Abstract: Can an emerging economy like India still choose its own interest rates? This paper approaches that question first empirically and then politically. Using monthly data on India’s ten-year government bond yield from 2000 to 2026, the paper regresses its monthly changes on changes in the United States ten-year yield and the global oil price, and then models the unexplained residual against India’s own policy rate and inflation. The US-yield coefficient is about 0.30 and statistically significant. The external variables together explain only about an eighth of the monthly variation, and India’s domestic policy rate moves its yield about as strongly as the US rate does. This pattern is termed constrained autonomy: India neither sets its long rates freely nor simply imports them. Looking across time refines the picture and rejects the initial hypothesis that the external influence would have increased as India’s market opened. It has not. The sensitivity was highest before the 2008 crisis and is lowest in the most recent years, and the India–US yield gap has narrowed considerably from its 2013 peak, a pattern consistent with India using improved monetary credibility to recover some policy room. The paper then argues that this statistical constraint reflects a political one as well. Drawing on the international relations of monetary power, and on the unusual 2026 episode in which the United States intervened directly in the Japanese yen, it suggests that the constraint is not only an emergent property of markets but something a dominant state can actively exercise, and that an emerging market such as India sits well below an advanced economy like Japan in that hierarchy. The finding places India closer to Rey’s “dilemma” than to a clean policy trilemma.

1 Introduction

On 22 May 2013 the chairman of the US Federal Reserve, Ben Bernanke, told a congressional committee that the Fed might soon slow the pace of its bond purchases (the episode widely known as the “taper tantrum”). He announced no policy change. He had only raised the possibility of one. The reaction in India was immediate and severe. Foreign investors pulled money out of Indian bonds and equities, the ten-year government yield jumped, and over the following months the rupee lost close to a fifth of its value against the dollar. India was soon grouped with Brazil, Indonesia, South Africa and Turkey as one of the “Fragile Five”, economies judged dangerously dependent on foreign capital¹. Nothing about India’s own fiscal or monetary policy had changed in those weeks. What changed was a sentence spoken in Washington.

That episode is the puzzle this paper is built around. If a remark by an official whom no Indian citizen elected can move the price at which the Indian government borrows, in what sense does India control its own monetary policy? Economists usually ask this as a technical question about interest-rate “spillovers”. This paper treats it as a political one, because the answer bears on something political scientists care about directly: the sovereignty of a democratic state over its own economy.

The paper makes two arguments. The first is empirical. India occupies a middle ground termed here constrained autonomy: it does not set its long-term rates freely, but it does not simply import them either. A sizeable share of the monthly movement tracks external conditions, while the rest is governed by domestic factors; and, as shown below, India appears to have widened that domestic room over the past decade rather than lost it.

The second part is interpretive, and makes an argument on domestic monetary autonomy citing recent developments in the international market. The constraint is not only an impersonal feature of global markets; it can be actively exercised by the state at the centre of the monetary system. The United States 2026 intervention in the Japanese yen makes this visible, and it implies that an emerging market like India, being lower in the hierarchy than a close ally such as Japan, can be pressed harder than what the historical sensitivity calculated using the data suggests.

The paper proceeds as follows. Section 2 sets out the relevant scholarship, which spans political economy, open-economy macroeconomics and the study of emerging markets. Section 3 states two testable hypotheses. Section 4 is the empirical core: the data and method, the main regression, its behaviour across time, and the domestic second stage. Section 5 reads the same channel qualitatively for emerging markets as a class. Section 6 turns to the recent Japan episode as a display of great-power monetary politics, and Section 7 connects that politics back to India. Section 8 discusses and concludes.

2 Literature Review

Three bodies of work bear on this paper, and the space between them is where its contribution lies.

The second body of work is open-economy macroeconomics. Its starting point is the Mundell–Fleming “trilemma”: with free capital movement a country can have an independent monetary policy or a fixed exchange rate, but not both (Mundell, 1963; Fleming, 1962). Obstfeld, Shambaugh and Taylor confirmed that the trade-off has genuinely bound countries, and that the exchange-rate regime shapes how much independence survives (2005). Hélène Rey then unsettled the consensus. She argued that a “global financial cycle” in capital flows and asset prices, driven substantially by US policy and global risk appetite, transmits worldwide regardless of the exchange-rate regime, so that for economies open to capital the choice is not a trilemma but a dilemma: independent monetary policy is possible only if the capital account is managed (2015). Miranda-Agrippino and Rey traced the mechanism through global banks and the pricing of risk (2020). That disagreement is exactly what a single country’s data can speak to.

The third body of work is specific to emerging markets. Calvo and Reinhart’s “fear of floating” showed that emerging economies which claim to float in fact intervene heavily, precisely because they fear the capital-flow consequences the dilemma describes (2002). Eichengreen and Gupta studied the taper tantrum directly and found that the countries with the largest and most liquid markets, India among them, were hit hardest, because they were the easiest to sell out of (2015).

The gap is this. The political-economy literature argues about sovereignty largely in words, while the macroeconomics literature measures spillovers precisely but rarely asks what they mean for democratic control. This paper runs one transparent empirical exercise on a single country, reads the result deliberately as evidence about sovereignty, and then uses a recent political episode to show that the constraint the regression detects can be a deliberate instrument rather than only an accident of markets.

3 Theory and Hypotheses

The debate reduces to two claims a regression can test.

H1 (the external constraint exists). There is a positive and significant co-movement between changes in India’s ten-year yield and the US ten-year yield, after controlling for the global oil price. If Rey is right, the pass-through should be real and detectable, and should not merely be a spillover of global commodity shocks.

H2 (domestic room survives). The domestic variables, such as the policy rate and inflation are expected to still explain a meaningful part of the movement that the external variables cannot explain. If they do, the constraint is only partial and India still has some hold on its own long rates.

If there is H1 but no H2, then sovereignty is only on paper; if there is H2 but no H1, then India sets its long rates largely on its own. The results support neither extreme, hence the term constrained autonomy; the strength of the external pull provides a measure of how much room there is in domestic policy.

4 Empirical Analysis

4.1 Data and method

The data are monthly, running from January 2000 to August 2026, which leaves 319 observations after differencing the series. The dependent variable is the change in India’s ten-year government bond yield. On the external side, the explanatory variables are the US ten-year Treasury yield and the dollar price of Brent crude, with the rupee–dollar exchange rate added as a further control. The second stage comprises two domestic variables, the RBI repo rate and consumer-price inflation. A supplementary check is provided by year-on-year GDP growth, which is available only quarterly. Series are from Reserve Bank of India, US Treasury, Federal Reserve, Ministry of Statistics, Energy Information Administration.

One choice about how the variables enter the regression is worth explaining plainly, as it decides what the coefficients mean. Take the oil price. A one-dollar rise from $60 to $61 is a move of about 1.7 per cent, while the same one-dollar rise from $90 to $91 is only about 1.1 per cent. Measured in raw dollars, a regression treats these as the same shock, a “plus one” in each case, even though the first value is a larger event for the economy than the second one. Taking the logarithm of the price fixes this. The monthly change in a logged series is, to a close approximation, its percentage change, so that $60?$61 and $90?$91 enter as different-sized moves. It also makes the interpretation easier for the coefficient. Because the yield enters in percentage points and the price in logs, the coefficient is a semi-elasticity: it reports how many percentage points India’s yield moves in response to a proportional move in the price (a 10 per cent rise in oil corresponds to roughly one-tenth of the coefficient), rather than a response to a one-dollar move whose meaning shifts with the price level. For both reasons, the paper uses the log change, or “log return”, for the two price series, oil and the rupee.

Yields, the policy rate and inflation are deliberately treated differently. They are already in percentage points so the natural unit of change is just the difference as a move from 7.0 to 7.2 per cent is a twenty basis point change, which is clearly interpretable as it stands. Taking the logarithm of a percentage would turn that transparent quantity into an opaque one.

A second requirement is about the behaviour of the series over time. An ordinary regression assumes that all the variables are stationary, meaning that their average level and their variability are roughly constant rather than drifting over time. This is important for the statistical validity, because two series that both trend up over time will look to be highly correlated in a regression, even in the absence of any relationship between them, just because they are both trending up, and the regression will then give a very good fit and significant coefficients that may be totally spurious.

The usual check is to test each series for a “unit root”, the statistical tool that helps to identify whether a series is stationary or not, using the Augmented Dickey–Fuller (ADF) test. On the levels, the Augmented Dickey–Fuller test comes back mixed: India’s yield and the repo rate marginally clear the bar, while the US yield, inflation, oil, and the rupee do not.

To avoid relying on these borderline results, the analysis uses only month-to-month changes, first differences for the yields and then log differences for the price series. Each of these rejects the unit root decisively (p < 0.001), which yields stationary variables and regression coefficients that can be meaningfully interpreted.

Table 1 reports summary statistics for the levels and for the transformed series actually used in regression.

Figure 1 plots the raw levels of India’s ten-year yield against America’s. Both fall over the period, India’s by more, so the gap between them, the premium investors want for holding Indian debt over US debt, ends the period well below its 2013 peak.

The estimation strategy has two stages. Stage one asks how much of India’s yield movement is imported:

with a variant adding the rupee log return r^fx_t. The coefficients are the sensitivities H1

Figure 1: India and US ten-year government bond yields, 2000–2026.

concerns and the residual eˆ_t eˆt is the part of India’s yield movement external conditions do not explain. Stage two models that residual on the domestic instruments:

eˆ_t = ?₀ + ?₁ ?repo_t + ?₂ ?CPI_t + u_t.      (2)

Stage two is intentionally restricted to this simple regression, along with a set of correlations. That domestic instruments move the residual is the only claim required here, and equation (2) tests it directly and legibly. All standard errors are Newey–West (HAC, six lags) to account for autocorrelation and heteroscedasticity in monthly financial data.

4.2 Main results

Table 2 reports the first stage in three specifications: the baseline over 2004–2026 (the window for which every series is populated), the same with the rupee added, and the baseline extended back to 2000.

The US-rate coefficient is the headline. At roughly 0.30, and significant at the five per cent level, it means that a one-point rise in the US ten-year yield over a month is associated with roughly 0.3 of a point on India’s, with oil held constant. Adding the rupee barely touches it (still 0.30), and over the longer 2000–2026 window it stays significant, though it slips to 0.25. So H1 holds, and it is robust across the alternative specifications: some clearly

Newey–West t-statistics in parentheses. *p < 0.1; **p < 0.05; ***p < 0.01.

non-zero share of the monthly move in Indian borrowing costs follows the US market. The mechanism is not mysterious. When US yields rise, investors can get more in the safest market there is, and Indian debt has to pay up to hold onto them. Of particular interest here is the political side of the same fact: part of the price the Indian government pays to borrow is set by conditions it does not choose, which leaves it exposed to swings that originate abroad. Figure 2 plots the relationship, one point per month.

Figure 2: Monthly co-movement of Indian and US ten-year yields, 2004–2026. The line is the fitted regression slope, ß ˜ 0.31.

The fit is worth a second look. The external variables are jointly significant, so they genuinely belong in the model, but they explain only about thirteen per cent of the monthly variation, which leaves roughly seven-eighths unaccounted for. That remainder is not noise to apologise for. It is the raw material of the second stage, and its size is the first hint that H2 will hold.

4.3 Has the sensitivity risen over time?

The initial expectation was the opposite of what the data show. As India’s bond market opened and deepened over these two decades, its exposure to US rates might have been expected to grow, producing a coefficient that climbed over time. No such pattern appears. Table 3 cuts the sample into five roughly equal windows and re-estimates how sensitively India’s yield tracks the US yield in each.

Bivariate regressions of ?y^IN on ?y^US, Newey–West t-statistics.

The coefficient peaks in the pre-crisis boom of 2004–2008, at about 0.77, then falls away through the 2010s, and by the most recent window, 2024–2026, it is down to roughly 0.14, statistically indistinguishable from zero. Figure 3 traces the same decline as a rolling thirty-six-month beta, a continuous line instead of bins. So India’s month-to-month sensitivity to US rates has not climbed over these two decades; if anything it has eased. The initial prediction is therefore rejected, which turns out to be the more interesting finding. The decline is not monotonic, however (the 2019–2023 estimate is the most precisely estimated of the five), and differences across windows have not been formally tested for structural breaks.

The finding admits two readings, and the data alone cannot settle between them. The optimistic reading is that the decline is real. Over these two decades India built a monetary framework markets take more seriously, above all by moving to flexible inflation targeting in 2015–2016 and standing up a Monetary Policy Committee, and a central bank trusted to keep inflation

Figure 3: Rolling 36-month sensitivity of India’s yield to the US yield. The tether does not tighten over time.

down can hold its own rates a little freer of the global cycle. The shrinking India–US gap fits that: it falls from an average near 5.9 points at the 2013 peak to about 2.5 in 2024–2026 (Table 3; Figure 4), so Indian borrowing costs have converged toward the world’s, which is roughly what growing credibility looks like. A more cautious reading is also possible. The recent window is only thirty-two months, the coefficient is loosely estimated, and the shocks that fill it, a higher-for-longer Fed and the odd oil spike, need not resemble the average stress the coefficient is supposed to summarise. On balance, the first reading is favoured, mainly because the narrowing spread points the same way, though the evidence is not conclusive.

Figure 4: The India–US ten-year yield gap, 2000–2026.

4.4 The domestic second stage

If the external variables leave seven-eighths of the movement unexplained, does India’s own policy govern that remainder? Table 4 regresses the Stage 1 residual on the change in the repo rate and the change in inflation.

Newey–West t-statistics. Dependent variable is the residual from Table 2, column (1).

Both domestic instruments move the residual, and both in the expected direction: when the RBI raises its policy rate, or when inflation rises, the part of India’s long yield that the global cycle does not explain rises as well. The clearest way to see H2 is in the raw correlations of the monthly change in India’s yield. It correlates with the change in the repo rate at 0.30, with the change in the US yield at 0.27, with the oil return at 0.28, with inflation at 0.12, and with the rupee at only 0.06. The first of these is the striking one: India’s own policy rate co-moves with its long yield at least as strongly as the US rate does. A quarterly check points the same way, with the average residual correlating with year-on-year GDP growth at about 0.17, the mild positive link a faster economy and its inflationary pressure would predict. These results support H2: domestic policy and inflation explain a statistically significant, though modest (R² ˜ 0.07), share of the movement that the global cycle leaves unexplained. One caveat applies: the repo rate is not fully exogenous, since the RBI itself responds to external shocks (as in 2013), so part of this domestic channel may in turn transmit global conditions.

5 From India to Emerging Markets: A Qualitative Reading

The regression is about India, but the channel it detects is general. Setting out how US and global yields reach emerging markets as a class is what makes India a general case rather than an exception.

The transmission, often called the “financial-tightening channel”, runs in four steps. When US rates rise, or when global investors turn cautious and reach for safety (also described as risk-off episodes) the return on dollar assets rises relative to riskier emerging-market assets. As a result, investors pull out money from emerging-market bonds and equities and move the proceeds into dollars, so money flows out from them. That outflow leads to large-scale sale of the local currency, which depreciates as a result and since emerging economies import much of their energy (oil) and other goods, many of which are priced in dollars, the weaker currency raises the import bill and thus directly feeds into inflation. The local central bank is then bound to respond. To defend the currency and contain imported inflation it has to tighten (increase the rates), or spend reserves, or both, even when its own economy would have ideally preferred an easier policy. The country ends up importing a monetary stance that was calibrated for conditions abroad. Analysts often summarise emerging-market fortunes through three external levers, the Fed’s rate cycle, which anchors global risk-taking capacity, the dollar, which is the channel of transmission, and the oil price, and all three are set well beyond any single emerging economy’s control. This is Rey’s global financial cycle viewed from the receiving end (Rey, 2015; Miranda-Agrippino and Rey, 2020).

Politically, two aspects of the channel matter more than its mechanics. It is asymmetric by design: capital floods in when the centre (the United States, as the dominant monetary power) is loose and rushes out when it tightens, and the economy on the receiving end does not control the timing, so its financial conditions are to a large degree a passive readout of decisions taken elsewhere. It is also counter-cyclical in a way, because the moments when an economy would most want to ease, a downturn or a crisis, are often the same moments of global risk-aversion when capital flees to the dollar and the constraint binds hardest.

6 Recent Developments: Great-Power Monetary Politics and the Japan Case

The account so far makes the channel sound impersonal, as though rates and flows simply move on their own. The events of 2026 are a reminder that they can also be steered. Over that year the United States took an unusually hands-on role in the Japanese yen. In late July, American authorities joined Japanese officials in buying yen, the first US purchases of the currency in roughly three decades, and the dollar fell from near 164 yen before the joint operation to around 153 yen after it.² The US Treasury Secretary, Scott Bessent, had also pressed the Bank of Japan to raise its interest rates and had coordinated the intervention directly with Japan’s finance minister. Defending this at a public event in Texas, Bessent said that “I am the house now,” explaining that his coordination with Tokyo gave him something close to inside knowledge of the Bank of Japan’s next moves, and inviting traders to bet against him.³

Read as economics, this is currency management. Read as politics, it is a demonstration of the structural power the theory describes. Three features stand out. First, a great power was setting the terms of another government’s monetary policy, not through a treaty or an institution but through market operations and open pressure on a nominally independent central bank. Second, the intervention was self-serving in a way that revealed the hierarchy. By buying yen with euros rather than forcing Japan to defend its own currency, Washington reportedly spared Tokyo from having to sell its large holdings of US Treasuries, a sale that would have pushed up American borrowing costs at an awkward moment.4 The favour to Japan was also a favour to the United States, and only the country at the centre of the network can arrange matters so that helping others helps itself, which is close to the essence of Farrell and Newman’s weaponised interdependence (2019). Third, the language itself is revealing. “I am the house” is how someone talks who knows the game is not played on level terms, and Kirshner’s account of monetary statecraft (1995) would recognise it at once.

7 From Japan to the Periphery: What It Means for India

The value of the Japan case for this paper is comparative, and it is a sobering comparison for an emerging economy. Japan is a wealthy, advanced, creditor nation and a close US ally, the second-largest holder of US government debt, with a currency held in reserves around the world. Even Japan found its currency and its central bank drawn into the orbit of American monetary decisions in 2026. India sits much further down the same hierarchy: smaller, more dependent on foreign capital, with a currency no one holds in reserve and a bond market that foreigners can exit far more easily than they can exit the yen. It serves as a reminder that if structural power can reach Tokyo, it can reach Mumbai with more force still.

Both the empirical finding and the political reading need to be read together with care because at a first glance they look to be going against each other. The regressions say India’s average monthly sensitivity to US rates is modest and more recently has been following a declining trend. The recent episode in Japan highlights that the dominant state’s latent power over the periphery is large and can be exercised at will. Both can be true because they describe different things. The empirical beta measures normal times co-movement while the Japan episode is a tail event that nonetheless calls for recalibration of what the centre can do deliberately, in a crisis or by design when it chooses to act rather than simply form the backdrop. A country can face a small average pull and still sit under a high political ceiling. India’s everyday autonomy is real; however, its limits can be severely tested during times of global turmoil.

What can a country in India’s position actually do? Past episodes, in 2013 and again in 2026, show the usual tools. It can sell foreign-exchange reserves to slow the rupee’s fall, and it can tighten or loosen the capital account to smooth the flows that carry the cycle in (a recent example is the RBI’s 2026 measure to attract FCNR deposit inflows). Neither tool is a remedy. Reserves run out, and spending them openly can further withdraw the support and add a drag on the currency. Capital controls can dampen the cycle at the same time they also scare off some of the investment that openness was supposed to bring in.

These tools do not restore sovereignty. What they really do is give the government a choice about where the pain lands and whether the shock is absorbed by the currency, reserves, or domestic rates. For an economy like India’s, that choice about where the shock lands is most of what monetary sovereignty means in practice.

8 Discussion and Conclusion

Putting it all together, the evidence places India at a somewhat specific and vulnerable spot. The effects of global monetary moves flow in indirectly but with less force than they once did as India established its monetary credibility and independence, a hopeful reading that the empirical analysis above supports. But the recent episode in Japan is a stark reminder that the progress is still fragile and the relationship can tighten again under the shadows of global monetary hegemony. A deliberate intervention can tighten the constraint on an economy like India’s far beyond what the regression results suggest and could wash out the institutional gains of the past decade in a single episode. The gains India has made are real, but whether it can hold on to them depends on decisions taken elsewhere.

The problem also needs to be viewed through a political lens, since external conditions drive a measurable share of the movement. A move in the international market shapes India’s long term rates and that move is imported. The RBI is then left with a choice whether to absorb it through the currency and reserves or to lean against it. Neither option is free; each comes at its own cost. This is the trade Rodrik describes (2011): India has accepted a degree of financial globalisation and has given up a measurable portion of the sovereignty with regard to its core economic lever. The choice looks rational as open capital brings investment and discipline and not just risk. However, one takeaway is that a credible central bank with monetary independence and a mandate focused on inflation targeting can win some of the lost control and drive the economic lever towards a domestically suited path.

To extend the work, the first step would be to estimate the same US-rate sensitivity for other emerging markets, such as Brazil, Indonesia, South Africa and Mexico, and try to empirically analyse whether India’s exposure and its apparent fall is normal or unusual. The second would be to add a measure of global risk appetite such as the VIX (global volatility) and split the sample into calm and stressed months. That would test whether the constraint tightens when the economic conditions are volatile in other words, whether it takes away the autonomy just when an economy would want some room to act. The third is to take the credibility story and try to look at whether globally central bank credibility measured, for example, by inflation-targeting adoption or forecast anchoring moves the needle on external sensitivity across countries. For India, this could be tied specifically to the 2016 inflation-targeting reform and test whether the external sensitivity fell around then, which would turn the pattern in Table 3 into a firmer claim.

References

Calvo, G. A. & Reinhart, C. M. (2002). Fear of Floating. Quarterly Journal of Economics, 117(2), 379–408.

Cohen, B. J. (1998). The Geography of Money. Ithaca: Cornell University Press.

Eichengreen, B. (2011). Exorbitant Privilege: The Rise and Fall of the Dollar and the Future of the International Monetary System. Oxford: Oxford University Press.

Eichengreen, B. & Gupta, P. (2015). Tapering Talk: The Impact of Expectations of Reduced Federal Reserve Security Purchases on Emerging Markets. Emerging Markets Review, 25, 1–15.

Farrell, H. & Newman, A. L. (2019). Weaponized Interdependence: How Global Economic Networks Shape State Coercion. International Security, 44(1), 42–79.

Fleming, J. M. (1962). Domestic Financial Policies under Fixed and under Floating Exchange Rates. IMF Staff Papers, 9(3), 369–380.

Frieden, J. A. (2015). Currency Politics: The Political Economy of Exchange Rate Policy. Princeton: Princeton University Press.

Keohane, R. O. & Nye, J. S. (1977). Power and Interdependence: World Politics in Transition. Boston: Little, Brown.

Kirshner, J. (1995). Currency and Coercion: The Political Economy of International Monetary Power. Princeton: Princeton University Press.

Miranda-Agrippino, S. & Rey, H. (2020). U.S. Monetary Policy and the Global Financial Cycle. Review of Economic Studies, 87(6), 2754–2776.

Mundell, R. A. (1963). Capital Mobility and Stabilization Policy under Fixed and Flexible Exchange Rates. Canadian Journal of Economics and Political Science, 29(4), 475–485.

Obstfeld, M., Shambaugh, J. C. & Taylor, A. M. (2005). The Trilemma in History. Review of Economics and Statistics, 87(3), 423–438.

Rey, H. (2015). Dilemma not Trilemma: The Global Financial Cycle and Monetary Policy Independence. NBER Working Paper No. 21162.

Rodrik, D. (2011). The Globalization Paradox: Democracy and the Future of the World Econ-omy. New York: W. W. Norton.

Strange, S. (1988). States and Markets. London: Pinter.

¹ The label was coined by Morgan Stanley strategist James Lord in an August 2013 research note CNBC, “Fragile Five: the new focus of currency wars,” 25 October 2013.
² Reporting in the Japan Times, Fortune and MarketWatch, 2026.
³ Remarks at Southern Methodist University, reported by Bloomberg and others, September 2026.
⁴ Fortune, “I am the house now,” September 2026.

About Authors

Ayaan Sood is a researcher and social entrepreneur whose work focuses on development economics, education policy and social impact. His research examines the relationship between public policy, institutional investment and real-world outcomes, with a particular interest in India’s education system. He is also the founder of Project DISHA and uses research and storytelling to explore how policy can translate into meaningful change.

Disclaimer : The opinions expressed in this article are the personal opinions of the author. The facts and opinions appearing in the article do not reflect the views of Indiastat and Indiastat does not assume any responsibility or liability for the same.