Skip to main content
SpringerLink
Log in

Logical Gaps: Path C—Empirical Results to the Real World

  • Chapter
  • First Online:
Rethinking Economics

Part of the book series: SpringerBriefs in Economics ((BRIEFSECONOMICS))

  • 1096 Accesses

Abstract

There are statistical criteria which are commonly used to determine the value of empirical results. Good performance under these criteria does not necessarily imply value for policy making. The rhetoric around such results can be influential, but may not be justified. Results are frequently misinterpreted. There are problems with some of the tests, as with hypothesis testing and the fallacy of the transposed conditional. A limited range of resulting policy options is considered. Numerous standard policy questions are not addressed, including ease of change and the costs and benefits of change. A further issue, the use of data to show advantage or disadvantage, is discussed.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution
Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 39.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 54.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Notes

  1. 1.

    For a recent contribution, see Wolf (2007).

  2. 2.

    See also Birks (2007).

  3. 3.

    While theoretical simplicity is commonly lauded, more complex analyses with longer equations, more advanced mathematics and/or larger data bases are also afforded high status.

  4. 4.

    Ziliak and McCloskey (2008) has been critically reviewed (Spanos 2008). However, the criticisms relate to proposed solutions to the problem. If anything, Spanos suggests that the problems themselves are more severe than suggested by Ziliak and McCloskey.

  5. 5.

    This point has been made also in relation to policy in education:

  6. 6.

    Based on their abstracts, econometrics was central to 10 of 11 articles in Applied Economics 39(21), December 2007. Of these, three drew some possible implications for decisions/policy. More recent issues include papers using other techniques, including experimental economics, but the emphasis is still on quantitative research, predominantly econometrics.

  7. 7.

    As a separate exercise, it may be interesting to see to what extent academic research impacts on policy decisions, and, if considered, whether the findings are correctly interpreted. Example 2 below is a case in point.

  8. 8.

    http://www.rbnz.govt.nz/research_and_publications/discussion_papers/2007/. Approximately 7 of the 15 papers in 2007 would fit into this category. The greater emphasis on policy implications from econometric analysis may be due to the Reserve Bank having defined objectives and a limited number of policy instruments. This to a degree constrains the range of analysis required, reducing the alternatives and associated costs and benefits to be considered. Nevertheless, none of the six discussion papers in 2013 directly relate to policy (http://www.rbnz.govt.nz/research_and_publications/discussion_papers/2013/).

  9. 9.

    Note that one of Dunn’s ‘modes of argumentation’ in policy debate is ‘reasoning from authority’ based on the achieved or ascribed status of the person presenting the information. (Dunn 2004, p. 395).

  10. 10.

    See also NZPA and Reuters (2007, 1 November) on obesity and cancer, ( Palmer 2007, 30 November) on job cancer risks, Perry (2008, 18 January) and Medical Update (2002). It may not be coincidental that so many examples are health related. The media considers reader interest and this often requires a personal angle (Hamilton 2004). The recommendations in the articles may shape perceptions, behaviour and policy.

  11. 11.

    This example is discussed further in (Birks 2002).

  12. 12.

    Note the discussion of control variables in Sect. 3.2.1.

  13. 13.

    For a discussion of the use of econometrics in law, including reservations and qualifications, see Harkrider (2005).

  14. 14.

    As described in Sect. 3.2.3 above, Stringer’s justification for action research is based on the limited value of studies seeking generalised patterns such as these.

  15. 15.

    There is scope to debate the criteria for determining whether results are ‘statistically significant’. What does the test result actually show (noting the fallacy of the transposed conditional described in Sect. 4.2.1), and is the result categorical, giving a firm yes or no answer given choice of levels of significance? This discussion takes statistical findings as given, looking at the subsequent stage of interpretation of results for policy purposes.

  16. 16.

    Douglas Adams, in The hitchhiker’s guide to the galaxy used the same flawed reasoning in his ‘proof’ of the non-existence of God ( Adams et al. 1981, p. 60).

  17. 17.

    Even that is not enough to fully internalise the externality as there is no compensation paid to the losers.

  18. 18.

    The so-called Coase Theorem asserts that, with no transaction costs, bargaining can result in an efficient outcome in the presence of externalities irrespective of the initial allocation of property rights. This was not the point that Coase was making, however. He recognised the significance of transaction costs and suggested that this might limit the negotiation. Consequently, he considered it important for efficient activity that the courts make the right allocation in the first place. This point has often been missed, as he indicated in his Nobel Prize lecture (Coase 1991), in which he objected to the interpretation by others of his findings with, ‘the infamous Coase Theorem, named and formulated by Stigler… I tend to regard the Coase Theorem as a stepping stone on the way to an analysis of an economy with positive transaction costs’.

  19. 19.

    The use of law for policy, including economic policy, is an important area in itself. Elected politicians make laws, and so see this as a policy instrument. Legal processes and legal reasoning are then important for implementation. There is also rhetoric in law, described as ‘judicial eloquence’ by Smith (1963).

  20. 20.

    Figure 1 of the study (Landhuis et al. 2007, p. 535) indicated fewer attention problems among those watching for 1–2 h per day compared to those watching less than 1 h per day, which suggests first that the relationship may be non-linear, and second that increased viewing may be beneficial for low watchers (if the relationship is causal).

  21. 21.

    By this I mean a textbook ‘ideal’. Note that criteria to be considered, methods of evaluation, and choice of decision rule are all open to question.

  22. 22.

    While these questions are raised in relation to econometric studies, they apply to all policy options where one (policy) variable is altered so as to bring about a change in another (target) variable.

  23. 23.

    This is one of the key concerns raised in Ziliak and McCloskey (2008). Note that effects on Y, and the policy significance of the resulting Y, may not always be continuous. This can cause particular problems, especially where variability of outcome is important. Consider the difference, for a non-swimmer standing in a tank of water, between a situation where the water level is exactly at shoulder height and one where the water level is, on average, at shoulder height. Econometric estimation gives average impacts only.

References

Download references

Author information

Authors and Affiliations

  1. Massey University, Palmerston North, New Zealand

    Stuart Birks

Authors
  1. Stuart Birks
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Stuart Birks .

Rights and permissions

Reprints and permissions

Copyright information

© 2015 The Author(s)

About this chapter

Cite this chapter

Birks, S. (2015). Logical Gaps: Path C—Empirical Results to the Real World. In: Rethinking Economics. SpringerBriefs in Economics. Springer, Singapore. https://doi.org/10.1007/978-981-287-176-3_4

Download citation

Publish with us

Policies and ethics

Search

Navigation