105 of 112 people found the following review helpful
- Published on Amazon.com
On April 15, 2013, a year and a half after I had first published this review a study by Thomas Herndon, Michael Ash, and Robert Pollin from the U of Massachusetts came out and refutted the authors main thesis that once a country reaches a Debt/GDP ratio of 90% sees its economic growth contract nearly automatically. This had become a covenant of libertarians such as Paul Ryan and Europeans promoting fiscal austerity. It turns out that Reinhart and Rogoff studies were completely wrong. R&R made numerous mistakes pointed out by the U of Mass team. The main one was to exclude three years out of the New Zealand data during a high Debt/GDP period. During those three excluded years New Zealand had grown very rapidly which contradicted R&R thesis. Once you make those corrections (including a few others that were minute by comparison), there is no statistical difference in growth rate between countries with high Debt/GDP ratio vs ones with lower ones. So much for Austerity. This is a devastating blow to what we thought was a classic study on the subject. Below see my original review. Notice that I had also observed many other flaws with their work but not the one mentioned above since I never saw the data firsthand.
This book is both fascinating and flawed. Starting with the flaws:
First, the book is mistitled. It covers the last 200 years not the last 800.
Second, their crisis framework is convoluted relative to the crystal clear framework of Charles Kindleberger in Manias, Panics, and Crashes: A History of Financial Crises (Wiley Investment Classics). The latter leans on the seminal work of Irving Fisher The Debt-Deflation Theory of Great Depressions and Hyman Minsky (the credit cycle exacerbates the business cycle) that the authors completely ignore.
Third, some of their analyses are obfuscating. They baffle the reader on how frequently emerging market countries default with surprisingly low external debt levels. Later, the authors clarify that debt levels are far higher when including domestic debt; then the baffling turns into the self-evident.
Fourth, in Chapter 16, their development of a crisis index measure is weak with no predictive power. The first two graphs capturing this index (ranging from 1 to 5) over the past 100 years have the wrong y-axis (ranging from 0 to 180?) rendering the graph incomprehensible (pg. 253, 254). Two pages later, they use the correct scale (1 - 5).
Fifth, the graph on page 267 denoting the % collapse of exports during the Great Depression has the wrong sign.
Sixth, some of their conclusions are already outdated. They advance that Greece, Portugal, Italy, and Spain are all doing better than in recent years. The book came out in 2009; didn't those countries show signs of fiscal stress? Since 1800, Greece suffered external debt defaults or rescheduling in over 50% of the years.
Seventh, their argument that large Current Account Deficits (CADs) fuel housing bubbles is not supported. When they show the magnitude of the rise in housing prices over 2002 - 2006 for many countries (Fig. 15.1), it is unclear if there are any relationship between high CAD and housing Bubbles. The housing bubble was far greater in many former USSR satellites than anywhere else (unclear if they had high CADs).
Moving on to the ambivalent OK parts:
1) Their early warning indicators of banking and currency crises (Table 17.1) are interesting. They indicate that 12 month changes in real housing and stock prices are good early signals for banking crises. They mention other metrics such as CAD levels. But, those indicators are unsupported by any statistical analysis.
Moving on to the good parts:
1) Their prototype sequencing of crises represents their best work. It shows how a nation can experience in succession financial deregulation, banking crisis, currency crash, inflation spike, and ultimately default. The tipping point is when a government faces an untenable choice between defending its currency (restrictive policies) and shoring up its financial sector (expansive policies). Governments invariably abandon supporting their currency.
2) Their historical data facilitate interesting observations:
2a) Crisis related to sovereign risks are so frequent, you wonder how countries ever manage to raise debt. While developed countries have "graduated" from defaults, they have not from banking crises. Since 1800, the UK, US, and France have experienced 12, 13, and 15 episodes of banking crises. Banking crises have been frequent since the 1980s. Developed countries are prone to banking crises because financial deregulation is a causal factor. In 18 of 26 banking crises observed since 1970, the financial sector had been liberalized within the preceding 5 years.
2b) Post WWII financial crises have been severe. On average, real housing prices decline by 35% over 6 years; stocks crash by 56% over 3.5 years; unemployment rate increases by 7 percentage points; GDP contracts by 9%; and, public debt rises by 86%.
2c) The US Subprime crisis was more severe than any other post WWII financial crisis. Its housing and stock market bubbles were more pronounced. The US CAD as a % of GDP was larger. The downturn in GDP was more severe. The resulting increase in public debt was faster. The ramp up of all mentioned indicators suggested a financial crisis was imminent. The authors remark that if the US had been an emerging market relying on external debt (in foreign currency), the US dollar value would have plummeted and interest rates soared.
3) When the authors move on to the US Subprime crisis, they note how the majority of experts, including Bernanke and Greenspan, were not concerned regarding the rising US Current Account Deficit (CAD) and rising housing prices. These experts stated the CAD and home price increases were associated with a World savings glut resulting from Asian export led economies. Meanwhile others (Rubini, Krugman, and the authors) were concerned about the CAD sustainability (absorbing 2/3d of World savings), housing prices (in real term rose by 92% between 1996 and 2006 or more than 3 x the 27% increase from 1890 to 1996! See graph pg. 207) and the massive increase in US household debt (rose from a norm of 80% of personal income to 130% by 2006).
If you are interested in this subject, I also recommend Raghuram Rajan's Fault Lines: How Hidden Fractures Still Threaten the World Economy [New in Paper].