ABSTRACT
This article presents estimates of labour values and prices of production following two approaches: the first is based on the classical and Marxian theory of value and distribution; the second on the so-called ‘new solution’ to the ‘transformation problem’ and its variant, the Temporary Single-System Interpretation (TSSI). The major advantage of the latter approach is its simplicity, along with the relatively low data requirements. Our empirical findings from the economies of China, Japan and South Korea suggest that both approaches give estimates of labour values and prices of production which are extremely close to each other as well as to actual market prices. On further examination, however, we conclude that our empirical findings are absolutely consistent with the theoretical requirements of the classical approach and contradict those of the TSSI.
Acknowledgments
We wish to thank, without implicating, Theodore Mariolis and three anonymous referees of this journal for their insightful comments on previous versions of the article.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 Vectors are indicated in boldface letters and a prime over these letters indicates their transpose. Matrices are indicated in capital and boldface letters.
2 For a comprehensive discussion of the same issue and a relevant numerical example, see Ochoa (Citation1984, pp. 58–70), and for a discussion within the context of Leontief’s price model, see Miller and Blair (Citation2009, Ch. 2).
3 For a theoretical critique along the classical lines of both the new solution to the transformation problem and the TSSI, see Petri (Citation2015) and the literature cited therein.
4 The data were accessed on 15 March 2016, and the link is www.wiod.org and the documentation is provided in Timmer et al. (Citation2015).
5 Frölich (Citation2011) argues that Diaz and Osuna (Citation2007, Citation2009) derive the alleged bias of the physical units of measurement by the inappropriate use of logarithms and their properties that are applicable to pure numbers.
6 We also tried regressions with percentage changes in both price–value deviations and VICC and the results were 100 per cent consistent with the classical model and the shortcomings of the TSSI model remained the same.
7 The results with circulating capital were very similar and are available from the authors on request. In effect, most of the discussions in this literature take place in terms of circulating capital for reasons that have to do either with the lack of adequate data on the matrix of fixed capital stock and the many zero rows that appear in such a matrix that it actually imposes its form on its vertical integrated expression that give rise to minimal subdominant eigenvalues (for details, see Mariolis and Tsoulfidis Citation2016a, Citation2016b).