Is there a transatlantic divide in undergraduate macroeconomics teaching?

Abstract

The global financial crisis triggered different policy responses in Europe and the United States. Interestingly, survey results suggest that there is also a significant difference in how undergraduate macroeconomics instructors responded to the crisis, with US instructors placing significantly more emphasis on financial topics than their European peers. This note considers whether such differences may be attributed to differences in instructors’ profiles and teaching environments. The results suggest that, rather than explaining this gap, the transatlantic divide becomes even wider when analysed in a multivariate setting.

Keywords:

• teaching
• undergraduate
• macroeconomics
• financial crisis

JEL Classification:

• A22
• E00

Notes

¹ For pertinent quotes from academia, see Gärtner et al. (2013), pp. 1–2. For examples of how popular media chimed in with similarly divided opinions, see Cohen (2009) and The Economist (2010). Specific proposals as to how undergraduate macroeconomics instruction should respond to the crisis are listed in Blinder (2010) and Shiller (2010). For a summary of how introductory textbooks changed after the crisis, see Madsen (2013).

² In the United States, about 25 000 bachelor’s degrees in economics are awarded every year, compared with 2500 pertinent master’s degrees and 1000 doctoral degrees. See Snyder and Dillow (2010). The instructors who responded to our survey reported that they teach some 50 000 students in their mandatory macroeconomics courses. And although education became a globalized industry over the past two decades, it seems still reasonable to assume that there is geographic persistency in the sense that the likelihood of being employed in the USA is higher for those being educated in the USA and vice versa, which potentially establishes a link between economics teaching and economic policies.

³ The classification of universities as top research institutions follows the ranking provided in Coupé (2003).

⁴ For details, see Gärtner et al. (2011).

⁵ Of course, there are many variables that may drive the results but are not incorporated in our model like personal income, tenure status, publication record or department policies, just to name a few. However, apart from the question of how to measure such variables, even to ask for them when conducting a survey obviously jeopardizes the response rate and thus the overall quality of the data. Facing this trade-off, we were cautious when asking for personal details and realized a comparatively high response rate.

⁶ The four regressions reported in Table 4 directly relate to the four columns shown in Table 3. The endogenous variables are sums over all the topics, however, and absolute numbers rather than percentages are used for a more straightforward interpretation.

⁷ The quantitative impact is small, however. Given the difference in enrolment reported in Table 2, class size explains 0.003 × 108.08 = 0.324 more financial topics in a European course.

⁸ This gap only exists on the programme level of the respective institutions, not on the level of the courses taught by the respondents. Since we have no information about those ‘other’ instructors in the programme who essentially appear to cause this model divide, we lack the data to move beyond descriptive statistics and look for an econometric explanation in a multivariate setting.

⁹ The number of topics taught in the curriculum before the crisis is not available. This may not matter too much, because the change in the number of topics taught by the respondents is minimal.

¹⁰ Multicollinearity, caused by the overlap between these two variables, does not call these results into question, since it leaves the coefficients unbiased and may only inflate the SEs and, thus, underestimate the significance levels.