unclear how these additional expenses are to be covered.
But let’s not be too critical beforehand. Let’s welcome the fact that fostering human capital and education has been firmly translated in some tangible policies. Figure 5 shows that there is much room for India to boost education and human capital vis-à-vis important emerging market peers. In our scenario, we assume that the policies under the NEP would generate human capital accumulation up to 2030, comparable to the results Thailand has been able to accomplish over the last twenty years. Under such a scenario, GDP growth would have an average annual boost of 1.8ppts over seven years, compared to our baseline (see Figure 6).
If policies turn out more favorable than expected or the Modi government decides to intensify education policy, human capital accumulation would resemble the trajectory of South Korea between 1975-1995. Consequently, average annual growth could lifted by 2.8ppts over seven years (Figure 6).
Scenario 2: fostering innovation and foreign direct investment
In a second scenario, we assume that the Indian government is stepping up its efforts by deepening its innovation agenda. PM Modi has mentioned the importance of innovation in many speeches and in 2016 the government launched the Atal Innovation Mission. One element of that program is to raise curiosity and innovation awareness among young Indians by setting up government-funded laboratories within schools. Other elements of the program are the establishment of incubator centers for high-tech start-ups and stimulation of mentoring networks. India definitely has improved its ranking in the Global Innovation Index, from 81st in 2015 to 52th in 2019 the GII concludes in 2019: “India will make a true impact on global innovation in the years to come.”
Still, there is room for improvement. For instance, the R&D intensity (R&D expenditure as % of GDP) shows a drop in 2015 and is lagging behind countries such as Brazil, China, Thailand and Turkey (Figure 7).
In our second scenario, we assume that India cranks up its innovation policy and that this would raise India’s share in total global knowledge stock (measured by the USPTO patent stock) from the current 1% to 8% in 2030 (see Figure 8). This would be no mean feat. Keep in mind that USPTO patents are only granted in case of technologies which are new to the US and, therefore, global market. For instance, China has invested billions of USD in its innovations system in order to crank up its share in global USPTO patent stock from 1.5% in 2014 to 3% in 2018. Ultimately, scenario 2 requires all elements within the entire innovation system to be upgraded on a markedly higher level. However, if India were to succeed, then, in combination with the ambitious human capital agenda from scenario 1, this would definitely result in a significant push of foreign direct investment towards India. There is indeed room for a significant improvement from an international perspective (Figure 9). We assume that India would be able to secure more FDI in the same fashion as Vietnam in the mid-90s until mid-00s.
Our calculations show that, in scenario 2, economic growth is pushed to more than 12% over the course of the decade (Figure 6). Over the period 2023-2030, average annual growth would end up being a staggering 5ppts higher than in our baseline scenario.
In this report we have shown that it will be very difficult for India to return to growth figures exceeding 7%. However, policymakers do have a choice to boost future growth by putting more emphasis on education and innovation. In the past, several emerging economies that went down that route managed to avoid the middle income trap, like South Korea, Japan, Ireland, Singapore, Hong Kong and Taiwan.
Acemoglu, D. and J.A. Robinson (2013). Why Nations Fail: The Origins of Power, Prosperity, and Poverty. Crown Business.
Ali, M., U. Cantner and I. Roy (2016). Knowledge spillovers through FDI and trade: the moderating role of quality-adjusted human capital. Journal of Evolutionary Economics, 26(4), 837-868.
Barro, R.J. and J.W. Lee (2013). A new data set of educational attainment in the world, 1950–2010. Journal of Development Economics, 104, 184-198.
Bassanini, A. and S. Scarpetta (2002). Does human capital matter for growth in OECD countries? A pooled mean-group approach. Economics Letters, 74(3), 399–407.
Belorgey, N., Lecat, R. and T.P. Maury (2006). Determinants of productivity per employee: An empirical estimation using panel data. Economics Letters, 91(2), 153–157.
Bitzer, J. and M. Kerekes (2008). Does foreign direct investment transfer technology across borders? New evidence. Economics Letters, 100(3), 355-358.
Bourlès, R. and G. Cette (2007). Trends in “structural” productivity levels in the major industrialized countries. Economics Letters, 95(1), 151–156.
Branstetter, L. (2006). Is foreign direct investment a channel of knowledge spillovers? Evidence from Japan’s FDI in the United States. Journal of International Economics, 68(2), 325-344.
Buccirossi, P., Ciari, L., Duso, T., Spagnolo, G. and C. Vitale (2013). Competition policy and productivity growth: An empirical assessment. Review of Economics and Statistics, 95(4), 1324-1336.
Cameron, G., Proudman, J. and S. Redding (2005). Technological convergence, R&D, trade and productivity growth. European Economic Review, 49(3), 775-807.
Coe, D.T., & Helpman, E. (1995). International R&D spillovers. European Economic Review, 39(5), 859–887.
Coe, D.T., Helpman, E. and A.W. Hoffmaister, A. W. (2009). International R&D spillovers and institutions. European Economic Review, 53(7), 723–741.
Cohen, W.M. and D.A. Levinthal (1989). Innovation and learning: the two faces of R&D. The Economic Journal, 99(397), 569–596.
Cohen, D. and M. Soto (2007). Growth and human capital: good data, good results. Journal of Economic Growth, 12(1), 51-76.
Eichengreen, B., D. Park and K. Shin (2013). Growth slowdowns redux: New evidence on the middle-income trap. National Bureau of Economic Research, no. w18673.
Engelbrecht, H.J. (1997). International R&D spillovers, human capital and productivity in OECD economies: An empirical investigation. European Economic Review, 41(8), 1479–1488.
Erken, H.P.G., P. Donselaar and A.R. Thurik (2016). Total factor productivity and the role of entrepreneurship. Journal of Technology Transfer, 43(6), 1493-1521.
Frantzen, D. (2000). R&D, human capital and international technology spillovers: A cross-country analysis. Scandinavian Journal of Economics, 102(1), 57–75.
Furman, J.L., M.E. Porter and S. Stern (2002). The determinants of national innovative capacity. Research Policy, 31(6), 899–933.
Griffith, R., Redding, S. and J. Van Reenen (2004). Mapping the two faces of R&D. Productivity growth in a panel of OECD industries. Review of Economics and Statistics, 86(4), 883–895.
Griliches, Z. (2000). R&D, education, and productivity. A retrospective. Cambridge (MA): Harvard University Press.
Grossman, G.M. and E. Helpman (1991). Innovation and growth in the global economy. Cambridge, MA: MIT Press.
Guellec, D. and B.P. Van Pottelsberghe de la Potterie (2004). From R&D to productivity growth: Do the institutional settings and the source of funds of R&D matter? Oxford Bulletin of Economics and Statistics, 66(3), 353–378.
Hejazi, W. and A.E. Safarian (1999). Trade, foreign direct investment, and R&D spillovers. Journal of International Business Studies, 30(3), 491-511.
Keller, W. (2010). International trade, foreign direct investment, and technology spillovers. In: Handbook of the Economics of Innovation, vol. 2, pp. 793-829.
Krammer, S.M. (2010). International R&D spillovers in emerging markets: The impact of trade and foreign direct investment. The Journal of International Trade & Economic Development, 19(4), 591-623.
Lucas, R.E. (1988). On the mechanisms of economic development. Journal of Monetary Economics, 22(1), 3–42.
Mankiw, N.G., D. Romer and D.N. Weil (1992). A contribution to the empirics of economic growth. Quarterly Journal of Economics, 107(2), 407–437.
Park, W.G. (1995). International R&D spillovers and OECD economic growth. Economic Inquiry, 33(4), 571-591.
Porter, M.E. and S. Stern (2000). Measuring the ‘‘ideas’’ production function: Evidence from international patent output. Cambridge, MA: National Bureau of Economic Research, NBER Working Paper, no. 7891.
Psacharopoulos, G. (1994). Returns to investment in education: A global update. World development, 22(9), 1325-1343.
Romer, P.M. (1990). Endogenous technological change. Journal of Political Economy, 98(5), S71–S102.
Subramanian, A. and J. Felman (2019). India’s great slowdown: What happened? What’s the way out? Centre of International Development, Working Paper No. 370, Harvard University, December 2019
Zhu, L., and B.N. Jeon (2007). International R&D spillovers: Trade, FDI, and information technology as spillover channels. Review of International Economics, 15(5), 955-976.
Appendix A: A framework for projecting Indian labor market developments
The contribution of labor to GDP growth in the long term is determined by structural employment, expressed in total hours worked. This is equal to the number of persons engaged times the amount of hours per worker. The easiest way to forecast the labor market for India would be to extrapolate both elements. However, this does not do justice to underlying labor market dynamics, as structural employment also consists of separate factors, such as growth of the working-age population, participation rates and the unemployment rate. The most accurate way is to make forecasts of these underlying components as well. In formulae:
where L is structural employment in hours, N is the number of persons engaged, HW is the number of hours per worker and t is year. Next, N can be expressed as :
where P15-65 is the population aged 15-65, PT15-65 is the participation rate of the population aged 15-65 and U is the number of unemployed in that age group. We chose to focus on the 15-65 age group, as data on participation of people younger than 15 or older than 65 are unreliable.
Both equations (A.1) and (A.2) can easily be transformed in equations expressed in differences. Ultimately, we translate our extrapolation to the total number of people engaged. There is, of course, the problem of the informal economy, which is substantial in India. However, we cannot take their contribution into account.
Forecasting the components in equations (1) and (2) is by no means easy. In fact, combining these parameters to break down the labor market into one solid decomposition is already an arduous task, as data from different sources simply do not add up. Therefore, we have worked our way backwards to generate a solid decomposition, starting with the parameter we are the least uncertain about: the growth of the working-age population. The quality of participation- and unemployment data is much more questionable. For the sake of this exercise, we choose to use participation data and simply generate our own unemployment series in order to fit in with the overall structural employment data in the national accounts (i.e. Penn World Tables data). Table A.1 shows all data that we use to conduct our analyses.
Our starting point is the growth of the working-age population. We use projections by the UN for five-year age cohorts (see Figure A.1). These cohorts are related to five-years age participation rates to derive the labor force, i.e. people active on the labor market between 15 and 64 years.
Next, we apply Hodrick-Prescott (HP) filters with a low lambda (λ = 10) on five-year age participation rates and extrapolate these filtered series by using AutoRegressive Integrated Moving Average (ARIMA) models. As the participation data by the OECD shows rather abrupt changes, we choose to smooth the data to fit the pattern of overall structural employment in India.
Finally, we backward-engineer unemployment levels based on the data for the working-age population, participation rates, structural employment and hours per worker. Our unemployment proxy expressed in differences is pretty much in line with the differenced job seekers series, based on NCS data (leaving out 2015 as an outlier).
Appendix B: A framework for analysing Indian TFP
Determinants of TFP
There is vast academic literature on determinants of TFP. Below we give a brief description of these determinants.
Human capital is defined as quality improvement of labor due to education and training. It is one of the most important driving forces behind technological progress and TFP, not only from a theoretical perspective (see Lucas, 1988; Romer, 1990), but also from an empirical one (e.g. Engelbrecht, 1997; Bassanini and Scarpetta, 2002; Krammer, 2008; Erken, Donselaar and Thurik, 2016). Eichengreen et al. (2013) also finds that developing high-quality human capital reduces the probability of a slowdown in growth.
The average years of (tertiary) education is usually used as an indicator to measure the amount of human capital in a country (see Barro and Lee, 2013). Cohen and Soto (2007) claim to produce data of a higher quality than Barro and Lee, as they use information from surveys based on uniform classification systems of education over time, and an intensified use of information by age groups. The Penn World Tables 9.1, from which we take the data on human capital, adopts several criteria in order to assess which human capital series to use for which county (see here) and we directly follow their approach. Moreover, when adopting the Barro and Lee data series, the Penn World Table also takes into account the return of education based on Mincer equation estimates (see Psacharopoulos, 1994).
Knowledge and innovation
Investment in knowledge and innovation is an important factor of TFP as well. Many studies use investment in Research & Development (R&D) as a proxy for private and public knowledge development (e.g. Coe et al., 2009; Coe and Helpman, 1995; Griffith et al., 2004; Guellec and Van Pottelsberghe de la Potterie, 2004). The problem with R&D data for India is that time series are available for only a limited amount of years. In order to circumvent this problem, we use an alternative proxy: patent capital. The idea to use patent capital as a proxy for innovative capacity is taken from Furman et al. (2002) and Porter and Stern (2000).
We define the cumulated knowledge stock on the number of patents granted by the US Patent and Trade Office (USPTO). To calculate the stock of patent capital, we use the perpetual inventory method (PIM):
In contrast to Porter and Stern (2000), we use a depreciation rate(δ) of 15% to take into account the obsolescence of knowledge. This depreciation rate is often used to calculate R&D capital, based on Griliches (2000, p. 54), who refers to this percentage as the “‘conventional’ 15 percent figure for the depreciation of R&D-capital”.
A complication to calculate formula (A.3) is that we do not know the initial capital stock. If we assume a constant growth rate of patent flows, we can derive the initial patent stock (S0) which can be calculated by:
In A.4, RD0 is the first observation of the patent flows and in A.5, g is the annual growth rate of the number of patents.
Foreign knowledge spillover
Next to human capital and domestic knowledge capital, knowledge developed abroad is an important factor for domestic productivity development (e.g. Coe and Helpman, 1995; Cohen and Levinthal, 1989). The idea is that countries with a relatively low level of technological development can benefit from knowledge developed elsewhere by using it in their own products or production processes. Through a process known as technological catching-up, low technological countries can narrow the gap in productivity with the technologically more advanced countries. We calculated the foreign knowledge stock by using equations (A.4) and (A.5) for USPTO annual granted patents for 150 countries.
But there is no such thing as a free lunch and two conditions should be met in order to benefit from foreign knowledge development. First, there needs to be a transmission channel, whether it be trade (Coe and Helpman, 1995; Grossman and Helpman, 1991, Zhu and Jeon, 2016), foreign direct investment (Branstetter, 2006; Bitzer and Kerekes, 2008) or both (Ali, Cantner and Roy, 2016; Hezaji and Safarian, 1999; Keller, 2010). In the current study we use the inward FDI stock as a percentage of GDP as a conduit of foreign knowledge spillovers.
The second condition which should be met is that domestic firms should be capable of using foreign knowledge by conducting knowledge activities themselves. This is in line with Cohen and Levinthal (1989), who claim that countries need an R&D base of their own in order to benefit from R&D conducted abroad. In the literature, this is often referred to as absorptive capacity.
Labor input generates adverse TFP effects (Belorgey et al., 2006; Bourlès and Cette, 2007; Erken, Donselaar and Thurik, 2016). High labor participation is often characterized by increased deployment of less-productive labor, which lowers labor productivity. Working fewer hours may have a positive impact on productivity if less fatigue occurs among workers or if employees work harder during the shorter number of active hours. Another factor that affects total factor productivity is business cycle effects, as labor and capital endowments are not immediately adjusted to business cycle volatility, which makes TFP susceptible to fluctuations of the business cycle. Finally, we argue that institutional development might generate positive TFP effects (Coe et al., 2009 and Acemoglu and Robertson, 2013). High-quality institutions provide the environment for economic convergence and development, and allow countries to achieve long-term income convergence. There is also substantial empirical evidence in the literature that the institutional quality has a beneficial effect on the development of productivity and growth. For instance, better regulation lowers administrative burdens and X-inefficiencies. In the current study we use data from the Fraser Institute to assess institutional quality. Finally, there are many more factors that determine TFP, such as competition and entrepreneurship, but due to limited data availability, we restrict our analysis to the factors discussed above.
A model for Indian TFP
The starting point for a model for Indian TFP is the human capital-augmented Solow model introduced by Mankiw et al. (1992):
where Y denotes gross domestic product, K and L denote (physical) capital input and labor, respectively. Furthermore, H represents human capital and A represents the level of (labor-augmented) technological change. Under the neoclassical conditions of perfect competition in product markets and constant returns to scale in the production factors of capital and labor, the marginal products of capital and labor are equal to the return on capital and the wage rate, respectively. It can be derived that the output elasticities of capital and labor are equal to the shares of capital income and labor income in total factor income. If we express (1) in first differences and take into account a varying output elasticity of capital, we arrive at the following equation:
where c is a constant term, wK is the share of capital income in total factor income (or stated differently, the share of capital income in the gross domestic product), D denotes mutation in first differences, t is a time index (i.e. year), log represents the natural log and e is an idiosyncratic error term. In addition, c1 picks up the effect of growth of domestic knowledge capital (S) on TFP growth. The coefficient c2 estimates the effect of growth of the foreign knowledge capital (S f) interacted with the FDI ratio (i.e. the inward FDI stock (F) as a % of nominal GDP (Y n)) and the growth of the domestic knowledge stock (S d). These two conditions are important, because a transmission channel (FDI) and absorptive capacity are necessary to benefit from foreign knowledge spillovers. The coefficient c3 estimates the impact of changes in human capital. Coefficients c4 and c5 pick up the impact of participation and hours worked per workers, respectively. Coefficient c6 is used to isolate the effect of the business cycle, which has a lead of one year, as business cycle shocks do not directly feed into TFP. As the impact of the global financial crisis was an anomaly which is not picked up fully by the regular business cycle variable, we use a dummy variable (c7) for 2009 to eliminate the bias this one-off event has on our estimations. Finally, c7 and c8 are institutional variables from the Fraser Institute. We experimented with cross corrections between all institutional variables (see Appendix C) with TFP and the two institutional factors that stand out are sound money (Im) and regulation (Ir). These two indices are included in the model.
Table A.2 shows the variables that we use in our US TFP model, as well as their data sources. We use simple OLS to estimate the model and use the various information criterions to obtain the optimal lag structure.
Our approach to model the impact of knowledge, human capital and foreign knowledge development on productivity within a dynamic setting fits a broad strand of literature (see Park, 1995; Frantzen, 2000; Griffith et al., 2004; Cameron et al., 2005; Buccirossi et al., 2013).
The first estimate in column (1) of Table A.3 consists of domestic patent capital, foreign R&D capital interacted with the FDI ratio and domestic patent capital, human capital and the GFC dummy. All variables, except the domestic patent stock, show a statistically significant impact on TFP growth as well as the expected sign. The estimation shows that domestic knowledge stock obviously has a much weaker correlation with TFP than foreign knowledge. This does not come as a complete surprise, as many emerging economies rely heavily on foreign technologies and are not capable of pushing the technological frontier by means of innovative activities at the home base. This implies that domestic knowledge activities are especially important to benefit from foreign knowledge (i.e. absorptive capacity) and incorporate technologies developed abroad in domestic production processes. The magnitude of coefficient c2 is quite substantial, but this is due to the fact that we have interacted the mutation of the knowledge stock with the mutation of the domestic knowledge stock as well as the foreign direct investment ratio. The human capital variable has a coefficient (c3) of 1.66, which means that an increase of the human capital index by 1% yields higher Indian TFP growth by 1.66ppts.
In the second estimation in column (2), we scratch out the insignificant domestic patent capital variable (c2) and add our two labor market variables: labor participation (c4) and hours worked per worker (c5). As expected, these variables both have a significant negative impact on TFP growth. The other variables are statistically significant as well and stable.
In column (3), we add the business cycle variable, which leads to a significant improvement of our model fit. In the final estimation in column (4), we add the institutional variables. Both show a statistically significant positive impact on TFP, but adding these variables does not bring much improvement in our model's explanatory power. The model in column (4) is the one we use to produce our forecasts for Indian TFP growth in the baseline, as well as in the two policy scenarios. The fit of the model is shown in Figure A.4.
Appendix C: Areas, Components, and Sub-components of the Economic Freedom of the World index (Fraser Institute)
1. Size of Government
A. Government consumption
B. Transfers and subsidies
C. Government enterprises and investment
D. Top marginal tax rate
(i) Top marginal income tax rate
(ii) Top marginal income and payroll tax rate
2. Legal System and Property Rights
A. Judicial independence
B. Impartial courts
C. Protection of property rights
D. Military interference in rule of law and politics
E. Integrity of the legal system
F. Legal enforcement of contracts
G. Regulatory costs of the sale of real property
H. Reliability of police
I. Business costs of crime
3. Sound Money
A. Money growth
B. Standard deviation of inflation
C. Inflation: most recent year
D. Freedom to own foreign currency bank accounts
4. Freedom to Trade Internationally
(i) Revenue from trade taxes (% of trade sector)
(ii) Mean tariff rate
(iii) Standard deviation of tariff rates
B. Regulatory trade barriers
(i) Non-tariff trade barriers
(ii) Compliance costs of importing and exporting
C. Black-market exchange rates
D. Controls of the movement of capital and people
(i) Foreign ownership / investment restrictions
(ii) Capital controls
(iii) Freedom of foreigners to visit
A. Credit market regulations
(i) Ownership of banks
(ii) Private sector credit
(iii) Interest rate controls / negative real interest rates
B. Labor market regulations
(i) Hiring regulations and minimum wage
(ii) Hiring and firing regulations
(iii) Centralized collective bargaining
(iv) Hours regulations
(v) Mandated cost of worker dismissal
 The middle income trap refers to the phenomenon that GDP per capita growth falls significantly after a certain threshold is reached, as wages start to rise and an emerging economy loses its comparative cost advantage (see Eichengreen et al., 2013). This threshold lies between USD 12,000 and USD 18,000 per capita, based on 2015 purchasing power parities.
 The term X-inefficiencies refers to slack in the production process and higher production costs than necessary, which are the result of lack of competitive pressure in the market.