Month: April 2009

Slate‘s Jacob Weisberg recently surveyed a range of sociopolitical issues from nuclear proliferation to the China Century where the expert consensus might be wrong.

Weisberg’s survey sample includes macroeconomic aggregates (home ownership, asset investment classes, international competition), geostrategic stability (China, nuclear proliferation), and longrun environmental issues (climate change, fossil fuels). In each, Weisberg contrasts a prevailing view, hypothesis or expert with a challenger.

Below are some thoughts on Weisberg’s analyses and observations on research methods in journalism.

· Selection and Framing of Experts: Weisberg mentions the late realist Samuel P. Huntington‘s thesis on political order in changing societies, to raise concerns about China’s near-future macroeconomic growth. He also refers to neorealist Kenneth Waltz‘s views that nuclear proliferation is inevitable. Huntington and Waltz both represent dominant traditions within international relations theory, and neither are as new or radical as Weisberg seems to portray. The selection and framing of experts is crucial: it would be even more interesting to compare their views with other schools of thought, such as liberal democratic, critical or constructivist theories, which have different deductive premises and levels of analysis. After all, neoconservative fears about Iraq were not just that nuclear weapons acquisition was a defensive action, but also the security orientation of ‘Axis of Evil’ regimes and their potential connections to non-state actors. Perhaps Weisberg could have checked with a nuclear proliferation specialist such as Graham Allison or Jessica Stern. Equally, a China specialist might convincingly show that Hu Jintao’s Chinese government is aware of Huntington’s thesis and has plotted a different future trajectory

· Heretics & Mavericks: Weisberg cites Freeman Dyson as a heretic of climate change models, although Dyson’s scientific expertise is primarily as a physicist and cosmologist. Other mavericks such as the late Federal Bureau of Investigation counterterrorism expert John O’Neill and United Nations weapons inspector Scott Ritter were ostracised in organisational politics, whilst economist Nouriel Roubini strengthened his reputation by foreseeing the global financial crisis. Perhaps it’s also a matter of luck, timing, and having an effective image makeover.

· Incomplete Deductive Arguments: Weisberg observes that market analysts are re-evaluating house ownership, stock investments and the global competitiveness of car manufacturers. Yet the common assumptions that Weisberg mentions are really incomplete deductive arguments with hidden premises. First, home ownership does not necessarily lead to greater community involvement, and the negative factors mentioned (financial risk, labour market mobility, commute time) are weighted during the purchase decision or emerge later in decision regret. Second, evaluating and comparing the risk premia of bonds versus stocks requires further details on the time horizon, sampling frame, weightings and volatility. Shocks such as the 1973 OPEC oil crisis, the 1995-2000 dotcom bubble, the 1997 Asian currency crisis and the current global financial crisis may affect the comparison of bond and stock returns. Third, although the Detroit Three have cut costs and launched several international joint ventures, their debts and liabilities are partly the result of earlier decisions. Weisberg’s argument that these balance sheet issues are the Detroit Three’s main barrier is not really new: asset management and private equity firms have targeted them for over two decades in their acquisition, reengineering and turnaround attempts. Perhaps that’s why the Obama administration has hired media banker Stephen Rattner.

· Inferences from Small Samples: In his sections on long-run stocks, climate change and fossil fuels Weisberg quotes from a single academic study. Whilst this establishes a challenger hypothesis it also probably means that the sample is too small for inferences that would establish a definitive Kuhnian paradigm shift in a knowledge field. Weisberg would have a more robust argument if he referenced meta-analyses which evaluated a group of studies for their sample size and other effects.