key: cord-1045434-k3o7snm5
authors: Goodhart, Charles; Masciandaro, Donato; Ugolini, Stefano
title: Pandemic Recession and Helicopter Money: Venice, 1629--1631
date: 2022-01-12
journal: Financial History Review
DOI: 10.1017/s0968565021000214
sha: 44637ff5b3e43044743e0094abde133e2663cb00
doc_id: 1045434
cord_uid: k3o7snm5

We analyse the money-financed fiscal stimulus implemented in Venice during the famine and plague of 1629--31, which was equivalent to a 'net-worth helicopter money' strategy -- a monetary expansion generating losses to the issuer. We argue that the strategy aimed at reconciling the need to subsidize inhabitants suffering from containment policies with the desire to prevent an increase in long-term government debt, but it generated much monetary instability and had to be quickly reversed. This episode highlights the redistributive implications of the design of macroeconomic policies and the role of political economy factors in determining such designs.

Whenever economic conditions become really critical, the concept of helicopter money frequently re-emerges -and the recent Covid-19 crisis was no exception (Masciandaro 2020) . The term "helicopter money" originates from Milton Friedman, who famously imagined (without endorsing it) a hypothetical situation in which freshly-issued central bank money would be randomly distributed to households. 1 Friedman's description, however, was characteristically short in detail on how such a monetary policy might concretely be implemented. As a result, while followers have agreed on the definition of helicopter money as "a money-financed fiscal stimulus […] that […] requires neither an increase in the stock of government debt nor higher taxes, current or future" (Galí 2020, p. 2), they have diverged in their views of the design of such a policy. On the one hand, most economists have interpreted helicopter money as a permanent increase in the liabilities of the central bank (see esp. Buiter 2014). As such a situation, which we might describe as "monetary-base helicopter money", has actually occurred relatively often in history, it represents a less radical option than it might appear at first sight. On the other hand, others have interpreted helicopter money as a permanent decrease in the assets (i.e., in the net worth) of the central bank (see esp. Masciandaro 2020) . Such a situation, which we might describe as "net-worth helicopter money", is a much more radical option than the previous one, and appears to have occurred only on exceedingly rare occasions in history.

In order to better understand the difference between "monetary-base" and "networth helicopter money", imagine a stylized balance sheet of the public sector made of the separate balance sheets of the fiscal authority (the Treasury) and of the monetary authority (the central bank: see Figure 1 ). The fiscal authority's assets consist of Treasury deposits with the central bank (TD) and of other Treasury assets (TA); its liabilities consist of the Treasury's net worth (TW), of marketable government bonds (TB), and of direct (unmarketable) loans from the central bank (TL). The monetary authority's assets consist of its portfolio of marketable government bonds (TB), 2 of its bullion and foreign reserves (BR), and of its direct (unmarketable) loans to the Treasury (TL); its liabilities consist of its net worth (BW), 3 "Monetary-base helicopter money" occurs in two steps. First, the Treasury issues new bonds and these are all bought by the central bank (on either the primary or the secondary market), which allows for a temporary increase in Treasury deposits with the central bank ( Figure 2 .1). Then, the Treasury proceeds to spend this money, which ends up being held by the private sector: the monetary base hence increases. On the fiscal authority's balance sheet, the money-financed fiscal stimulus reduces the Treasury's assets without reducing any liability except the Treasury's net worth, and can therefore be regarded as a loss (Figure 2 .2). On the whole, the difference between this "monetary-base helicopter money" strategy and a classical fiscal expansion is quantitative rather than qualitative, as the boundary between the two is defined exclusively by the size of the government debt monetization by the central bank. 3 The monetary authority's net worth (BW) will include central bank equity when the latter formally exists. Note that central bank equity may be owned by the fiscal authority -in which case, it will be included in the Treasury's assets (TA) -but this will not always necessarily be the case. 4 Note that the monetary authority's and fiscal authority's balance sheets might substantially differ in terms of their size. By contrast, a "net-worth helicopter money" strategy is qualitatively different from a classical fiscal expansion. Here, the fiscal authority instructs the monetary authority to credit money on the private sector's accounts with it. This increase in the central bank's liabilities should be theoretically matched by an increase in its assets -viz., in its direct loans to the Treasury (Figure 3 .1). In truth, however, these loans simply do not formally exist -which is economically equivalent to a Treasury default on its direct borrowings from central bank. As a result, the money-financed fiscal stimulus here reduces the monetary authority's (instead of the fiscal authority's) assets without reducing any of its liability except the net worth, and can therefore be regarded as a loss (Figure 3 .2). From a long-term perspective, "net-worth helicopter money" is therefore irreversible for the monetary authority, whereas "monetary base helicopter money" might potentially be (partially or totally) reversed by selling marketable Treasury bonds to the private sector. The biggest difference between the two is however of short-term nature, as "net-worth helicopter money" fragilizes the solidity of monetary authority and directly reduces its margins for manoeuvre in the management of the value of issued money, thus rendering it less able to counteract an inflationary outburst. 5 5 The literature has generally only considered long-term implications, arguing that the only difference between irreversible "monetary base helicopter money" and "net-worth helicopter money" is the fact that the former affects the Treasury's expenditures while the latter affects its revenues (as the Treasury is the only recipient of the central bank's dividends, which are negatively impacted by the latter strategy: see esp. Buiter 2014). This conclusion is based on the assumption that the monetary authority is economically indistinguishable from the fiscal authority -which may not actually be always true, as it is the case today e.g. in the Eurozone, and as it was very often the case in the past. By contrast, the short-term implications of each strategy on monetary policy implementation have not been considered so far. Theoretically, it is unclear whether one version of "helicopter money" is always necessarily superior to the other. Empirically, the implementation of "net-worth helicopter money" has been so rare in history that it has been practically impossible to analyse it. In this article, we study one rare historical episode in which "net-worth helicopter money" appears to have actually been implemented. This happened in the Republic of Venice in 1629-31interestingly, precisely in the context of a full-fledged pandemic which bears many resemblances with Covid-19. We describe the historical context in which fiscal monetization took place, explain why it can be considered as an example of "net-worth helicopter money", and analyse its consequences. In particular, we ask what were the political economy conditions which generated the choice of this very peculiar version of expansionary fiscal policy.

The remainder of this article is organised as follows. In Section I we briefly describe the Venetian monetary and fiscal systems during the first decades of the seventeenth century.

In Section II we analyse the huge real economic shocks of 1629-1631, as well as the containment policies put in place by the Venetian government during this period. Section III focuses on the "net-worth helicopter money" strategy implemented in order to finance expenditure, and offers a political perspective to help explain why the Republic chose this radical option rather than more classical ones. Section IV concludes. In Venice, the relative size of the public sector with respect to the domestic economy was extraordinarily high for coeval standards. In the real sector, the government, via the Grain Office and then the Fodder Office, was active in the grain market, in order to address and to stabilize the volatility in the supply and price of food, which was an extremely sensitive political issue. As for the financial sector, the financial district of Venice -the Rialto area -was entirely owned by the Republic, which rented to the private bankers the benches at which they used to operate in what was considered as a public concession. Private bankers' books were considered public records, the bank transfers being a legal way to discharge debt under Venetian law (Ugolini 2017, pp. 37-9 ). Yet the ideal that inspired the Venetian government was that, as far as possible, the State should not replace private initiative in markets (Dunbar 1892, p. 308). The Senate was the body that designed the Republic's economic policy, with its committees responsible for overseeing every aspect of production and distribution (Rapp 1976, p. 139) . With the Republic being firmly controlled by an oligarchy of merchants, the government's goal was just to provide the services that were essential for business but too expensive or too risky to be provided by the businessmen themselves (Ugolini 2017, p. 38) .

The Republic's tax system, featuring both "direct" and "indirect" taxes, was rather muddled and complicated. Reflecting the balance of power, the system had a marked "regressive" nature. From the second half of the sixteenth century until the 1620s, the Republic of Venice gradually increased per capita taxation, and this growth accelerated much more markedly in the aftermath of the 1630 plague (Alfani and Di Tullio 2019, pp. 24-33 and 147). The Republic had a long and generally good record with public debt management, which had been implemented since at least 1164. In early times, Venice had regularly resorted to forced loans from the patricians; by the seventeenth century a more market-friendly approach had prevailed thus increasingly attracting foreign investors, yet the government debt was still overwhelmingly held by the domestic upper class (Fratianni and Spinelli 2006, pp. 262-3; Alfani and Di Tullio 2019, p. 172) . 6 The Republic issued both floating and funded (long-term) debt, though usually the role of floating debt was limited; in the course of the sixteenth century, the management of the funded debt had been entrusted to the State-owned Mint (Pezzolo 2003a ). In tranquil years, the Republic generated substantial fiscal surpluses in order to repay all its debt (a goal which had been fully attained, for instance, in 1600), thus enhancing its creditworthiness (Sissoko 2002, p. 8; Fratianni and Spinelli 2006, p. 263; Alfani and Di Tullio 2019, p. 172) . Therefore, at the beginning of the seventeenth century the state of public finance in Venice was good (Pezzolo 2003a, p. 103) , and the government borrowed at an interest rate which was substantially lower than the average rate of return from trade (Sissoko 2007, p. 6) .

By the seventeenth century, Venice had reached a degree of monetization unknown for centuries anywhere else. As usual in early-modern times, the official monetary anchor consisted of commodity money (coins) issued by the Mint (Al-Bawwab 2021); however, inhabitants commonly used cheques and bank transfers (even among the lower middle class), which allowed them to economize on coins (Mueller 1997, p. 24; Fratianni and Spinelli 2006, p. 271 ). This had been made possible by the development of the private deposit banks operating on the Rialto square, but by the end of the 16 th century all such banks had gone bust and no serious private initiative was any longer available for running the business on the conditions imposed by regulation (Ugolini 2017, pp. 35-43) . As a result, the State eventually took the initiative on itself, and on April 1587 Venice's first public bank (called "Banco della Piazza di Rialto") started its operations (Soresina 1889, p. 7; Roberds and Velde 2014, p. 19; Bindseil 2019, pp. 207-10) .

In theory, the Banco di Rialto was supposed to represent a case of quasi-narrow banking, given that it was obliged by law to accept only deposits in coins, and cash was always to remain available at the request of depositors (Anonimo 1847, p. 564; Dunbar 1892, p. 321), defining tendentially a policy of 100 per cent reserves (Sissoko 2002, pp. 7 and 10) ; transfers had to be made simultaneously between creditors and debtors (Dunbar 1892, p. 321; Roberds and Velde 2014, p. 20) . In practice, however, because the coins into which the Banco's liabilities were formally convertible had been withdrawn from circulation in 1588, bank liabilities were de facto inconvertible into the new circulating coins (Roberds and Velde 2014, pp. 21-2; Ugolini 2017, pp. 225-6) . Coins were the main item on the asset side of the Banco's balance sheet, also private commercial debt was allowed to a certain extent (Ugolini 2017, p. 44 (Sissoko 2002, p. 8; Roberds and Velde 2014, p. 21) . During these years the two legal tenders -commodity and scriptural money -were imperfect substitutes, the conversion rate between the two being determined on the market (Dunbar 1892, p. 318; Inclimona 1913, p. 149; Fratianni and Spinelli 2006, p.271) . The existence of a positive premium ("agio") for Banco money with respect to coins was almost a constant in the Venetian experience (Siboni 1892, p. 291; Magatti, 1914, pp. 285-9; Roberds and Velde 2014, p. 17) . As a matter of fact, scriptural money at the time was safer for obvious reasons, and the quality of available commodity money was especially poor during the "bullion crisis" of the late sixteenth and early seventeenth century (Dunbar 1892, pp. 309, 321, and 330-1; Ugolini 2017, pp. 225-7) .

In May 1619 the government created a new public bank -the Banco del Giro -with floating (short-term) public debt and coins on the asset side of its balance sheet and deposits on the liability side (Soresina 1889 , p. 9, Siboni 1892 , p. 288, Inclimona 1913 p.152, Roberds and Velde 2014 Bindseil 2019, pp. 215-7) . In general, the State's creditors were likely to become floating debt holders using the transfer mechanism of the State bank. This sort of mechanism had been first introduced in the thirteenth century when the Grain Office and Salt

Office had started providing transfers to their creditors, and also the Fodder Office had resorted to it from 1608 to 1614 (Pezzolo 2003a, p. 63; Roberds and Velde 2014, p. 24; Ugolini 2018, p. 6) . While the Banco della Piazza di Rialto was a deposit bank, the Giro bank was a device to make the public debt easily transferable, turning it into a means of payment (Roberds and Velde 2014, p. 22) , and "paying deposits at the call of the depositor, like the existing Banco di Rialto", with the possibility of deposit overdrawing, i.e. making loans (Dunbar 1892, p. 325).

The account holders were floating debt holders; the Giro bank was allowed to accept deposits of private individuals only after the closure of the Banco della Piazza (Sissoko 2002, p. 11) .

The functioning of the Giro bank was simple (Ugolini 2017, p. 43) : the government opened accounts to merchants having credits to the Republic and to public officials as well (Soresina 1889 p.16, Inclimona 1913 p.1146 , that could be converted into coins upon authorization; merchants and public officials became Republic's debt holders. The credit of one accountholder could always be freely transferred on demand to another accountholder, and the corresponding amount would continue to circulate until the final repayment to the last bearer cancelled it out (Soresina 1889, pp. 12 and 16) . Banknotes were not issued (Siboni, 1892, p. 291) . The Giro bank liabilities were legal tender for any payment greater than one hundred ducats, while its clearing activity was possible also for payments lower than one hundred ducats (Soresina 1889 p.12) . Moreover, from July 1627 the account holders could pay import taxes using Giro bank transfers (Soresina 1889, p. 20; Siboni 1892, p. 294 In general, war, famine and epidemics (the so-called "Three Horsemen of the Apocalypse")

were often associated in the preindustrial world (Alfani 2013, p. 43), and this episode was no exception to the rule. Starting from March 1629, French and Imperial troops crossed the Alps to participate in the conflict (Alfani and Percoco 2019, p. 1171). In the Italian states, the first decades of the seventeenth century were characterized by severe food shortages (Alfani 2018, p. 152), and the already meagre supply of food was further jeopardized by war. In general, in

Venice famine episodes were less intense than elsewhere thanks to the Republic's capacity to collect grain from the rest of the Mediterranean (Todesco 1989, p. 11 ), but this came at a sizable cost for the government, which sold the grain at subsidized prices to the population (Ugolini 2017, p.37) . From a fiscal viewpoint, therefore, the war and famine put considerable pressure on public finances already. As early as April 1629, the governors of mainland cities in the city occurred between September and December 1630 -20,923 deaths -with a peak in October 1630 (Ell 1989, p. 130) , and in total 43,088 deaths were recorded over just three years; the population of Venice was 141,625 in 1624 and became 102,243 in 1633, a reduction of nearly 30 per cent (Lazzari et al., 2020, p.3) . Such figures are consistent with the 35 per cent estimated mortality in Northern Italy, and should be compared to an estimated average annual mortality of between 2.7 and 3.7 per cent in normal times (Lazzari et al. 2020, p. 3) . Many different indicators agree that the shock to the domestic economy was colossal, disrupting many diverse aspects of economic life including the arts industry -music production shrank by 40 per cent (Gonzaga Band 2018), while the average price of paintings collapsed by 81 per cent (Etro and Pagani 2010). While traditionally this epidemic has been considered a short-run disaster with limited long-run impact (Rapp, 1976, p. 154 To address the pandemic, the Senate had to decide immediately its policy action.

During the previous 1576 pandemic the Venetian government had reacted slowly, both denying the plague and downsizing the number and nature of deaths (Palmer 1978, pp. 238-75; Preto 1979, p. 123 ). The first and most urgent issue consisted of protecting public health by designing and implementing a containment policy. In the Venetian territories urban mortality rates during 1630-1 were severe (ranging from 433 per thousand in Chioggia to 615 per thousand in Verona), while in Venice itself the mortality rate was 330 per thousand, pointing to a certain success of the strict lockdowns implemented there (Alfani and Di Tullio 2019, p. 115). Venice had passed its first legislation to address epidemics in 1423, and a Health Office had been established in 1490 (Palmer 1978, pp. 51 and 85) . The Health Office used its authority to close shops, as well as to prohibit auctions and markets (Allerston 1996, p. 279).

These measures hit the majority of Venetians, who became unable to work during epidemics (Pullan 1960, p. 26; Biraben 1973, p. 145; Cipolla 1976, p. 42; Allerston 1996, pp. 292-5) .

The containment measures "were carried into effect on a colossal scale with full resources of the state" (Palmer 1978, p. 142) . The government was aware of such negative signaled in the Venetian mainland (Preto 1979, p. 141 effects on economic activity, so it then tried to alleviate the inhabitants' losses (Pullan 1964, p. 410; Cipolla 1976, p. 41; Biraben 1973, p. 57) . 10 In normal times, social expenditures were very low in Venice: for example, available data for 1602 and 1633 -i.e. before and after the pandemic -show that social expenditures were negligible, amounting respectively to 0.2 and 0.4 per cent of total expenditures; in the same years the service of debt amounted respectively to 8.2 and 19.9 per cent (Alfani and Di Tullio 2019, p. 167). But during a pandemic, things were different. In practice, during the 1630 pandemic the government bought necessary goods from merchants to distribute them to the confined population, 11 as it had already done during the 1575 plague (Pullan 1964, p. 409) . 12 When districts were put into quarantine, their inhabitants were provisioned by the State (Palmer 1978, p. 143) . The overall fiscal effort to help inhabitants included subsidies and other fiscal help given to affected communities, and distributions of free rations of grain (Alfani 2018, p. 162). When decisions to destroy supposedly infected goods were made, compensation was paid -though not always in full (Pullan 1964, pp. 251 and 319; Allerston 1996, pp. 278 and 287) . Moreover, the Venetian government influenced employment and nominal wages in the sectors under its total or partial control, including all the activities strictly related to health (esp. body clearers, who got paid huge salaries during pandemics: Allerston 1996, p. 296) and defence (esp. Arsenal workers, who had their wages paid despite being confined to their homes: Pullan 1964, p. 420 ). These fiscal transfers helped directly or indirectly inhabitants in trouble, particularly those in the lowest classes, which represented the largest part of the overall population: such people were often in debt, and at the greatest risk of crossing the boundary between subsistence and 10 A textile merchant pleaded for the quarantine to be lifted, given that "an incomparable greater number of people has died purely as a result of unemployment than of typhus or any other contagious disease" (Pullan 1964, p. 409) ; Verona was reported to be suffering more from the lockdown than from the disease itself (Palmer 1978, p. 275) . Bribery episodes were registered, merchants being anxious to get their goods into Venice (Palmer 1978, pp. 231-5) . Also the regular activity of the Mint suffered during the plague episodes (Stahl 2001, pp. 42-53) . 11 On December 3, 1630, the civil authorities of the city of Bergamo expressed increasing concern about the growing debt that the municipality was accumulating to address the costs of the pandemic (Archival source: Biblioteca Angelo Mai, -60, class. 1.2.3.1-60 1629 dicembre 15 -1632 maggio 22 "Consiliorum" pp. 108-9). 12 Between December 1628 and November 1629 (i.e., during the early times of the famine), the Venetian Fodder Office ("Provveditori della Biave") had already sent to the mainland territories more than 236,000 ducats worth of grain, a sum which roughly corresponded to 10% of the Republic's total expenditures during the "normal" year 1633 (Pezzolo 2006, p. 84) . The authorities of Vicenza and Treviso did not obey letters from Venice that ordered them to send grain to other territories (Lombardini 1963, p.27 poverty (Alfani and Di Tullio 2019, pp. 62-3). In the absence of subsidization, these people would have been most likely to revolt against the government's containment measures, as the opportunity cost of rioting would have been extremely low to them.

The 1629-31 famine and plague thus obliged the Venetian government to implement subsidies on a colossal scale, but how was such a huge fiscal expansion financed? Taxes were actually increased; 13 however, the plague had made tax collection more difficult, increasing the fiscal pressure (Pezzolo 1994, pp. 322-3; Alfani and Di Tullio 2019, p. 29) . To help alleviate such pressure, the government also allowed Jews to lend on collateral outside of the ghetto (Preto 1979, p. 144; Allerston 1996, p. 293) . But this was far from enough for financing the fiscal expansion. To deal with its worsening deficit, the Republic had to resort to fiscal monetization through the scriptural money issued by the Banco del Giro: in fact, the government paid its creditors by merely crediting their current accounts with the bank. The size of the Giro bank's liabilities, which had generally been less than one million ducats in the 1620s, rose to 2,071,168 ducats in April 1630, and surpassed 2,666,926 ducats in June 1630 (Soresina 1889, pp. 19, 23, and 29; Siboni 1892, p. 290; Roberds and Velde 2014, p. 24) . While the Giro bank's liabilities kept increasing in concert with the famine and the bubonic plague (Soresina 1889, p. 29) , the asset side of its balance sheet underwent a serious deterioration: the rise in scriptural money held by the public was theoretically matched by an increase in floating (non-marketable) government debt, but in reality such a debt did not actually exist, so that in fact the bank's assets were actually decreasing with respect to its liabilities. This situation, which de facto generated a loss for the issuing bank, precisely corresponds to the situation described in Figure 3 above. Fiscal monetization was actually decreasing the net worth of the monetary authority: in view of this, we interpret this episode as an historical illustration of the "net-worth helicopter money" strategy.

The choice of implementing such a radical strategy triggered a number of serious consequences. First, the government-induced growth of the Banco del Giro's business 13 On August 4, 1629 a new lump-sum wealth tax was introduced, and a second one was added in 1630 (Preto 1979, p. 144; Pezzolo 2003a, p.69; Pezzolo 2003b, p. 110; Pezzolo 2021, p. 74) . Moreover, the "duties" -i.e. the set of taxes that affected the transit and consumption of goods (Alfani and Di Tullio 2019, p. 24) were increased twice in 1629, and once more in 1630 (Pezzolo 2021, p. 71). crowded out the activity of the Banco della Piazza di Rialto (Ugolini 2017, p. 44) . The balance sheet of the Banco della Piazza had reached its peak of 1.7 million ducats in 1618, i.e. one year before the establishment of the Banco del Giro; in 1630, the amount of deposits with the former had dropped to 56,185 ducats only, making the bank moribund until its final demise in 1637 (Siboni 1892, p. 290; Sissoko 2002, p. 8; Pezzolo 2018, p. 155) . Second, the monetary expansion entailed a stark depreciation of the value of Giro bank money, which the bank had no way to counteract. Up to 1625 the premium between Giro bank money and coin had been positive and substantial; then from 1625 the premium began to fall, slowly at first and then precipitously in 1630, eventually turning negative (Soresina 1889, p. 29; Roberds and Velde 2014, p. 24) : the agio was equal 20 per cent in 1624, then it dropped to 19.5 per cent in early 1629 and fell into negative territory (-10 per cent) in 1630 (Pezzolo 2018, p. 156) . The depreciation of Giro bank money can be confirmed by looking at its market value in terms of silver. As shown in Figure 4 , during the 1629 famine Giro bank money had already lost 10 per cent of its silver value, and at the peak of the plague (in September 1630) it was down as much as 25 per cent with respect to its pre-crisis valuation. The overexpansion of the money supply thus quickly triggered a substantial monetary depreciation, which the bank had no means to counteract in view of the weakness of its balance sheet. This depreciation forced the government to backtrack on its "net-worth helicopter money" strategy and reform its monetary policy setting. Already in July 1630, a . This reflected the distribution of political rights, which was strictly reserved to the upper class. However, when calamities occurred, the patricians had to take into account the expectations of the lower classes. Urban populations were watchful of rulers, and they were ready to riot if they became convinced that the government was not doing all it could, (and should) have done to ensure the availability of food, guaranteeing the "right to bread" (Alfani 2018, p. 162). Politicians had much to fear, also in terms of personal safety, from riots motivated by distributional reasons -"injustice" -so that incentives to act were really strong.

Indeed, most popular riots in early modern Venice seem to have been caused by "political" claims. Inhabitants' preferences mattered, although even full citizenship did not guarantee full political rights, as those were reserved for the patriciate (Alfani and Di Tullio 2019, pp. 13 and 61 real subsidies without being sensitive to the "monetary externalities" generated by the policy. 15 In this respect, the government decision can be considered consistent with the aim of pleasing the majority of inhabitants during a big macroeconomic downturn, thus enhancing consensus and avoiding riots. The theoretical framework behind this interpretation is illustrated in the Appendix. At one point, however, monetary instability became too strong and the government risked losing control of the useful monetization mechanism. At that point, those that were mostly sensitive to "monetary externalities" (i.e. the ruling upper class) were ready to accept some "sacrifices" in order to put an end to currency depreciation. The "sacrifices" consisted of accepting a conversion of liquid sight debt (Giro bank money) into illiquid long-term debt (the "Mint deposits"). Acceptance of this conversion (somewhat akin to a debt restructuring) by patricians allowed for reversing part of the monetization and putting an end to instability. Importantly, the "sacrifices" were such only in the short term, as in the long term high interest rates were duly paid to the holders of the newly-created funded debt. Because the conversion was actually financed through earmarking future tax revenues, and because the tax system was strongly regressive in Venice, the short-term redistribution in favour of the poor was offset by a long-term redistribution in favour of the wealthy. Indeed, the 1629-30 money-financed fiscal expansion did not cause a permanent change in the condition of poor inhabitants, since analysis of income distributions shows that, contrary to the Black Death of 1348, the bubonic plague of 1630 did not trigger a phase of sustained inequality decline (Alfani and Di Tullio 2019, p. 116; Alfani 2020, p. 204) .

In view of all this, we can tentatively provide some speculative interpretations why a "net-worth helicopter money" strategy was first implemented and then abruptly reversed in

Venice during the 1629-31 shock. At the beginning of the shock, the ruling class (which held the overwhelming share of Venetian public debt) was apparently unwilling to see the funded debt increase, thus forcing the Treasury to monetize the politically inevitable stimulus. After realizing the substantial costs of excessive monetization, however, consensus was found among patricians on the need for the funded debt to grow in order to limit money depreciation. Contrary to the "monetary-base helicopter money" strategy, the "net-worth helicopter money" might have prevented the long-term government debt from growing in the short term, but the latter's impact on monetary stability was probably much more violent than 15 For a definition of "monetary externality", see the Appendix. the former's. As a result, the "net-worth helicopter money" strategy had to be quickly reversed and replaced by a much more conservative policy.

Recently, monetary policy theory highlighted the relevance of the way interventions are orchestrated between the central bank's and the government's balance sheets (Sims 2004; Reis 2015; Orphanides 2016; Benigno and Nisticò 2020) . In particular, radical options like the so-called "net-worth helicopter money" have been discussed as a possible strategy to cope with extraordinary macroeconomic shocks such as pandemics. However, episodes in which the "net-worth helicopter money" strategy has been actually implemented appear to be very rare in history. In this article, we have focused on one historical example in which a pandemic recession was actually addressed through the implementation of this strategy.

In the Republic of Venice, the 1629 famine and the 1630-1 plague caused a unique negative macroeconomic shock, that the oligarchic government addressed using a particularly radical form of fiscal monetization that corresponds to the modern notion of "net-worth helicopter money". As a matter of fact, the expansion of the bank of issue's liabilities was associated with a deterioration in the quality of its assets, thus producing capital losses to the issuer and reducing its net worth. This policy entailed monetary depreciation and instability, so that the government had to reverse it very quickly -i.e. before the end of the macroeconomic shock. Backtracking on the "net-worth helicopter money" strategy implied transferring the losses suffered on the monetary authority's balance sheet to the fiscal authority's balance sheet: this actually took place through a de facto government bailout of the bank of issue, as the bank's sight liabilities were converted into long-term public debt.

Why did the government choose to implement this suboptimal strategy? We argue that they did so to avoid political disturbances and popular rioting, while also preventing the long-term government debt from growing. Thus, in the short term, potential political and social consequences of the macroeconomic shock were minimized by implementing a redistributive policy from the rich to the poor. But in order to keep the monetization mechanism viable and prevent a complete debasement of bank money, the quality of the assets side of its balance sheet had to be restored: losses had to be covered through a de facto bailout, which made the inevitable growth of long-term government debt eventually occur.

into long-term debt actually reversed the distributional effects triggered by the "net-worth helicopter money" strategy: as taxes were strongly regressive in Venice, the ensuing conservative fiscal policy which allowed repaying the debt entailed a redistribution of resources from the poor to the rich.

The history of the Venetian reaction to the 1629-31 famine and pandemic echoes many aspects of the Covid-19 crisis. For one thing, it proves that nowadays' extraordinary fiscal expansion to cope with a pandemic were far from unprecedented. Moreover, it suggests that nowadays' central bankers' refusal to embark into some "net-worth helicopter money" experiment may have been a good idea after all. More generally, the Venetian experience with "net-worth helicopter money" highlights the redistributive implications of the design of macroeconomic policies, as well as the importance of political economy factors in the choices underlining such designs.

What happened in Venice in the period 1629-1631 can be also described using a model, using and modifying the theoretical settings introduced in Masciandaro and Passarelli (2019), Masciandaro (2020) and Favaretto and Masciandaro (2021) .

The economy consists of a population of inhabitants, and a policymaker that controls both fiscal and monetary policies. In the case of Venice, the Senate represents the incumbent policymaker. Initially we use the standard assumption that the policymaker is a benevolent player (Acemoglou et al. Argente et al. 2020; Brotherhood et al. 2020; Glover et al. 2020) .

This assumption will be modified later. All players are rational and share the same information;

they maximize their utility in a simultaneously and one-shot way. Moreover, the population size is one, then total and per-capita amounts are the same for all the variables.

Since the population size is one, maximizing the utility of the average representative inhabitant amounts to maximizing social welfare. Without macroeconomic shocks, fiscal monetization is not needed. In Venice in normal times the Republic produced systematic fiscal surpluses. If a pandemic and/or a famine occurs, the policymaker sets her economic policies to maximize social welfare. Venice was hit by an adverse macroeconomic shock from April 1629 -when a famine started -to September 1631 -when a bubonic plague ended.

A pandemic shock triggers a special recession, because inhabitants' incomes can be hit in an heterogeneous way for three reasons. First, the effects of both the disease and the corresponding public policies are unequally distributed (Glover 2020; Bloom et al. 2021 ).

Second, the less the policymaker is involved in supporting the economy during the pandemic recession through fiscal transfers, the more likely are negative second-round effects on the well-functioning of the economic and financial system after the pandemic (pandemic externalities) (Acharya and Steffen 2020; Anderson et al. 2020; Deb et al. 2020; Bloom et al. 2021 ). Third, the fiscal policy financing introduces the possibility of monetary stability risks (monetary externalities). Therefore the policymaker will choice its economy policy design maximizing a welfare function with three terms (Masciandaro and Passarelli 2019; Masciandaro 2020) :

Where U(β,δ,τ), F(β) and M(β,δ) are respectively the inhabitants' utility, the pandemic externalities and the monetary externalities, while τ, β, and δ represent the key economy policy variables: taxation, fiscal spending and fiscal monetization. Then the first step is to explore how inhabitants behave.

We assume that inhabitants are risk neutral, and they maximize their overall utility from consumption and disutility from effort. In our setting inhabitants' utility is associated with Inhabitants expect that when a pandemic occurs their incomes can be hit.

Containment measures save lives, but in parallel impose limitations on several economic activities. People suffer because lockdown measures and quarantines reduce their incomes and expenditures (Baker et al. 2020a; Carvalho et al. 2020; Cox et al. 2020) . At the same time, inhabitants expect that the policymaker will help those of them who are suffering with an injection of a lump-sum fiscal transfer to mitigate the pandemic costs (Acemoglou et al. Argente et al. 2020; Brotherhood et al. 2020; Glover et al. 2020) .

As much as inhabitants' losses due to the pandemic can be heterogenous, the same will be true for the distribution of the fiscal transfers (Bayer et al. 2020; Glover et al. 2020) .

The pandemic shock and the consequent fiscal transfer policy influence inhabitants' welfare in an unequal way, producing a special case of income heterogeneity (Auerbach et al Bayer et al. 2020; Gertler et al. 2020; Glover et al. 2020; Kaplan et al. 2020) .

Given the risk that an activity is frozen during a pandemic due to the containment policies -pandemic or "quarantine shock" (Bayer et al. 2020) -we can distinguish between safe activities and risky activities. Occupations that do not suffer losses during a pandemic produce safe incomes, while risky earnings are associated with activities that are negatively influenced when a pandemic happens (De Vito and Gomez 2020; Elenev et al. 2020) . Safe incomes can be taxed. As extreme examples of the two situations during the 1630 pandemic, consider the case of the second-hand clothiers as inhabitants involved in risky activities, being systematically hit by blockades and closures, were the Arsenal workers were individuals with a safe occupation, having their salaries in any case, due to the Senate subsidies.

Therefore the representative unhabitant's utility is:

(1 − ) − ( ) + ( )

Assuming a normalized productivity to one, the first term is the after-tax income, while the second term is a standard increasing and convex effort function. Given taxation, the inhabitant chooses her optimal effort. The optimal condition yields inhabitants' labour supply, decreasing in the taxation, and this rate represents her elasticity to tax distortion. Since both population size and productivity are equal to one, the labour supply represents the total safe income.

The third term represents the risky earning utility. There is only one risky activity, and it is measured using the variable r, which parameterizes the risk that the inhabitants bear. The return of the risky activity is θ(r), with θ'(r)>0 and θ''(r)<0. We normalize this return to one. If a pandemic occurs, with probability p, lockdown and quarantine are implemented. The inhabitants know that the value of the risky activity will fall -for simplicity to zero -and they will bear the full cost. At the same time, the inhabitants expect that the policymaker will design a fiscal transfer policy to address their losses, in a proportion β of such losses, and with a corresponding monetization δ.

Safe incomes and risky earnings finance consumption. The overall budget constraint of the representative average inhabitant will be:

With risky earnings being the only source of heterogeneity among inhabitants, such a metric allows us to highlight in the clearest and simplest way the redistributive effects of the fiscal transfer and its monetary financing. Given that the economy policy design influences the inhabitants' welfare, now we turn again our attention to the policymaker choices.

When a pandemic breaks out, the policymaker designs a containment policy, facing an unpleasant dilemma between two public goals (Baldwin and Weder di Mauro 2020) . The policymaker needs to protect public health by implementing a containment policy with the aim of minimizing the expected loss of life. In parallel, containment policies save lives, but, given the interactions between economic decisions and epidemics (Eichenbaum et al. 2020) ,

any containment policy has short-term economic and financial costs (Deb et al. 2020; Ludvigson et al. 2020) . These costs simultaneously affect aggregate supply (Del Rio-Chanona et al. 2020; Koren and Peto 2020) and aggregate demand (Andersen et al. 2020; Del Rio-Chanona et al. 2020) .

The policymaker can address the pandemic recession by implementing an extraordinary fiscal transfer policy using a lump-sum distribution (Bloom et al. 2021; Glover et al. 2020) , with the aim of mitigating the negative effects of containment measures (Beck 2020; Bénassy-Quéré et al. 2020; Brunnermeier et al. 2020; Deb et al. 2020; Drechsel and Kalemli-Ozcan 2020; Gros 2020a; Kahn and Wagner 2020; Segura and Villacorta 2020) . The fiscal transfer can come in many forms, as income subsidies, work insurance, equity injections, loan guarantees (Céspedes et al. 2020; Didier et al. 2020; Elenev et al. 2020) . Transfer payments can be unconditional (Kubota et al. 2020) and/or conditional on the inhabitant's status (unemployed and/or liquidity or credit constrained individuals) (Bayer et al. 2020) . In Venice the Senate systematically implemented fiscal transfer policies during the pandemic recessions.

Then two questions arise.

First, how large should this fiscal policy be? Two opposite options arise. At one extreme, the policymaker is completely absent, and inhabitants suffered income losses. At the other extreme, the fiscal expansion helps the suffering inhabitants. The policymaker injects resources in the economy, and the metric of this fiscal action is a proportion, β∈(0,1) of the inhabitants' losses, which is the policy variable that parameterizes fiscal policy.

Second, how can such a fiscal policy be financed? The policymaker can raise taxation (Bloom et al. 2021; Eichenbaum et al. 2020) , issue debt or money. The policymaker defines the optimal fiscal transfer policy,  *, and this policy can be financed by issuing new debt, charging a regressive lump-sum tax  on the safe income for servicing the issued debt, and through monetization. Assuming no default risk, the policymaker budget constraint is:

(1 + (1 − )) =

where  is the lump-sum tax, l is the safe income of the inhabitants before taxes, i is the interest paid on the public bond and  is the fiscal monetization where   1 , 0   . For any unit of debt issued, the policymaker repays

negatively associated with the degree of fiscal monetization. When a higher monetization is implemented (i.e. higher  ), a lower portion of funded debt will be sold to inhabitants.

Therefore fiscal transfer and its monetization influence inhabitants' consumption. But the pandemic policy can produce long standing effects on inhabitants' welfare. So we introduce the possibility of monetary and pandemic externalities, internalizing future negative spillovers due to the economic policy action that can affect the economy when the pandemic ends.

First, fiscal monetization is not a free lunch: it may create monetary externalities.

Monetary externalities can depend on the association between central bank seigniorage and monetary stability risks. The more traditional channel is the relationship between seigniorage and inflation tax (Buiter 2007) , that increases both national inflation (Friedman 1969; Aizenman 1992 ) and, via exchange rate devaluation, international inflation (Hamada 1976) .

Moreover, monetary externalities can also include banking (Bianchi, 2010) and financial (Stein 2012; Cesa Bianchi and Rebucci 2017) imbalances, or more generally it is a device to take into account the risk of monetary policy multiple equilibria and their costs (Gliksberg 2009; Airaudo and Bossi 2017) . 

The monetary externality aversion -i.e. the parameter Φ -is homogenous among inhabitants. With this assumption it will be evident that it is sufficient to have just two sources of heterogeneity among inhabitants -fiscal transfer and its bond financing -to have a multiple equilibria setting in terms of political consensus. With further heterogeneity sources the results should be even stronger.

Second, the less the policymaker is involved in supporting the economy, the more likely are negative second-round effects on the well-functioning of the economic and financial system after the pandemic (Acharya and Steffen 2020; Anderson et al. 2020; Deb et al. 2020) .

The absence of active public policies can have adverse economic effects that spread out over time and into the longer term, as a reduction in return to human capital or negative structural changes in terms of trading patterns and stalling development (Bloom et al. 2021) . On this respect, and notwithstanding the public action, the 1630 pandemic recession was a crucial negative turning point in the history of the Most Serene Republic.

To capture in the simplest way this channel, let the pandemic externality function be:

The pandemic externalities are increasing and convex in the amount of losses, and they are lower the higher are the fiscal transfers,  . Also the pandemic externality aversionthe parameter ε -is homogenous among inhabitants, for the same motivations above expressed.

Inhabitants and policymaker simultaneously optimize their choices. The average representative inhabitant, optimizing the goal function (2), and given her elasticity to tax distortion, η, identifies the optimal effort, l* and the optimal risk assumption, θ*. The corresponding safe incomes and risk earnings finance consumption. This assumption is particularly relevant during a pandemic: shutdowns and quarantines lockdowns produce material deprivation and households can draw on all their net available resources to address the shock (Baker et al. 2020a; Carvalho et al. 2020; Cox et al. 2020) . Moreover the available resources depend on the design of taxation, fiscal transfer and income from bond financing.

The policymaker maximizes the social-welfare function (1) (Sargent and Wallace 1981) : monetary policy is not independent. Focusing on the optimal level of fiscal monetization, *  , its social optimal value is:

The optimal level of monetization, *  , will increase: a) the more taxation is distortionary; b) the more the cost of debt servicing is high; c) the more monetary externality aversion is low.

Given the decision in terms of fiscal monetization, the final step is its implementation.

Here a central bank -or a state bank issuer, the Venice Giro Bank, in the case of Venice -comes in as a public institution with its goal, which comes from somewhere or someone (Reis 2013) - consensus in favour of containment policies is to use fiscal retributive policies to reduce the costs to those whose resources are threatened by shutdowns and quarantines (Glover 2020) .

Therefore, we need to explore the inhabitants' preferences regarding the policy mix designed by the policymaker. Two different dimensions are relevant: inhabitants can be or not subsidy recipient individuals, and/or they can be or not monetization prone agents.

Which are the subsidized inhabitants? Let us consider any inhabitant j, being + the amount of risky earnings in her balance sheet. With > 0 inhabitant j will be a subsidized inhabitant relative to the average inhabitant (subsidization gain) . Let ( ) be the distribution of the subsidized inhabitants across the population. With risky earnings being a proxy for the fiscal transfer, these resources in the balance sheet of the median inhabitant tell us whether the subsidized inhabitants represent the majority or a minority of the population.

What about monetization propension? Inhabitants can be heterogeneous also as funded debt holders. In this case, the more a inhabitant j is a debt holder, the more she will be monetization adverse, given that more monetization implies lower interest rates. Let 

where the last two terms on the right-hand side account for the two forms of heterogeneity of individual j relative to the average inhabitant. Each inhabitant's preferences can differ from those of the benevolent policymaker because of these two terms. Focusing on monetization preferences, the optimal fiscal monetization for inhabitant j is:

Assuming equation (10) holds as an equality, solving it yields:

By comparing equation (7) with the social optimal monetization (11), it is evident that, given a fiscal policy 0   , a political distortion can arise between a inhabitant's preferred policy and the social optimal monetization: − * = −

The political distortion ^*   − will reflect inhabitants' preferences. The direction of the political pressure depends on who the median inhabitant is. For example, we can assume that financial asset/wealth holdings are very skewed, concentrated among a small segment, the rich, of the population. Therefore monetary policy, influencing asset returns, produces redistributive effects that benefit the holders of such as assets (Krishnamurthy and Vissing-Jorgensen 2011; Brunnemeier and Sannikov 2014) .

We interpreted the monetary stance in Venice in early modern times as an extraordinary "hard helicopter money" with redistributive effects. Then a question arises:

which inhabitants like fiscal monetization? Among all the possible equilibria, all Venetian subsidized inhabitants like helicopter money, but when they were monetization adverse individuals. In parallel, all monetization prone inhabitants like helicopter money, but when they are not subsidized inhabitants. In other words, uncertainty is present in cases without a clear-cut net benefit, as subsidized merchants that are also bond holders.

But how relevant are the median inhabitant's preferences for the incumbent policymaker? Taking inspiration from Passarelli and Tabellini (2017) and Favaretto and Masciandaro (2021) , we assume that the monetary policy decisions are associated with political consensus, because consensus depends on the median inhabitant's preferences through economic and psychological group-thinking mechanisms. The risks of political unrest can influence incumbent policymakers, and these risks can be motivated by facts and emotions. If the policymaker considers the median inhabitant's preferences as a relevant proxy for riot risks, political pressures may be relevant in shaping fiscal monetization choices. The link between inhabitants' preferences, political pressure and political choices can emerge also in an oligarchy of merchants, as the Venice Republic was at that time. In fact, ordinary Venetians used collective actions to influence patrician choices, especially during crisis periods.

This situation is captured in the simplest way assuming that the actual monetary policy decision is such that:

where 0 < < 1 represents the relevance of the political pressure. In Venice, the 1630 extraordinary monetization over-expansion created inflation and currency depreciation. In parallel, the population expected "whatever it takes" myopic fiscal policies, and the politicians tended to please inhabitants' preferences, given the threat of riots. Political pressure and helicopter money were likely to be two sides of the same coin.

The analysis can be enriched in several directions: a) Monetary externality sensibility and inhabitants' heterogeneity: monetary instability is assumed to be an homogeneous social cost. But inhabitants can be heterogeneous in their ability to address such risks through hedging, with some individuals facing -or feeling that they face -higher costs due to monetary instability (i.e. inflation-adverse citizens). In other words, we could explicitly take into account the redistributive effect of inflation, that have long been recognized in the traditional literature (Keynes 1923; Bresciani-Turroni 1937; Friedman and Schwartz 1963) , and that has been discussed again recently (Doepke and Schneider 2006; Colbion et al. 2012) . Also in early modern Venice the ruling elites constantly demanded a stable currency (Al-Bawwab 2021). Allowing for this kind of heterogeneity would lead to a straightforward prediction: the smaller the mass of inflation risk-adverse citizens, the stronger the political pressure to engage in fiscal monetization.

Moreover, we could introduce heterogeneity in the propensity to consume, that can influence the effect of the fiscal transfer in stimulating consumption (Andreolli and Surico 2021). This change could help to explain the empirical estimates of the marginal consumption propensity during a pandemic (Baker et al. 2020b; Chetty et al. 2020; Coibion et al. 2020; Karger and Rajan 2020; Kim and Lee 2020; Kubota et al. 2020) . Finally, we can assume heterogeneity in inhabitants' marginal propensity to take risk (Kekre and Lenel 2021) , and also this heterogeneity can influence the distribution of the fiscal transfer effect. b) Taxation and inhabitants' heterogeneity: safe income taxation has been assumed to be the same for all individuals. In the presence of taxation heterogeneity, the distributional effects are likely to increase. For example, given the decisions regarding the fiscal policy and its monetization, if richer inhabitants are likely to have higher tax burden, all else equal, they would prefer lighter fiscal policies. The income and/or taxation heterogeneity can be relevant in strengthening or weakening political pressure in favour or against the fiscal monetization. c) Public debt and interest rates: public debt is only issued to address the pandemicrelated recession and the interest-rate level remained constant. Assuming an initial debt, or interest rate endogeneity depending on the debt stock, would exacerbate the policy trade-offs and, consequently, the relevance of the political distortions.

Una finanza in guerra

L'Inestinguibile Sogno del Dominio: Francesco Morosini. Venice: Istituto Veneto di Scienze

Le grandi pesti dell'età moderna: 1575-77 e 1630-31

Poverty, charity and the reason of state: some Venetian examples

Bollettino dell'Istituto di Storia della Società e dello Stato Veneziano

Wage-earners and the Venetian economy

Industry and Economic Decline in Seventeenth-Century Venice

When does a central bank's balance sheet require fiscal support

Early public banks

Il Banco Giro di Venezia: a proposito di alcune recenti pubblicazioni

Fiscal aspects of central bank independence

European Monetary Integration

The political economy of private paper money: institutional development in Europe up to 1800

The disappearance of deposit banks: an explanation

Il Banco Giro di Venezia

The Venetian mint in the age of the Black Death

Material Culture and Cultural Materialism in the Middle Ages and Renaissance

Andamento demografico della nobilità veneziana allo specchio delle votazioni nel Maggior Consiglio (1297-1797)

The Evolution of Central Banking: Theory and History

The people's prince: popular politics in early modern Venice

Optimal targeted lockdowns in a multi-group SIR model

Stress tests for banks as liquidity insurers in a time of Covid. VoxEu

Consumption externalities and monetary policy with limited asset market participation

Competitive externalities and the optimal seigniorage

The Zecca mint: a self-enforcing monetary constitution in historic Venice

Bureaucrats or politicians? Part I: a single policy task

Consumer responses to the Covid-19 crisis: evidence from bank account transaction data

A simple planning problem for Covid-19 lockdown and smart tracing

How does household spending respond to an epidemic? Consumption during the 2020 Covid-19 pandemic

Mitigating the COVID Economic Crisis: Act Fast and Do Whatever It Takes

Dealing with the Next Downturn: From Unconventional Monetary Policy to Unprecedented Policy Coordination

It's all in the mix: how monetary and fiscal policies can work or fail together

The Coronavirus stimulus package: how large is the transfer multiplier? CEPR Discussion Paper Series n

Finance in the times of Covid-19: what next

Mitigating the COVID Economic Crisis: Act Fast and Do Whatever It Takes

Covid-19: a proposal for a Covid credit line

Non-neutrality of open market operations

Some thoughts on monetary policy in Japan

What tools does the Fed have left? Part 3: helicopter money

Brookings Institution

Credit externalities: macroeconomic effects and policy implications

Modern infectious diseases: macroeconomic impacts and policy responses

Helicopter money: the illusion of a free lunch

The Economics of Inflation

An economic model of the Covid-19 pandemic with young and old agents: behavior, testing and policies

Covid-19: inflation and deflation pressures

A macroeconomic model with financial sector

Seigniorage. NBER Working Paper Series n. 12919

Tracking the Covid-19 crisis with high-resolution transaction data

How do central banks control inflation? A guide for the perplexed

Does easing monetary policy increase financial instability

The macroeconomic of a pandemic: a minimalist model

Optimal fiscal and monetary policy

How did Covid-19 and stabilization policies affect spending and employment? NBER Working

How did consumers use their stimulus payments?

Initial impact of the pandemic on consumer behaviour: evidence from linked income, spending, and savings data

Covid-19: preventing a corporate cash crunch among listed firms

Supply and semand shocks in the Covid-19 pandemic: an industry and occupation perspective

Fiscal shocks and helicopter money in open economy

Financing firms in hibernation during the Covid-19 pandemic

Inflation and the redistribution of nominal wealth

Are standard macro policies enough to deal with the economic fallout from a global pandemic? Economics for Inclusive Prosperity Policy Brief

The macroeconomics of epidemics

Can the Covid bailouts save the economy? CEPR Discussion Papers n

Populism, financial crises and banking policies: economics and psychology

The effects of a money-financed fiscal stimulus

A macroeconomic model with financial panics

Monetary policy and multiple equilibria with constrained investment and externalities

Health versus wealth: on the distributional effects of controlling a pandemic

A strategic analysis of monetary interdependence

Liquidity provision during a pandemic

The great lockdown and the big stimulus

Heterogeneity in the marginal propensity to consume: evidence from Covid-19 stimulus payments. Federal Reserve Bank of Chicago Working

Monetary policy, redistribution, and risk premia

A Tract on Monetary Reform

Can stimulus checks boost an economy under Covid-19? Evidence from South Korea

Business disruptions from social distancing

The aggregate demand for Treasury debt

Consumption responses to Covid-19 payments: evidence from a natural experiment and bank account data

Covid-19 and the macroeconomic effects of costly disasters

Populism, political pressure and central bank (in)dependence

Covid-19 helicopter money: economics and politics

Combatting Eurozone deflation. VoxEu

Eurozone recovery: there are no shortcuts. VoxEu

Is a money-financed stimulus desirable? Working paper

Central bank design

The fiscal footprint of macroprudential policy

Some unpleasant monetarist arithmetic

Federal Reserve Bank of Minneapolis Quarterly Review

Policies to support firms in a lockdown: a pecking order

Monetary policy as financial stability regulation

Debt, money and Mephistopheles

Methods of policy accommodation at the interest-rate lower bound