Historical Patterns of Inequality and Productivity around Financial - PowerPoint PPT Presentation

Historical Patterns of Inequality and Productivity around Financial Crises Pascal Paul Federal Reserve Bank of San Francisco 1 1 The views expressed herein are solely mine and do not necessarily reflect the ones of the Federal Reserve Bank of San Francisco or the Federal Reserve System.

Research Questions • Are financial crises more likely after particular macroeconomic developments? a • Once a crisis occurs, what determines the severity of the following recession?

Research Questions • Are financial crises more likely after particular macroeconomic developments? a • Once a crisis occurs, what determines the severity of the following recession?

U.S. around Great Recession 10 20 46 5 19 45 0 18 44 17 43 -5 2000 2003 2006 2009 2000 2003 2006 2009 2000 2003 2006 2009 3 4 4 2 3 3 1 2 2 0 1 1 -1 0 -2 -1 0 2000 2003 2006 2009 2000 2003 2006 2009 2000 2003 2006 2009

Data • Long-run historical perspective • 17 advanced economies • Annual data, 1870 − 2013 • Combining data sets a 1. Jordà, Schularick, and Taylor (2017) → Macrofinancial Data 2. Bergeaud, Cette, and Lecat (2016) → Productivity Data 3. World Wealth & Income Database → Inequality Data

Data • Long-run historical perspective • 17 advanced economies • Annual data, 1870 − 2013 • Combining data sets a 1. Jordà, Schularick, and Taylor (2017) → Macrofinancial Data 2. Bergeaud, Cette, and Lecat (2016) → Productivity Data 3. World Wealth & Income Database → Inequality Data

Main Results Buildup Period • ∆ Top income shares and productivity growth are robust crises predictors • Predictive power of top income shares explained by asset price booms a Recovery Period • Recessions preceded by rising inequality or low productivity growth associated with deeper and slower recoveries

Main Results Buildup Period • ∆ Top income shares and productivity growth are robust crises predictors • Predictive power of top income shares explained by asset price booms a Recovery Period • Recessions preceded by rising inequality or low productivity growth associated with deeper and slower recoveries

Literature & Channels Inequality • Inequality and Financial Crises (Morelli and Atkinson, 2015; Kirschenmann et al., 2016) • Credit channel (Kumhof, Rancière, and Winant, 2015; Bordo and Meissner, 2012) a Productivity • “Bad credit booms” associated with low productivity growth (Gorton and Ordoñez, 2016)

Literature & Channels Inequality • Inequality and Financial Crises (Morelli and Atkinson, 2015; Kirschenmann et al., 2016) • Credit channel (Kumhof, Rancière, and Winant, 2015; Bordo and Meissner, 2012) a Productivity • “Bad credit booms” associated with low productivity growth (Gorton and Ordoñez, 2016)

Macrohistory Database Jordà, Schularick, and Taylor (2017) • Macrofinancial variables • Financial crises dates Definition Dates • Recession dates 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1870 1890 1910 1930 1950 1970 1990 2010

World Wealth & Income Database 60 60 60 60 40 40 40 40 20 20 20 20 0 0 0 0 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 60 60 60 60 40 40 40 40 20 20 20 20 0 0 0 0 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 60 60 60 60 40 40 40 40 20 20 20 20 0 0 0 0 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 60 60 60 60 40 40 40 40 20 20 20 20 0 0 0 0 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000 1880 1900 1920 1940 1960 1980 2000

Long-term Productivity Database Bergeaud, Cette, and Lecat (2016) Y t = A t K α t H 1 − α , LP t = Y t t H t Details Utilization-adjusted TFP 6 5 4 3 2 1 0 -1 -2 1900 1920 1940 1960 1980 2000

Part I: Predicting Financial Crises

Typical Patterns around Financial Crises Good Booms, Bad Booms

Logit Regressions � P [ FC k , t = 1 |· ] � = α k + β 1 ∆ h Z k , t − 1 + β 2 ∆ h X k , t − 1 + u k , t log P [ FC k , t = 0 |· ]

Predicting Financial Crises - Inequality (1)*** (2)*** (3)*** (4)*** (5)*** ∆ 4 log ( Credit ) t − 1 -0.772*** -0.817*** -0.677*** -(0.333)*** -(0.339)*** -(0.333)*** -[0.018]*** -[0.013]*** -[0.013]*** ∆ 4 Income Share 1 % t − 1 -0.887*** -0.954*** -(0.258)*** -(0.316)*** -[0.019]*** -[0.015]*** ∆ 4 Income Share 10 % t − 1 -0.784*** -0.738*** -(0.251)*** -(0.279)*** -[0.018]*** -[0.014]*** Number of crises -24*** -24*** -24*** -24*** -24*** Observations -670*** -670*** -670*** -670*** -670*** Countries -13*** -13*** -13*** -13*** -13*** Country FE - � *** - � *** - � *** - � *** - � *** *** p-value -0.623*** -0.869*** -0.878*** -0.735*** -0.855*** Pseudo R 2 -0.086*** -0.109*** -0.090*** -0.159*** -0.130*** AUROC -0.758*** -0.772*** -0.751*** -0.825*** -0.810*** -(0.053)*** -(0.053)*** -(0.052)*** -(0.049)*** -(0.047)*** AUROC Figure

Predicting Financial Crises - Inequality (1) (2) (3) ∆ 4 Income Top 1 % t − 1 -0.928*** -(0.341)*** -[0.016]*** ∆ 4 Income Top 1 % Labor t − 1 -0.253*** -0.180*** -(0.339)*** -(0.456)*** [-0.004]*** [-0.002]*** ∆ 4 Income Top 1 % Capital t − 1 -1.084*** -1.044*** -(0.286)*** -(0.346)*** -[0.015]*** -[0.014]*** Credit, Stock Prices, House Prices - � *** Number of crises -8*** -8*** -8*** Observations -302*** -302*** -302*** Countries -8*** -8*** -8*** Country FE p-value Pseudo R 2 -0.100*** -0.174*** -0.184*** AUROC -0.747*** -0.813*** -0.778*** -(0.105)*** -(0.071)*** -(0.080)***

Predicting Financial Crises - Inequality (1) (2)** (3)*** ∆ 4 log ( Total Capital Income ) t − 1 -0.060 -0.053** -0.078*** -(0.235) -(0.072)** -(0.094)*** -[0.001] [-0.001]** [-0.001]*** ∆ 4 Income Top 1 % t − 1 -0.704** -(0.294)** -[0.015]** ∆ 4 Income Top 1 % Labor t − 1 -0.008*** -(0.317)*** [-0.000]*** ∆ 4 Income Top 1 % Capital t − 1 -1.048*** -(0.324)*** -[0.013]*** Number of crises -9 -9** -5*** Observations -404 -322** -230*** Countries -7 -7** -6*** Country FE p-value Pseudo R 2 -0.000 -0.053** -0.152*** AUROC -0.511 -0.698** -0.796*** -(0.101) -(0.090)** -(0.087)***

Robustness - Inequality • Excluding country-fixed effects Details • Ordinary Least Squares Details • Pre- and Post-WWII samples Details • Excluding Great Depression and Great Recession Details • Financial vs. Nonfinancial Recessions Details • Change in log of income share Details • Demeaned moving average Details • Alternative Crises Dates Details • Interpolation over small gaps Details • Wealth-to-income ratio Details

Controls - Inequality (1)*** (2)*** (3)*** ∆ 4 Income Share 1 % t − 1 -0.667*** -0.732*** -0.699*** -(0.222)*** -(0.223)*** -(0.219)*** -[0.011]*** -[0.011]*** -[0.008]*** Set of controls - � *** - � *** - � *** GDP & Global GDP - � *** - � *** Stock & House Prices - � *** Number of crises -29*** -29*** -27*** Observations -781*** -781*** -738*** Countries -14*** -14*** -14*** Country FE - � *** - � *** - � *** *** p-value -0.748*** -0.811*** -0.366*** Pseudo R 2 -0.163*** -0.176*** -0.227*** AUROC -0.814*** -0.824*** -0.859*** -(0.042)*** -(0.039)*** -(0.030)*** Individual Controls

Predicting Financial Crises - Productivity (1)*** (2)*** (3)*** (4)*** (5)*** ∆ 4 log ( Credit ) t − 1 -0.458*** -0.537*** -0.597*** -(0.126)*** -(0.142)*** -(0.147)*** -[0.013]*** -[0.014]*** -[0.015]*** ∆ 4 log ( TFP ) util . − adj . -0.269*** -0.407*** t − 1 -(0.129)*** -(0.132)*** [-0.008]*** [-0.011]*** ∆ 4 log ( LP ) t − 1 -0.303*** -0.512*** -(0.115)*** -(0.124)*** [-0.009]*** [-0.013]*** Number of crises -53*** -53*** -53*** -53*** -53*** Observations -1552*** -1552*** -1552*** -1552*** -1552*** Countries -17*** -17*** -17*** -17*** -17*** Country FE - � *** - � *** - � *** - � *** - � *** *** p-value -0.963*** -0.983*** -0.981*** -0.946*** -0.943*** Pseudo R 2 -0.044*** -0.024*** -0.025*** -0.058*** -0.066*** AUROC -0.695*** -0.626*** -0.636*** -0.711*** -0.724*** -(0.035)*** -(0.040)*** -(0.038)*** -(0.035)*** -(0.032)*** AUROC Figure Horse Race

Robustness - Productivity • Excluding country-fixed effects Details • Ordinary Least Squares Details • Pre- and Post-WWII samples Details Comparison • Excluding Great Depression and Great Recession Details • Financial vs. Nonfinancial Recessions Details • Input Factors Y t A t = K α t H 1 − α Details t • Alternative Crises Dates Details