In a very precise paper in Investigación Económica, Thomas ^{Palley (2021)} makes a critical assessment of New Developmentalism. Our first reaction was one of incredulity, since Palley was an early supporter of New Developmentalism: On 29 September 2010 he signed the “Ten theses on New Developmentalism” (see https://www.scielo.br/j/rep/a/CFZ8xg7rqKrBGcF5fxmSy5H/?lang=en), a list of 10 core principles of New Developmentalism agreed with all those who signed the manifesto. In his new article, Palley criticizes many of the principles he had previously agreed with. To paraphrase John Maynard Keynes, you are always entitled to change your mind when in error or when the facts change. In either case, Palley should begin his critical assessment by explaining why he has changed his mind about New Developmentalism.

Our first general comment on his assessment is that, like others critical of New Developmentalism, Palley identifies the whole school of thought with only certain references: He ignores important material, such as the 2015 textbook on New Developmentalism (^{Bresser-Pereira, Oreiro, and Marconi, 2015}) and the fact that New Developmentalism is not some kind of religion, with one single view of the “truth”. Indeed, one of the authors of this paper has disagreed with some of Professor Bresser-Pereira’s views, mainly regarding the industrial equilibrium exchange rate concept, as can be seen in a recent article published in the Brazilian Journal of Political Economy (^{Oreiro, 2020}).

A second general comment is that, unlike ^{Palley (2021, p. 30)}, we do not understand New Developmentalism to be a new theory of economic development, but rather a synthesis of classical theory of economic development, Latin American structuralism and Kaldorian theory of demand-led growth (^{Oreiro, Martins da Silva, and Dávila-Fernandez, 2020, pp. 27-29}), as regards the main drivers and determinants of economic development. We see no opposition between Classical Developmentalism (CD) and New Developmentalism, but recognize that these theories are designed for different historical stages of Latin American countries’ development. While the basic issue for Classical Developmentalism was how to overcome the poverty trap by means of industrialization-led structural change starting from an infant industry, the basic issue for New Developmentalism is how to overcome the middle-income trap in semi-mature economies by introducing a new macroeconomic policy regime (together with industrial policy) to replace the new macroeconomic consensus that has prevailed in both developed and developing countries since the early 1990s. Accordingly, Palley’s claim that New Developmentalism somehow compromises with Neo-Liberalism seems unfair.

New Developmentalism at its earlier stages (^{Bresser-Pereira, 2006}, ²⁰⁰⁷, ²⁰⁰⁹) took it for granted that Latin-American countries had already surpassed the infant-industry phase of their economic development, which meant that further development of manufacturing industry called for a shift from “Import Substitution Industrialization” to “Export Promotion of Manufacturing Goods” (and sophisticated services). That proposal is very similar to the one advanced by ^{Kaldor (1967)}, to whom the most successful cases of industrialization in world history were precisely ones where countries increased their share in world manufactured goods exports, as was the case with the United States, Germany, Japan and Italy (nowadays, China and South Korea). This is the fundamental difference between New Developmentalism and Classical Developmentalism.

Moreover, the world where the Classical Theory of Economic Development (1950s and 60s) arose was very different from the one where New Developmentalism has emerged (early twenty-first century). Classical Developmentalism was formulated, and implemented in Latin America, under the aegis of the Bretton Woods System, which provided policy space for developmental policies. Since then, globalization and financial liberalization have made it much more complicated to adopt the tariff protection and multiple exchange rate instruments widely used by Latin American countries to boost industrialization from the 1950s to the 1970s. New financial products have enabled commodity price booms to translate into capital inflows to countries that are rich in natural resources, leading to real exchange rate overvaluation from purely financial causes (See ^{Nalin and Yazima, 2021}; ^{Botta, 2015}). In the same vein, ^{Akyuz (2020)} shows that, in the global financial cycle, capital flows are pro-cyclical and correlate strongly with commodities prices listed on the international market, in what he denominated the “commodity-finance nexus”. Indeed, this means that the severity of “Dutch Disease” is directly tied to financial liberalization, a fundamental aspect of neo-liberal policy.

There are two parts to Palley’s critical assessment. In the first, he gives a reasonably fair account of the basic principles of Bresser-Pereira’s version of New Developmentalism, as presented in some of his recent papers. However, he completely disregards Bresser-Pereira’s earlier writings or those of other members of the New Developmentalist School. The second part of the paper, with its critique of New-Developmentalism (ND), seems highly unfair not only to Bresser-Pereira, but to the whole school of thought. As argued here, Palley’s paper makes mistaken claims about New Developmentalism’s internal logic and its relationship to Neo-Liberalism. These misconceived claims cannot be attributed to “ND’s failure to model the economy analytically”, because new developmentalist principles have inspired the development of many mathematical models in recent years (^{Gabriel, Jayme Jr, and Oreiro, 2016}; ^{Santana and Oreiro, 2018}; ^{Oreiro, Martins da Silva, and Dávila-Fernandez, 2020}).

^{Palley (2021)} argues that ND has four main components and challenges: The Dutch disease problem; its criticism of the growth-cum-foreign savings strategy; the need to develop a manufacturing sector that was internationally competitive and technologically advanced; and getting macroeconomic prices right.

Palley’s account of Dutch Disease is very similar to the arguments presented by ^{Diamand (1972)}. Natural resource-rich countries had an unbalanced productive structure where primary goods production and exports are competitive on international markets at a higher (real) exchange rate than manufactured goods. In Bresser-Pereira’s framework, this means that there are two equilibrium exchange rates: The first (e2 in Figure 2 of Palley’s article) is the industrial equilibrium exchange rate, the (real) exchange rate at which domestic manufacturing firms operating with state-of-the-art technology are competitive on international markets, while the second, the current account equilibrium exchange rate, is the (real) exchange rate compatible with a zero long-run account deficit (or zero foreign saving). For countries with a closed capital account, the current equilibrium exchange rate is real exchange rate at which the trade account is balanced (e3 in Figure 2 of Palley’s paper). In this case, the actual real exchange rate will fluctuate around the current account equilibrium level under the influence of foreign currency demand and supply forces on the foreign exchange market, in the long run generating an overvalued exchange rate for the manufacturing industry. In this case, the economy suffers from Dutch Disease (DD).

For a country with a closed capital account, DD is relatively simple to deal with. The domestic monetary authority can intervene in foreign exchange markets, buying international reserves so as to hold the real exchange rate at an undervalued level and hence achieving a trade account surplus. In this case, the real exchange rate can be adjusted to a level that is compatible with the industrial equilibrium rate, but also extremely profitable for primary goods exports. Consequently, the profit rate from primary goods production and exports will be much higher than from the manufacturing sector, thus causing real resources to flow from the manufacturing sector to primary goods production, increasing primary exports in the long-term and hence aggravating the DD problem. Here is where export taxes on primary goods are important: An export tax will reduce primary sector profits to more “normal” values, thus reducing incentives for private investment and greater production in the primary sector which result from maintaining an undervalued exchange rate (and which otherwise would aggravate the DD problem) and hence increasing government revenue from commodity exports. In other words, export taxes, by reducing the (post-tax) profitability of producing primary goods for export, discourage investment in that sector.

Palley criticizes export taxes, arguing that such taxation “[…] redistributes rents from the primary sector to the State. Export volumes are unaffected, and so is trade balance” (^{Palley, 2021, p. 16}). As we have already argued, in the absence of such taxes, if the monetary authority attempts to achieve a more competitive exchange rate for manufacturing activities by accumulating reserves, producers of primary goods will obtain supernormal profits, making the attempts to neutralize DD self-defeating in the long run.

However, if the economy is operating at near full capacity utilization, devaluation of the nominal exchange rate as a result of foreign reserve accumulation can produce inflationary pressures due to excess aggregate demand. That case is not considered by Palley, because his assessment of ND considers prices to be constant (^{2021, p. 6}). In a full-employment scenario, domestic price increases can prevent nominal exchange rate depreciation from transforming into real exchange rate depreciation. That is why, in a full-employment environment, government should pursue contractionary fiscal policy in order to neutralize DD: If government saves all export tax revenues in a form a sovereign fund (^{Bresser-Pereira, Oreiro, and Marconi, 2015, p. 146}), then redistribution of income from primary goods producers to the State will increase the average savings rate in the economy (supposing that the rate of savings among primary sector entrepreneurs is considerably lower than one), influencing the trade balance positively without inflationary effects. Spending those revenues on government consumption would reduce the average saving, increasing the inflationary effects of nominal exchange rate devaluation.

In an economy with an open capital account, the situation is a little more complicated. First of all, the current equilibrium exchange rate now requires a trade surplus large enough to balance the current account (supposing a positive external debt). This means that the current equilibrium exchange rate will be at the left of e3 in Figure 2 of Palley’s article, meaning that DD will be smaller than in the case of zero capital mobility. On the other hand, access to world financial markets enables a country to run up current account deficits with excess of foreign capital inflows. It is here that the model of growth cum foreign savings enters the picture. Neo-liberalism, grounded in traditional neoclassical economics, considers domestic and foreign savings to be complementary (rather than substitutes, as argued by ND). That differentiation is missing from Palley’s review. Accordingly, a neo-liberal policymaker would be likely to encourage capital inflows by setting the domestic interest rate higher than the external equilibrium level (given by the sum of the international interest rate and country risk premium). That policy would also allow the policymaker to achieve a low rate of inflation (see ^{Oreiro, Martins da Silva, and Dávila-Fernandez, 2020, p. 33}). Capital inflows, an autonomous source of exchange rate appreciation, cause current account deficits. It is important to stress that, to ND, growth cum foreign savings is not a market result, but a policy choice: Policymakers choose to set the domestic interest rate at a higher than equilibrium level in order to (i) increase foreign saving in a (self-defeating) attempt to increase aggregate saving and hence investment and growth¹; (ii) achieve a low inflation rate compatible with the inflation target set by the monetary authorities (see ^{Bresser-Pereira and Gala, 2007}). That choice (to liberalize the capital account) -supported by the exchange rate populism endemic to Latin American countries, under left- or right-wing governments- is guided by neo-liberal ideology. It, therefore, makes no sense to associate ND with that ideology.

Solving the problem of growth cum foreign saving thus involves fundamentally setting the domestic interest rate at the right level, as given by the sum of the international interest rate, plus the country risk premium, so as to attract less of the capital inflows that tend to result in domestic currency appreciation. Palley is thus misleading when he writes that “[ND] has no policy prescription for interest rates in the form of an interest rule or interest rate target” ^{(2021, p. 14)}. For more on this issue see ^{Santana and Oreiro (2018)} and ^{Oreiro et al. (2021)}.

On the other hand, the nature of peripheral economies’ external vulnerability has changed since the end of the Bretton Woods era. Although current account, and especially terms-of-trade, shocks have remained significant (particularly in commodity-dependent economies), the monetary and financial dimensions have gained increasing weight in the relationship between center and periphery (^{Ocampo, 2001}). Monetary asymmetry or hierarchy² is intertwined with the financial dimension of global asymmetries. While monetary asymmetry comprises the negative consequences for peripheral economies of their inability to borrow abroad in their own currency, financial globalization relates to the magnitude and patterns of international capital flows to peripheral economies. Capital flows towards peripheral emerging economies mainly depend mainly on exogenous sources (^{Rey, 2015}), which renders them permanently vulnerable to reversal by changes in monetary conditions in advanced countries, as well as to increased risk aversion of global investors. In that setting, international financial markets are highly volatile, resulting in boom-bust cycles. To be successful, any developmentalist strategy has to address monetary and financial asymmetry issues, a problem that New Developmentalism seeks to solve and that Palley and social-developmentalists tend to neglect (see ^{Fritz, Paula, and Prates, 2022}). The new developmentalist approach has a clear and well-developed strategy, which focuses on shielding the economy from external shocks.

Palley correctly identifies exports as the driver of economic growth in the ND growth model. As already said at the beginning of this note, ND is based on Kaldor’s demand-led growth model, where export demand is the only long-run source of autonomous demand growth. Palley, however, argues, according to the so-called supermultiplier model, that government spending can be an important source of autonomous demand (^{Palley, 2021, p. 28}). ND strongly rejects not only government spending, but domestic demand generally, as the main (or only) driver of autonomous demand growth in middle-income countries in their “catching-up” strategy, because (i) such countries -unlike the USA- have no reserve currency from which to finance balance of payments deficits for indefinitely long periods (what former French President Valéry Giscard d’Estaing called the “exorbitant privilege”); and (ii) import income elasticity is generally greater than one. This means that simply in order to meet Thirlwall’s balance of payments constraint, exports must grow faster than domestic output, meaning that growth must be export-led in order to be sustainable in the long-run (^{Thirlwall and Dixon, 1979, p. 174}).

Apart from these general comments, we have some specific criticisms of Palley’s article. On page 21, he argues that “The rate of accumulation then depends on the difference between the expected profit rate and the interest rate […]. That is a substantially Neoliberal view of the accumulation process, and it contrasts with the CD view in which the state occupies a far more activist position”. In the standard ND model (^{Bresser-Pereira, Oreiro and Marconi, 2014, p. 66}), the investment function is the Neo-Robinsonian equation for desired rate of capital accumulation (^{Blecker and Setterfield, 2019, p. 136}). It thus makes no sense to associate specification of the investment function with Neo-Liberalism of any kind.

Palley also criticizes ND’s emphasis on industrialization as the engine of growth. Although Palley is right that the trend in high-income countries in the past three decades has been to de-industrialize ^{(2021, p. 290)}, he ignores the problem of premature de-industrialization, defined as a reduction of the share of manufacturing in added value and employment in economies that have not yet reached the “Lewis turning point”, that is, the situation in which there is an unlimited supply of labor for the modern sector of the economy -which is the focus of ND’s concern regarding middle-income countries. ^{Rodrik (2016)}, who has shown public sympathies with some ND ideas, explains that manufacturing tends to experience relatively stronger productivity growth and technological progress over the medium to long term. That given, premature de-industrialization closes off the main avenue for achieving fast economic convergence in low- and middle-income countries. It was the industrialization process that permitted non-Western nations (Japan in the late nineteenth century, South Korea, Taiwan and others, in the twentieth century, and China in the twenty-first century) to catch up and converge with the West³. There thus seems to be no basis for Palley’s claim that “Those empirical facts cast doubt on ND’s framing of the development solution in terms of industrialization” (^{Palley, 2021, p. 21}), a statement that can be applied to advanced economies.

Another unfair criticism by Palley regards the role of public investment in economic development. There is nothing in the ND literature that denies the important role of investment in infrastructure for economic development. Indeed, ^{Bresser-Pereira and Oreiro (2010)} advocated separating the fiscal budget into current and capital accounts, arguing that government must run up capital budget deficits in order to finance a sustainable increase in public investment (see also, ^{Paula, 2008}). To ND, sound fiscal policy is policy that allows public investment to increase without an unsustainable increase in public debt as a proportion of Gross Domestic Product (GDP). To achieve that goal, many fiscal rules can be applied, though, from a cyclically-adjusted budget target for a fiscal primary surplus to an overall government target for a current account surplus, which excludes government investments, but includes interest payments on public debt.

We also disagree with Palley’s support for the Dilma administration’s budget deficits. As ^{Oreiro and D’Agostini (2017)} have shown, during Dilma Rouseff’s first term, the Brazilian economy was suffering not from a problem of insufficient aggregate demand, but from loss of economic dynamism due to premature deindustrialization. In such a setting, increasing aggregate demand by fiscal stimulus only increases imports, and has little impact on economic growth. One of the main problems of Roussef’s first term was the frequent changes in economic policy, which was sometimes contractionary, sometimes expansionary, and mainly lacked coordination in its countercyclical economic policies. For instance, from 2011 to 2015, public investments in Brazil grew at an average rate of -5.2%⁴ (^{Orair, 2016, p. 16}), while fiscal policy was expansionary over most of the period (with some exceptions), in part due to tax relief for industries, with dubious effects on economic growth.

Regarding the nature of the demand regime in the Brazilian economy, Oreiro and Araujo (2013) used a neo-Kaleckian growth and distribution model to show that the demand regime is dependent upon the state of real exchange rate misalignment. In periods when the exchange rate is over-valued, the demand regime is profit-led; and in the periods of undervalued exchange rate, the demand regime is wage-led. Since Brazil has tended to over-valued exchange rates in the last 20 years, the demand regime in the Brazilian economy is predominantly profit-led.

We have one last criticism of Palley’s arguments. On page 27, he writes that more egalitarian distribution requires a higher wage-share, which is not exactly or necessarily true. If a high wage share is associated with greater wage dispersion (for example, by a higher ratio between average and median wages), then it is possible for income distribution to be more concentrated than in an alternative scenario where the wage share is lower, but the ratio of average to median wage is also lower. It is important to point out that more egalitarian income distribution can be achieved by means of tax reforms that increase the burden of income taxes relative to indirect taxes in total government revenues, and with higher tax rates for higher-income groups. Personal income distribution can be dramatically changed without major changes in the functional distribution of income.

Finally, we do agree with Palley’s critical assessment of ND on one point: ND certainly has underestimated the role of industrial policies in economic development. This was due to a theoretical and case-specific explanation. The theoretical explanation involves the industrial equilibrium exchange rate concept employed by Bresser-Pereira in his writings. For Bresser-Pereira, the industrial equilibrium exchange rate is the real exchange rate level that enables firms using state-of-the-art technology to be competitive on international markets. The problem is that, in developing countries, most firms in the manufacturing sector operate short of the technological frontier. This makes it necessary to redefine the concept of industrial equilibrium exchange rate as the real exchange rate that, for a given technology gap, will keep the manufacturing share of output constant over time (^{Oreiro, Martins da Silva, and Dávila-Fernandez, 2020}; ^{Oreiro, D’Agostini, and Gala, 2020}). Once industrial equilibrium is defined this way, there is a role for industrial policies in economic development. That role is precisely to reduce the technology gap to allow the industrial equilibrium exchange rate to appreciate without jeopardizing manufacturing firms’ price-competitiveness, thus enabling the real wage rate to rise sustainably.

Brazil’s industrial policy experience during the Lula and Dilma Rouseff administrations was very far from successful, however. On the contrary, the policies implemented from 2003 to 2015 were unable to prevent premature deindustrialization in the Brazilian economy and the return to primary goods as its main exports, a process termed the “reprimarization” of exports. To ND, the failure of Brazil’s industrial policies from 2003 to 2016 was proof that industrial policies can never be considered a substitute for a competitive real exchange rate, although this does appear to be the thinking of many Brazilian heterodox economists. Rather than being substitutes, industrial policies and a competitive real exchange rate (that is, an exchange rate at its industrial equilibrium level) complement each other as tools for achieving economic development through structural change towards activities offering high added-value per worker, most of which are still in the manufacturing sector.