La science économique au service de la société

Documents de travail

Working papers serie, Chaire Risque macroéconomique

WP n°2024-12 - Geography Versus Income : The Heterogeneous Effects of Carbon Taxation

PNG - 14.3 ko

Par Charles Labrousse (Paris School of Economics) et Yann Perdereau (Paris School of Economics)

Executive Summary
Carbon taxes stand among the most effective instruments for mitigating greenhouse gas emissions. However, they induce strong distributional costs, as energy represents a larger share of expenditures for low-income and rural households. These distributional effects are likely to reduce the political acceptability of carbon taxation, as shown in France with the Yellow Vests protest and the subsequent carbon tax freezing. These asymmetric costs, and the lack of political acceptability that follows, pose a significant risk to the green transition. Therefore, a socially acceptable carbon taxation design should account for its redistributive effects.

In this working paper, Charles Labrousse and Yann Perdereau develop a dynamic general equilibrium model with both income and geographic heterogeneities, to capture that energy expenditures heavily depends on living area and revenue. Both imported fossil energy and locally produced cleaner energy are consumed as a non-homothetic final good by households and an intermediate input by firms. The model is precisely calibrated using French micro data, to match the energy bundle composition within each income quintile and living area. A gradual, permanent increase in carbon taxes on fossil energy used by firms and households is simulated, possibly at different rates. The model computes the aggregate and distributional welfare costs associated with this transition, considering various revenue-recycling policies.

The paper’s results highlight the distributive and political risks associated with the green transition. Firstly, geography outweighs income or wealth in determining the distributive effects of carbon taxation. While the fiscal burden is relatively evenly distributed across income quintiles, it varies significantly across living areas. Rural households bear approximately twice the cost of urban households due to their higher incompressible energy needs.

Secondly, taxing households’ emissions is considerably more regressive than taxing emissions from firms. Taxing households’ energy consumption is regressive, because of the non-homotheticity of energy consumption, disproportionately affecting low-income and rural households. Conversely, taxing firms’ energy consumption reduces both capital and labor income, affecting high-income households to a greater extent. Thirdly, it is possible to reduce emissions and make the policy progressive with respect to income. A 250 €/tCO2 carbon tax with a uniform lump-sum rebate reduces CO2 emissions by 18% per year, while enhancing overall welfare and reducing income inequality. However, this uniform transfer widens the rural-urban gap. Compensating for the loss experienced by rural households through targeted transfers entails a trade-off between equity and climate efficiency, as rural households exhibit a higher marginal propensity to consume energy. Compensating rural households is welfare-improving but comes with a 6% increase in total emissions compared to the uniform lump sum transfer.

In conclusion, the paper contributes to understanding the political risks associated with the green transition and proposes a more equitable and socially acceptable framework for carbon taxation. The paper argues that targeted transfers are crucial for communication and political acceptability. These transfers explicitly distinguish carbon tax revenue from government budget, clarifying that the tax aims to alter behavior rather than finance public deficits. Finally, this research emphasizes the paramount importance of geography in comprehending the aggregate and distributional effects of carbon taxes, hence suggesting that future carbon tax designs should take geographical factors into account.

WP n°2024-11 - The Flight to Safety and International Risk Sharing

PNG - 13.6 ko

Par Rohan Kekre (The University of Chicago Booth School of Business) et Moritz Lenel (Princeton University)

Executive Summary
The U.S. sits at the center of the international monetary system. There are two defining features of this role. The first concerns its currency. Relative to bonds denominated in the currencies of equally high-income countries, dollar bonds pay well when equities pay poorly, and have low expected returns when output has been declining. These imply that dollar bonds are a hedge whose value rises in bad times. The second concerns the U.S. international investment position. The U.S. is positively exposed to equities and negatively exposed to the dollar exchange rate. As such, it serves as the ’’world’s insurer’’ and transfers wealth to the rest of the world in bad times.

This paper proposes a quantitative two country business cycle model with nominal rigidities to jointly capture these patterns and study their implications. The two key ingredients of the framework are a time varying demand for safe dollar bonds and a higher risk tolerance of the U.S. relative to the rest of the world, bridging a growing literature emphasizing the safety and liquidity value of U.S. Treasuries with another strand of the literature that argues that the U.S. has a greater capacity to bear risk than the rest of the world.

In the model, an increased demand for safe dollar bonds, a flight to safety, implies that the relative return on all other assets has to increase. If US interest rates do not fall sufficiently to achieve this return differential, this instead is achieved by a decline in global consumption and investment as well as immediate dollar appreciation, increasing the returns on risky assets and foreign bonds going forward. The goods market and foreign exchange market responses are linked by a larger fall in U.S. output than output abroad, appreciating the U.S. terms of trade. As dollar bonds thus pay well in endogenously ’’bad’’ times, they earn a negative risk premium versus foreign bonds, and relatively risk tolerant agents insure the risk averse against such a shock. If agents in the U.S. are more risk tolerant than those abroad, this implies that U.S. net foreign assets fall on impact of the shock. In the periods which follow, the dollar depreciates, excess foreign bond and equity returns are high, global output recovers, and U.S. net foreign assets improve. These patterns are consistent with observed comovements in the data, but cannot be delivered by productivity and disaster risk shocks. Flight to safety shocks therefore provide a resolution to the ’’reserve currency paradox’’ elucidated by Maggiori (2017).

The quantitative model disciplines the demand for safe dollar bonds to match spreads in financial markets, and differences in risk tolerance across countries to match the sensitivity of U.S. net foreign assets to excess equity returns. The model generates untargeted comovements between relative bond returns, equity returns, output, and U.S. net foreign assets quantitatively in line with the data. It allows the authors to study global business cycles and the transmission of macroeconomic policy. Absent the time-varying demand for safe dollar bonds, global output would be roughly 15% less volatile, particularly so in the U.S. Absent the U.S.’ greater capacity to bear risk, its net foreign assets would be only as volatile as net exports, but net exports would in turn bear a greater burden in external adjustment and the U.S. would no longer earn positive average returns on its external position. Both the flight to safety and greater U.S. risk-bearing capacity played important roles in the Great Recession. Finally, the creation of safe dollar liquidity, such as via the dollar swap lines employed by central banks in recent crises, is globally stimulative but revalues wealth in the U.S.’ favor.

WP n°2023-10 - Tail risk in production networks

PNG - 11.5 ko

Par Ian Dew-Becker (Northwestern University, Kellogg School of Management)

Executive Summary
This paper describes the response of the economy to large shocks in a nonlinear production network. While arbitrary combinations of shocks can be studied, it focuses on a sector’s tail centrality, which quantifies the effect of a large negative shock to the sector – a measure of the systemic risk of each sector. Tail centrality is theoretically and empirically very different from local centrality measures such as sales share – in a benchmark case, it is measured as a sector’s average downstream closeness to final production. The paper then uses the results to analyze the determinants of total tail risk in the economy. Increases in interconnectedness in the presence of complementarity can simultaneously reduce the sensitivity of the economy to small shocks while increasing the sensitivity to large shocks. Tail risk is strongest in economies that display conditional granularity, where some sectors become highly influential following negative shocks.

WP n°2023-09 - Efficient Allocations under Ambiguous Model Uncertainty

JPEG - 15.6 ko

Par Chiaki Hara (Kyoto University), Sujoy Mukerji (Queen Mary University of London), Frank Riedel (University of Bielefeld) et Jean-Marc Tallon (Paris School of Economics)

Executive Summary
In this paper, Chiaki Hara, Sujoy Mukerji, Frank Riedel and Jean-Marc Tallon investigate consequences of ambiguity on efficient allocations in an exchange economy. Ambiguity is embodied in the model uncertainty perceived by the consumers : they are unsure what would be the appropriate probability measure to apply to evaluate consumption plans, and keep in consideration alternative probabilistic laws. Importantly, the degree of ambiguity aversion can vary across consumers. This heterogeneity has key implications regarding (1) the efficient allocation, and (2) its associated pricing kernel.

The paper focuses on environments where, under expected utility, the efficient consumption sharing rule is a linear function of aggregate endowment. In contrast, when consumers feature smooth ambiguity preferences with heterogeneous ambiguity aversion, it is shown that the efficient sharing rule systematically deviates from the linear benchmark. Intuitively, it becomes efficient to provide a smoother expected utility-across models-to the more ambiguity-averse consumers. Consequently, the efficient sharing rule favors the most ambiguity-averse consumers in the worst models - think recessions - while the least ambiguity averse consumers are favored in the best models-think expansions. Thus, under regularity conditions, the efficient allocation tends to allocate a larger share of resources to more ambiguity-averse consumers in recessionary times.

The authors then characterize the representative consumer and use it to find implications of heterogeneity in ambiguity aversion for the pricing kernel. As the assortative matching between ambiguity aversion and worse models in the efficient allocation may suggest, if consumers are heterogeneously ambiguity averse then the representative consumer features decreasing ambiguity aversion-and not constant ambiguity aversion, as assumed in common practice. The decreasing ambiguity aversion of the representative consumer implies that the market price of risk varies more pronouncedly between states associated with worse models and states associated with more optimistic models. In other words, the market price of risk is higher in recessionary states and lower in good states. This property is empirically compelling since the Sharpe ratio for the U.S. aggregate stock market is countercyclical and highly volatile. More generally, ambiguity aversion is shown to increase the elasticity of the pricing kernel, thereby increasing the Hansen-Jagannathan bound.

These results are particularly relevant to analyze households who need to forecast variables such as rainfall or temperature in the context of climate change. One may also consider households who try to forecast an unobservable hidden state, high growth or low growth, given observed quarterly realizations of the GDP. A third example is that of decision making in the face of a contagion engendered by a novel virus, for which decisions have to be made before learning the exact behavior of the virus.

WP n°2022-08 - The Curious Incidence of Monetary Policy Across the Income Distribution

PNG - 23.4 ko

Par Tobias Broer (Paris School of Economics, IIES, Stockholm University and CEPR), John Kramer (IIES, Stockholm University) et Kurt Mitman (IIES, Stockholm University, CEPR and IZA)

Executive Summary
How does monetary policy intervention affect the earnings and employment prospects of individuals across the income distribution ? Does the unequal incidence of monetary policy across the distribution amplify or dampen the response of aggregate consumption to changes in interest rates or future consumption ? The burgeoning heterogeneous-agent New Keynesian (HANK) literature has identified labor income as an important channel through which household heterogeneity impacts the transmission of monetary policy. Answering the foregoing questions is key to understand the transmission of monetary policy to the aggregate economy. However, the literature is still lacking direct empirical evidence on these transmission channels.

In this paper, Tobias Broer, John Kramer and Kurt Mitman first study empirically the heterogeneous effects of monetary policy surprises on labor earnings across the income distribution. Using high-frequency data on labor earnings and labor market status from Germany, they show that monetary policy has significantly larger effects on the earnings of low-income workers. This is mainly because their job-loss risk responds more strongly to changes in interest rates than that of high-income workers. This unequal incidence significantly reduces income inequality in response to monetary expansions and has long-lasting effects on employment rates of poor workers, which remain elevated even years after the initial shock. In particular, the authors find that an unexpected interest rate cut leads the Gini coefficient of labor earnings to fall significantly. In addition, monetary policy has significant effects on medium-run employment prospects : individuals who become unemployed in the month of a monetary policy expansion find jobs significantly faster, have significantly higher earnings, and remain employed significantly longer.

The authors then use a structural model to show how this heterogeneous incidence of monetary policy strongly amplifies its effect on aggregate demand. Relative to a model where unemployment risk is homogeneous across the distribution, heterogeneous incidence further amplifies the unemployment-risk channel because monetary policy affects more strongly the riskier workers who account for the bulk of precautionary savings. This positive association of level and cyclicality of risk in the cross-section makes aggregate precautionary savings more responsive to monetary policy. Their analysis suggests quantitatively important results : consumption increases by about a third after a monetary policy intervention.

WP n°2022-07 - Exchange rate policy and firm heterogeneity

PNG - 13 ko

Par Masashige Hamano (Waseda University) et Francesco Pappadà (Paris School of Economics and Banque de France)

Executive Summary
In a recent contribution, Obstfeld (2020) looks back at "The Case for Flexible Exchange Rates’’ made by Harry G. Johnson in 1969, and explores whether his argument survives the most recent academic critiques of exchange rate flexibility. He concludes that none of the arguments against exchange rate flexibility convincingly undermines the case for a flexible exchange rate. Nonetheless, policymakers have recently adopted exchange rate policies aimed at limiting the fluctuations of the exchange rate, as documented in Ilzetzki et al. (2019). In this paper, Masashige Hamano and Francesco Pappadà provide a rationale for managed exchange rate policies that protect industries and workers in the export market from exchange rate fluctuations.

The main contribution of this paper is to highlight the unexplored role of firm heterogeneity and nominal rigidities on the exchange rate policy trade-offs. In this economy, external demand shocks produce fluctuations in the nominal exchange rate that modify the selection of exporter firms. When firms are small on average and homogeneous in terms of productivity, the fluctuations on external demand may induce a large fraction of firms to enter or exit the export market. In presence of wage rigidity, large fluctuations in external demand translate in high wage mark-ups. In this context, the optimal exchange rate policy reduces the fluctuations of the nominal exchange rate and hence the uncertainty in the export market. These results therefore suggest that a managed exchange rate is welfare improving when firm heterogeneity is low, that is when many firms are subject to fluctuations in external demand. Instead, when firms are large on average and more heterogeneous, the benefits of dampened fluctuations in the exchange rate do not compensate for the costs associated with the high wage mark-ups of domestic firms. The optimal monetary policy therefore responds less to external demand shocks, letting the exchange rate free to float.

The two-country setup of this paper fits the description of two large economies (e.g. US and China) which both attempt to manage exchange rate fluctuations in favor of their own exporting sector. Further, this model relates to the case of one economy that has to choose the exchange rate policy vis-a-vis the currency of its main trade partner. For instance, consider the case of a country outside the Euro Area, which exports all of its goods in the Euro Area with producer currency pricing. This country has to choose whether to let its exchange rate to freely float with respect to the euro, rather than manage it or peg. This model shows to what extent demand fluctuations and the size of the exporter extensive margin may affect the choice of the exchange rate policy in the presence of nominal rigidities and imperfect financial markets. In particular, it shows that there might be an incentive for policymakers to use actively the exchange rate policy to insulate the demand in the trade sector from exchange rate fluctuations.

WP n°2022-06 - Herding through Booms and Busts

PNG - 10.7 ko

Par Edouard Schaal (CREI, ICREA, UPF, BGSE and CEPR) et Mathieu Taschereau-Dumouchel (Cornell University)

Executive Summary
Business cycle history is replete with examples in which new technologies have led to periods of massive investment that ended in severe economic downturns. One salient example is the 1990s boom in information technologies that culminated in the stock market crash of 2001 (“dot-com bubble”). While extreme enthusiasm about new technologies was initially fueled by the high volume of investment and rising valuations of IT companies, a crash eventually followed as some of the expected returns failed to materialize. Recent technological advances have reignited the debate over the potential negative spillovers that innovations can have on macroeconomic stability.

A common view is that shifts in expectation play a key role in shaping boom-bust cycle episodes. In his seminal work on the origin of economic fluctuations, Pigou (1927) emphasized the importance of beliefs in shaping the business cycle. In his view, booms can be caused by waves of optimism among business executives, and crashes arise when their lofty expectations turn out to be mistaken. This hypothesis has been explored in modern business cycle theory by the news-driven business cycle literature, pioneered by Beaudry and Portier (2004). According to this view, investors receive news about the future profitability of their investments, which sometimes turn out to be false. Boom-bust cycles arise after an initial sequence of positive news is later contradicted by experience.

These theories, however, remain mostly silent on the technological, social and psychological determinants that drive the evolution of beliefs. In most of these studies, investors’ beliefs obey an exogenous law, and boom-bust cycles occur after a specific sequence of shocks—first positive, then negative. In other words, these cycles remain attributed to unexplained factors, precluding a deeper understanding of the key drivers of business cycle fluctuations. What explains that beliefs follow a particular—and perhaps systematic—pattern which evolves from a phase of rising optimism to all-out pessimism ? Is growing optimism during the boom the consequence of luck or the result of particular interactions between investors that lead to instability and inefficiencies ? What causes precipitate the economy into a bust ? Providing answers to these questions is essential for our understanding of business cycles and for the design of stabilization policies.

This paper explores the role of rational herding as a source of macroeconomic fluctuations. In the theory, investors learn about the quality of an investment opportunity by observing the decisions of their competitors and can be tempted to invest when they see other market participants expand their operations. The introduction of a new technology of uncertain quality can trigger a slow-rising boom followed by a sudden crash, in line with the experience of the dot-com era. In the boom phase, the initial optimism of investors translates into high levels of aggregate investment, and high investment, in turn, leads to further increases in optimism. This self-reinforcing process can fuel a long-lasting expansion of the economy, which comes to an end when an overly optimistic view of the technology is no longer supported by the data. Investment collapses, taking down the rest of the economy.

The authors embed this mechanism into a dynamic stochastic general equilibrium model of the macroeconomy, amenable for quantification and policy analysis. The model is calibrated to the US economy, using in particular macroeconomic expectation surveys to discipline the learning block of the model. The theory is able to generate realistic boom-bust cycles in line with the experience from the dot-com era. The role of monetary policy and other macroeconomic stabilization tools are also studied. A key lesson emerges from the theory : rational herding provides a novel justification for the use of leaning-against-the-wind stabilization policies which reduce the incidence of boom-bust cycles and limit the severity of the ensuing crises.

WP n°2021-05 - Sovereign default and imperfect tax enforcement

PNG - 12.2 ko

Par Francesco Pappadà (Paris School of Economics and Banque de France) et Yanos Zylberberg (University of Bristol, CESifo and the Alan Turing Institute)

Executive Summary
Sovereign default risk typically decreases in response of fiscal consolidations. However, the response of sovereign default risk to fiscal policy is dampened when tax enforcement is weak. A fiscal consolidation leads to an expansion of the informal sector, thereby limiting fiscal surpluses, but also hampering future tax collection and failing to reduce default risk. For instance, during the European sovereign debt crisis of 2009 - 2014, several economies with relatively low tax enforcement implemented fiscal consolidations that led to significant welfare costs but limited effects on default risk.

In this paper, we study the dynamics of fiscal policy and default risk when tax enforcement is imperfect. The contribution of the paper is threefold. We first document stylized facts about tax compliance and its dynamics in economies with imperfect tax enforcement, most notably the relationship with fiscal policy and default risk. We then provide a model of sovereign debt with limited commitment in order to understand how the dynamics of tax compliance---disciplined by the empirical moments---affects optimal fiscal policy and default risk. Finally, we quantify the ignored, yet important, welfare cost associated with imperfect tax enforcement : a responsive tax compliance significantly constrains optimal fiscal policies, which, ultimately, has an impact on consumption smoothing.

We uncover novel empirical facts about the dynamics of tax compliance and its impact on default risk. First, we show that tax compliance is volatile and there is large heterogeneity in volatility across countries. Tax compliance is volatile because it strongly responds to fiscal policy and business cycle fluctuations. The heterogeneous volatilities across economies reflect large heterogeneity in such responses. In some economies with imperfect tax enforcement, a larger share of taxpayers hide their activity in downturns and in periods of austerity. In contrast with the standard behavioral response, the magnitude of fluctuations in tax compliance implies sharply decreasing returns to taxes, and some economies display an extreme form of fiscal fatigue. Second, the response of tax compliance to fiscal policy alters the relationship between fiscal policy and default risk. We find that fiscal consolidations are associated with a marked decrease in default risk, but only in countries where tax compliance is inelastic. Instead, when tax compliance strongly responds to taxes, this adjustment directly affects default risk and significantly limits the returns to fiscal consolidations.

We explore the implications of fluctuations in tax compliance on the dynamics of optimal fiscal policy in a model of sovereign debt where a benevolent government uses fiscal policy as a consumption-smoothing instrument. In our quantitative analysis, we evaluate how the dynamic properties of tax compliance affect optimal fiscal policy and welfare by comparing two economies differing along the tax compliance response to fiscal policy and business cycle fluctuations around the (same) steady-state level. The baseline economy differs from the low-response economy in two important dimensions. First, the baseline economy is ten times more likely to experience a default (with a yearly probability of 0.2%, and a yearly probability to be excluded from financial markets of 1.8%). Default is more likely, even though the baseline economy accumulates far less debt on average (10% of output versus 21%). Second, fiscal policy in the baseline economy is less able to smooth fluctuations in consumption : household consumption is much more volatile around the same average levels. We use the model to quantify the costs of such fluctuations and find that they are equivalent to a 2.2% decrease in certainty equivalent consumption. These findings illustrate that fluctuations in tax compliance constrain the set of feasible fiscal policies and significantly lower welfare.

WP n°2020-04 - Waiting for the Prince Charming : Fixed-Term Contracts as Stopgaps

PNG - 15.7 ko

Par Normann Rion (Paris School of Economics, ENS-PSL)

Executive Summary
Over the last decades, fixed-term employment has raised in the European Union. Fixed-term employment provides flexibility to firms in front of workload change, but it is costly to the workers : fixed-term workers face a higher unemployment risk than open-ended workers. Moreover, fixed-term employment also causes an inefficiently high turnover. As a response, European governments have carried out many reforms of employment protection legislation. Prominent questions remain, though. In particular, how do fixed-term contracts impact unemployment and welfare ? What are the main risks associated with the implementation of these reforms ?

To tackle these issues, I extend the classic model of Mortensen and Pissarides to add open-ended and fixed-term contracts. On the job creation side, firms post vacancies while unemployed workers search for a job. Vacancies and unemployed workers meet according to a matching function. New firm-worker matches differ in productivity. They optimally choose between going back to search or starting production under a fixed-term contract or an open-ended contract. On the job destruction side, firms lay off open-ended workers that become unprofitable and pay firing costs. Fixed-term matches only split with an exogenous probability at zero cost.

The future value of a match is uncertain to searching firms and workers. Two polar cases arise. If the initial productivity of the match is low, the firms and the worker go back to search. On the contrary, if the initial productivity is high, the firm and the worker lock up the match to make the most out of it ; the immediate gains overcome the potential firing costs in the future. The option of offering a fixed-term contract strikes a balance. It provides a production opportunity for the best rejected matches and a quick return to searching for a more productive match.

Two opposite mechanisms shape the equilibrium. On the one hand, higher firing costs encourage substitution towards fixed-term contracts. On the other hand, higher firing costs discourage the destruction of open-ended matches, which reduces unemployment and job creation of both open-ended and fixed-term contracts. As such, the response of unemployment and welfare is non-monotonous. I calibrate the model on the French labor market and find that small changes in firing costs cannot jointly enhance employment and welfare. The optimal reform consists in a large cut in firing costs and the ban of fixed-term contracts.

Post-reform dynamics are costly in the short run. Consider a 50 % cut in firing costs combined with the ban of fixed-term contracts. On impact, it leads to an increase in unemployment and decrease in open-ended employment. Unemployment remains above its initial value for at least one year before eventually reaching a lower value. Similarly, open-ended employment stays below its initial value in the short-run. Overall, an in-depth reform may improve open-ended employment and unemployment in the long-run at the expense of an adverse transition.

WP n°2020-03 - Financial Cycles with Heterogeneous Intermediaries

PNG - 15.4 ko

Par Nuno Coimbra (Paris School of Economics) et Hélène Rey (London Business School, CEPR and NBER)

Executive Summary
In the aftermath of the great financial crisis, the question of excessive risk-taking by banks and other financial intermediaries became very important in the minds of policy makers and academics alike. Understanding its causes and origins is essential in designing adequate regulations, since risk is also a fundamental and inherent aspect of the financial sector. One important concern is understanding whether monetary policy affects risk-taking in the financial sector and also when and how the two are linked. Did the period of historically low rates before the crisis worsen financial stability ? Is the current low interest rate environment also leading to the build-up of systemic risk ?

This paper looks at how competition between different financial intermediaries can generate a strong connection between policy rates and financial stability. Interest rates affect the composition of the financial sector and how assets are distributed between riskier and less risky financial intermediaries.

When interest rates become very low, riskier financial intermediaries find it very cheap to grow using leverage and competition can push safer, more conservative ones out of risky financial markets. For example, a cautious bank who wanted to provide mortgages during the boom might find it hard to compete with a neighboring bank which is willing to provide No-Income-No-Job-no-Assets (NINJA) mortgages at very cheap rates. As the name suggests, these were very risky mortgages where the lender did not even require the borrower to provide any information on their job, income or wealth. The easier it is for the riskier bank to find cheap funds to provide these mortgages, the harder it will be for the conservative bank to find clients for his own mortgages. This means that low interest rate environments can lead to lower financial stability through its effect on the composition of the financial sector.

A macroeconomic model with heterogeneous financial intermediaries helps clarify the full effect of interest rates on the composition of the financial sector. When interest rates fall, there are two competing effects. The first, already explained above, means that competition for financial assets drives up asset prices and decreases expected returns, which can lead to a decrease in the market share of safer banks and financial intermediaries. This leaves the financial sector more concentrated into riskier, highly-levered intermediaries generating low risk-premia and higher systemic risk. The second effect works through the cheaper cost of liabilities. If the cost of liabilities falls, then more conservative intermediaries might find it now worthwhile to increase their holdings of risky assets and so the market becomes less concentrated and financial stability is improved. The first effect tends to dominate when interest rates are already low to begin with, while the second one dominates for environments with higher rates. So an expansionary monetary policy worsens financial stability in low interest rate environments, but can actually improve it when rates are high to begin with.

In conclusion, when interest rates are not very low, central banks who need to stimulate the economy do not face any trade-off in terms of financial stability. But when interest rates are already low, further cuts might come at a cost in terms of systemic risk and central banks would do well to recognize this.

WP n°2019-02 - From Microeconomic Favoritism to Macroeconomic Populism

PNG - 15.2 ko

Par Gilles Saint-Paul (Paris School of Economics, ENS-PSL and NYU-AD)

Executive Summary
Why would people support macroeconomic policies that are likely to lead to sovereign crises, balance of payments crises, and the like ? A rational explanation is based on favoritism – an institutional feature of society implying that some social groups have better access to public goods than others. A favored group that bears a low fraction of the costs of a crisis but benefits in the short-run from unsound policies is likely to support fiscal indiscipline. This paper formalizes the role of favoritism for public spending, indebtedness, and crisis in an illustrative model based on Saint-Paul et al. (2017), and studies support for political parties implementing it, so-called « populists ». It argues that favoritism shaped the recent history of French pension reforms and confirms its effect on macroeconomic policy across a panel of countries.


Favoritism generates fiscal indiscipline if the decisive voter is favored relative to the mean in crisis times. When the government’s fiscal capacity is insufficient to cover its obligations and society enters a fiscal crisis, people’s access to their entitlement of publicly provided goods must be rationed. Under favoritism, this adjustment is mostly burdened on unfavored groups. By pursuing unsound fiscal policies, the favored decisive voter can engineer future crisis and manages to have the public good on average financed by others. For example, increasing the level of public debt implies more rationing in crisis times but relatively less so for favored groups who also benefit from increased private consumption possibilities through higher debt. Absent crisis, Ricardian equivalence holds and debt has no effect on society. Thus, the incentive to raise more debt stems from states of fiscal crisis only. It becomes stronger the greater the probability of a crisis. Similarly, favoritism increases public spending. Since favoritism need not be a structural property of society, the paper then studies how favoritism arises as an outcome of collective choice between either a populist or a technocrat. The populist implements favoritism regardless of fiscal and macroeconomic conditions. The technocrat sticks to anonymity and rations access to publicly provided goods only in a crisis. It is shown that the support for the populist is greater, the greater greater the likelihood of a crisis.


The recent history of French pension reforms is used as an example for this paper’s mechanisms. In particular, it was rational for French public sector employees to support the reduction in the retirement age from 65 to 60, implemented by the Mitterand administration following his 1981 presidential victory, despite overwhelming evidence that it was fiscally unsustainable. Civil servants, having their own special pension system, had good reason to anticipate that subsequent adjustments were likely to hit other social groups proportionally more. Indeed, the first attempt to balance the accounts of the pension system, the 1993 Balladur reforms, made it more difficult for private sector employees only to retire at the age of 60 by raising the duration of their contributions from 37.5 to 40 years. Formal statistical evidence in favor of the model’s predictions is provided by merging four country-level databases, the IMF’s World Economic Outlook for macro indicators, the Institutional Profiles Database (IPD) for indicators of favoritism at the institutional level, the Database of Political Institutions (DPI) for indicators of party ideology, and the CRAG-Bank of Canada database of sovereign defaults to get proxies for fiscal crises. Overall, the results support the theory. Unequal treatment from administrations, a proxy for favoritism, is more likely to generate high debt, high public expenditures, and high deficits, as well as (indirectly through debt) sovereign default. Furthermore, adverse fiscal conditions such as high public debt, high deficits, and low fiscal capacity are more likely to lead to a populist government.

WP n°2019-01 - Trading ambiguity : a tale of two heterogeneities

PNG - 17.8 ko

Par Sujoy Mukerji (Queen Mary University of London), Han N. Ozsoylev (Koç University) et Jean-Marc Tallon (Paris School of Economics, CNRS)

Executive Summary
The financial literature largely assumes that investors know the distribution of asset returns. In most real-world situations, however, decision makers are uncertain about the data-generating process underlying asset returns. This has important implications for portfolio choice, because investors may prefer portfolio allocations that are robust across the set of return distributions believed to be possible. Novel to the literature, the paper shows that such ambiguity potentially explains several puzzling cross-sectional regularities.

Two heterogeneities are key : First, the heterogeneity in the uncertainty about the mean of an asset’s return distribution. This uncertainty parameter encapsulates the ambiguity of assets. It is high, for example, for stock returns of new-technology companies whose risks have not yet been fully learned. Second, investors differ, additionally to their risk aversion, in their tolerance for ambiguity. Together, these heterogeneities give rise to a parsimonious extension of the standard mean-variance framework (referred to as robust mean-variance) in which investors face a three-way trade-off between expected return, variance, and ambiguity. The paper considers, in turn, the implications for portfolio choice, equilibrium prices and returns, and trade upon the arrival of public information.

More ambiguity-averse investors are shown to hold less of the more ambiguous assets. This finding not only confirms the failure of the classical mutual funds theorem (Tobin, 1958) in the presence of heterogeneous ambiguity-aversion, but the direction of departure is also empirically compelling. Indeed, conservative investors are commonly encouraged to hold more bonds, relative to stocks. Such financial planning advice is inconsistent with standard mean-variance investors (asset allocation puzzle), but can be accommodated in this framework.

Turning to equilibrium prices, the authors show that despite the failure of the mutual funds theorem, a single-factor pricing formula emerges. As in the standard consumption asset pricing model (CAPM), the single factor is the excess return of the market portfolio. However, the CAPM beta is adjusted by the extent to which the ambiguity of the asset return is correlated with the ambiguity of the market portfolio. Two uncertainty premia explain the cross-section of expected returns : a risk premium and an ambiguity premium. The latter has the potential to explain the size and value premia documented by Fama and French (1992, 1993). High book-to-market firms, which tend to be in financial distress, and small-cap firms, due to their over-reliance on external financing, likely carry a high ambiguity premium.

In the dynamic extensions of the model implications of public signals for trading volumes are analysed. Earning announcements or aggregate uncertainty shocks are shown to induce trading if and only if agents are heterogeneously ambiguity averse. Such trade occurs because public signals change the return-risk-ambiguity trade-off, making investors seek a different allocation across ambiguous assets depending on their different ambiguity tolerances. Trade results from uncertainty sharing considerations and leads to no or very small price movements, which is consistent with the empirical literature.

The paper concludes by proposing strategies to estimate the ambiguity of individual assets returns. Since ambiguity about the return distribution is taken to be the uncertainty about the mean of the return distribution, a measure can be obtained from a Bayesian estimate of this parameter.