The café economystructural transformation in Greece in the wake of austerity and “reforms”

Picture a café terrace in Athens during the summer of 2019. Every table is full, mostly with tourists. Waiters are moving fast, earning not much. If you were an economist watching that scene, you might have said: this is the recovery. Greece came through the adjustment programs, labor markets opened up, barriers came down, and here is the result — activity, employment, and growth. Nikiforos, Missos, Pierros, and Rodousakis looked at that same scene and reached a different conclusion. What grew in Greece after the crisis wasn't a sign of structural modernization. It was a trap. The "café economy," as they call it — hotels, restaurants, bars, and the Accommodation and Food Service Activities sector — absorbed hundreds of thousands of workers into one of the least productive corners of the economy. They argue that the reforms designed to prevent exactly that outcome are part of what caused it. That framing deserves some pushback. The reform logic was not invented to harm Greece; it came from a standard and well-supported playbook. The three Economic Adjustment Programs, signed with the troika — the European Commission, the European Central Bank, and the International Monetary Fund — began in 2010, then again in 2012 and 2015. They combined deep fiscal consolidation with a broad structural agenda. Labor market liberalization meant easing hiring and firing, promoting temporary contracts, weakening collective bargaining, and explicitly cutting the minimum wage. Product market deregulation affected energy, transport, pharmaceuticals, and professional services. The Organisation for Economic Co-operation and Development produced seven hundred seventy-five reform recommendations across sectors; by 2019, over ninety percent had been implemented. The fiscal consolidation alone was enormous — real government expenditure fell by about thirty percent between 2009 and 2017. The premise was coherent: remove rigidities, lower unit labor costs, improve competitiveness, and trigger an export-led recovery. Internal devaluation served as a substitute for a currency adjustment that you cannot make inside the eurozone. Right, and what did the data reveal? The data showed a collapse. Real gross value added for the whole economy fell about twelve point five percent between two thousand nine and two thousand twenty-three. Aggregate labor productivity dropped sixteen percent. Unemployment peaked at roughly twenty-seven percent. Average real wages fell somewhere between twenty-six and thirty-four percent, depending on the deflator you use. That is not a recovery story by any measure I would use. The part I would emphasize is where the employment actually went. If labor markets flexed and workers moved into more productive activities, the model still works even if the transition was painful. That is precisely the test. What actually happened is that of the roughly seven hundred forty thousand jobs created since the two thousand thirteen employment trough, about three hundred twenty thousand, or forty-three percent, went into Accommodation and Food Service Activities. The share of total employment in this sector nearly doubled, from seven point seven percent to thirteen point eight percent. Its share of value added rose from five percent to seven percent. The average for Accommodation and Food Service Activities employment in the European Union is five percent. Greece is now almost three times that. And what about AFSA's productivity? Down about forty-one percent since two thousand nine, measured per employee. Down thirty-five percent per hour worked. It was already among the lowest productivity sectors before the crisis. So the sector that absorbed the most labor is also the one that became less productive the fastest. That is the core paradox the paper documents. Nikiforos and colleagues use a shift-share decomposition to make this precise and to put a number on exactly how much of the aggregate productivity damage came from this specific dynamic. The decomposition separates two sources of change in economy-wide productivity: improvements or declines within individual sectors, and the movement of workers between sectors of different productivity levels. When workers move toward high-productivity sectors, the reallocation term is positive. When they migrate into low-productivity Agricultural and Farm Support Activities, it creates a drag. Applied to Greece for the years 2009 through 2023, that drag is measurable and significant. The collapse in productivity within the Agricultural, Forestry, and Seafood Activities sector, along with the reallocation of workers into that sector, accounts for a reduction of 5.2 percentage points out of the total decline of 16.13 percentage points — nearly one-third of the overall productivity loss. Based on the hours worked measure, the combined contribution of this sector is a reduction of 4.5 percentage points out of the total decline of 11.87 percentage points. Nearly a third from one sector. One sector that the reform programs did not specifically target for expansion. So the question I keep coming back to is, why did firms rationally expand there instead of in higher productivity sectors where competition was supposedly being unleashed? That is the mechanism the paper argues is not accounted for in the standard reform model. There are two channels. The first is demand. The fiscal consolidation wiped out roughly a quarter of gross domestic product in a very short period. The Kaldor-Verdoorn relationship, which is the empirical regularity that productivity growth tracks output growth, means that when demand collapses that severely, you lose the scale economies, the learning by doing, and the induced investment that drives productivity forward. It’s not just that output fell; the collapse itself removed the incentive and opportunity to upgrade. The Kaldor-Verdoorn mechanism is something I can work with. The second channel is where I would want more precision. The second channel is the cost of labor. Structural reforms and mass unemployment drove wages down sharply, by sixty percent in AFSA specifically. The paper argues that when labor is that cheap, profit-maximizing firms do not invest in labor-saving technology. They hire people instead. The incentive to automate, to upgrade, or to adopt more productive techniques disappears when you can staff a restaurant for next to nothing. The standard model states that lower labor costs improve competitiveness and attract investment. Investment in what, though? The paper distinguishes between investment that competes on price in a low-margin sector and investment in productivity-enhancing capital. Wage compression makes the first rational and the second unnecessary. The authors frame this explicitly as a Lewis-type dual economy, which consists of a high-productivity modern sector and a low-productivity traditional sector that absorbs surplus labor. The Agricultural Food Sector Alliance is the traditional sector. The Lewis mechanism states that the traditional sector continues to grow as long as there is surplus labor to absorb and wages remain low enough to make it profitable. The Lewis framing is interesting, but Lewis was describing pre-industrial economies with subsistence agriculture. Greece in two thousand ten was an advanced economy with a services base. The paper acknowledges this. Their claim isn't that Greece became a developing economy — it's that the structural dynamics of dualization described by Lewis can be reproduced in any context where a sufficiently large productivity gap opens between sectors, and wage compression makes low productivity expansion rational. They demonstrate that the gap exists and has widened. Let me try to state the version of the reform argument that survives this. Tourism did improve something real — the travel surplus rose from three point four percent of GDP in 2009 to eight point three percent in 2023. The goods trade deficit narrowed by roughly eight percentage points of GDP between 2008 and 2016. Those are external balance improvements that were the explicit goal of internal devaluation. The paper counts them. It doesn't dispute the improvement in external balance. What it argues is that those gains coexist with overtourism, environmental strain, housing pressure, and the structural limitation that the African Financial Services Act generates low value added with decreasing returns. You can fix the current account and still damage the productive capacity of the economy. That framing — fixing the current account while damaging productive capacity — is actually where the paper's two mechanisms meet. Exactly. Demand collapse through fiscal consolidation killed the Kaldor-Verdoorn engine for productivity growth. Wage compression through labor market reform removed the firm-level incentive to invest in technology. Both operated simultaneously, and both pushed in the same direction: making AFSA expansion the rational response for firms and workers. The reforms didn't fail despite their logic; the logic produced the outcome. The concluding implication is hard to dismiss given those numbers. Greece's Agricultural Food Sector Authority employment share is nearly 14 percent compared to an EU average of 5 percent. Its productivity is 40 percent below 2009 levels, and aggregate labor productivity remains 16 percent below 2009 after 14 years. These are not transition costs. This looks like an equilibrium. The paper refers to it as a risk of permanent low-productivity equilibrium. Their policy prescription follows from this diagnosis: restoring demand, deliberately promoting higher-productivity sectors, and supporting technological upgrading. Not more horizontal deregulation. If the mechanism that drove AFSA expansion was cheap labor combined with a demand collapse, then neither stabilization alone nor further labor market liberalization can reverse it. You need the Kaldor-Verdoorn engine restarted — output growth that makes productivity investment worthwhile again. Whether that policy shift is politically or institutionally achievable within the current European framework is a different question. However, the empirical case for why the café economy emerged — and why the reforms that promised competitiveness delivered this instead — is harder to wave away than it was twenty minutes ago. That's the sharpest contribution of the paper. It's not just about documenting the outcome; it also traces the mechanism that made it rational. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.