Skip to main content

Capital Input Cost

  • Chapter
  • First Online:
Productivity

Part of the book series: Contributions to Economics ((CE))

  • 457 Accesses

Abstract

Unlike labour productivity change, the measurement of total factor productivity change (or difference) crucially depends on the measurement and decomposition of capital input cost. This chapter discusses in detail the basics of its measurement and shows that one can dispense with the usual neo-classical assumptions. Various models are reviewed, as well as the role played by the ‘rate of return’.

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

Access this chapter

eBook
USD 16.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 129.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info
Hardcover Book
USD 129.99
Price excludes VAT (USA)
  • Durable hardcover 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.

    Parts of this chapter are based on joint work with Dirk A. van den Bergen at Statistics Netherlands in the years 2000–2010.

  2. 2.

    It is of course a simplification to assume that the economically maximal service life of an asset type is some given constant. Actually a time superscript should be added. See Diewert (2009) for some theoretical considerations.

  3. 3.

    Fatica (2017) multiplies the first and second component by a factor accounting for corporate taxation.

  4. 4.

    According to Rymes (1983) this naming goes back to Pigou.

  5. 5.

    SNA (2008, par. 6.130) implicitly prescribes that for non-market units the interest rate r t must be set equal to 0.

  6. 6.

    It should be noted that all the operational costs associated with the use of a particular asset are accounted for as intermediate or labour inputs cost. For investment decisions expected values of such cost components must of course be considered together with expected user cost.

  7. 7.

    The slightly different concept of ex ante user cost goes back to Hicks (1946, 193–194).

  8. 8.

    In the case of irreversible investments, such as electricity networks, basically a slight modification of expression (3.6) must be used, as argued by Diewert et al. (2009, Chapter 10). The first part remains as it is; in the second part, now called amortisation amount, instead of the actual end-of-period price the expected end-of-period price must be used; and the tax component disappears.

  9. 9.

    The factor (1∕2)r t is meant as an approximation to (1 + r t)1∕2 − 1, and the factor \((1/2)\tau _{ij}^{t}\) as an approximation to \(((1 + \tau _{ij}^{t}/P_{i,j-0.5}^{t^-})^{1/2} - 1)P_{i,j}^{t}\).

  10. 10.

    Authors such as Hulten (1990) suggest that “in equilibrium” this will be the case. Hulten (2010) calls it a “strong assumption.”

  11. 11.

    In the literature one sometimes finds the productive capital stock at current prices defined as \(\sum _{i=1}^{I}P_{i0}^{t}\left (\sum _{j=1}^{J_{i}}\phi _{ij}I_{i0}^{t-j}\right )\).

  12. 12.

    See SNA (2008, par. 20.22–24) for a simple proof.

  13. 13.

    Notice that there is no underlying probability distribution assumed.

  14. 14.

    See Erumban (2008) on the estimation of expected lifetimes for three types of assets in a number of industries.

  15. 15.

    Adding total (= unanticipated plus anticipated) revaluation to profit would be consistent with SNA (2008)’s prescription for non-market units.

  16. 16.

    Lovell and Lovell (2013) compared the value-added based TFP indices with exogenous and S-K rates of return for the Australian Coal Mining industry over the period 1991–2007 and found that the differences were almost negligible.

  17. 17.

    See Schreyer (2010) or Görzig and Gornig (2013). The last authors specifically consider the effect on the S-K rate of return of including intangible capital at the input and output side of the production units.

  18. 18.

    This corresponds to the SNA (2008, par. 12.87) advice on calculating “neutral holding gains.”

  19. 19.

    See Diewert (2009) for a model justifying such a profile.

References

  • Blades, D. 2015. A Short-cut Method of Estimating Capital Stocks: When Can It Be Used and How Well Does It Work? The Review of Income and Wealth 61: 373–393.

    Article  Google Scholar 

  • Coremberg, A. 2008. The Measurement of TFP in Argentina, 1990–2004: A Case of the Tyranny of Numbers, Economic Cycles and Methodology. International Productivity Monitor 17: 52–74.

    Google Scholar 

  • Diewert, W.E. 2009. The Aggregation of Capital over Vintages in a Model of Embodied Technical Progress. Journal of Productivity Analysis 32: 1–19.

    Article  Google Scholar 

  • Diewert, W.E., and D. Lawrence. 2000. Progress in Measuring the Price and Quantity of Capital. In Econometrics and the Cost of Capital: Essays in Honor of Dale W. Jorgenson, ed. L. Lau. Cambridge MA: MIT Press.

    Google Scholar 

  • Diewert, W.E., D. Lawrence, and J. Fallon. 2009. The Theory of Network Regulation in the Presence of Sunk Costs (Economic Insights Pty. Ltd., Hawker ACT, Australia).

    Google Scholar 

  • Diewert, W.E., and P. Schreyer. 2008. The Measurement of Capital. The New Palgrave Dictionary of Economics, ed. S.N. Durlauf, and L.E. Blume, 2nd ed. Palgrave Macmillan.

    Google Scholar 

  • Erumban, A.A. 2008. Lifetimes of Machinery and Equipment: Evidence from Dutch Manufacturing. The Review of Income and Wealth 54: 237–268.

    Article  Google Scholar 

  • Fatica, S. 2017. Measurement and Allocation of Capital Inputs with Taxes: A Sensitivity Analysis for OECD Countries. The Review of Income and Wealth 63: 1–29.

    Article  Google Scholar 

  • Görzig, B., and M. Gornig. 2013. Intangibles, Can They Explain the Dispersion in Return Rates? The Review of Income and Wealth 59: 648–664.

    Article  Google Scholar 

  • Hicks, J.R. 1946. Value and Capital. 2nd ed. Oxford: Clarendon Press.

    Google Scholar 

  • Hulten, C.R. 1990. The Measurement of Capital. In Fifty Years of Economic Measurement, ed. E.R. Berndt, and J.E. Triplett. Studies in Income and Wealth, vol. 54. Chicago and London: The University of Chicago Press.

    Google Scholar 

  • Hulten, C.R. 2010. Growth Accounting. In Handbook of the Economics of Innovation, ed. B.H. Hall, and N. Rosenberg, vol. 2. North Holland: Elsevier.

    Google Scholar 

  • Lovell, C.A.K., and J.E. Lovell. 2013. Productivity Decline in Australian Coal Mining. Journal of Productivity Analysis 40: 443–455.

    Article  Google Scholar 

  • OECD. 2009. Measuring Capital; OECD Manual. Second edition (OECDiLibrary).

    Google Scholar 

  • Oulton, N. 2007. Ex Post versus Ex Ante Measures of the User Cost of Capital. The Review of Income and Wealth 53: 295–317.

    Article  Google Scholar 

  • Rymes, T.K. 1983. More on the Measurement of Total Factor Productivity. The Review of Income and Wealth 29: 297–316.

    Article  Google Scholar 

  • Schreyer, P. 2010. Measuring Multi-Factor Productivity when Rates of Return are Exogenous. In Price and Productivity Measurement, Volume 6, ed. W.E. Diewert, B.M. Balk, D. Fixler, K.J. Fox, and A.O. Nakamura. Trafford Press, www.vancouvervolumes.com, www.indexmeasures.com.

  • SNA. 2008. System of National Accounts (European Commission, International Monetary Fund, Organisation for Economic Co-operation and Development, United Nations, World Bank).

    Google Scholar 

  • Walras, L. 1874. Elements of Pure Economics (trans: Jaffé, W.). (Richard D. Irwin, Homewood IL, 1954).

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Appendices

Appendix A: Decompositions of Time-Series Depreciation

Time-series depreciation of an asset of type i and age j over period t is, according to expression (3.6), defined as \(P_{i,j-0.5}^{t^-} - P_{i,j+0.5}^{t^+}\), which is the (nominal) value change of the asset between beginning and end of the period. This value change combines the effect of the progress of time, from t to t +, with the effect of ageing, from j − 0.5 to j + 0.5. Since value change is here measured as a difference, a natural decomposition of time-series depreciation into these two effects is

$$\displaystyle \begin{aligned} \begin{array}{rcl} & &\displaystyle {P_{i,j-0.5}^{t^-} - P_{i,j+0.5}^{t^+} =} \\ & &\displaystyle \qquad (1/2)\left[(P_{i,j-0.5}^{t^-} - P_{i,j-0.5}^{t^+}) + (P_{i,j+0.5}^{t^-} - P_{i,j+0.5}^{t^+})\right] + \\ & &\displaystyle \qquad (1/2)\left[(P_{i,j-0.5}^{t^-} - P_{i,j+0.5}^{t^-}) + (P_{i,j-0.5}^{t^+} - P_{i,j+0.5}^{t^+})\right]. {} \end{array} \end{aligned} $$
(3.79)

This decomposition is symmetric; it has the same structure as the decomposition of a value difference into Bennet indicators. The first term at the right-hand side of the equality sign measures the effect of the progress of time on an asset of unchanged age; this is called revaluation. The revaluation, as measured here, is the arithmetic mean of the revaluation of a j − 0.5 periods old asset and a j + 0.5 periods old asset, and may be said to hold for a j periods old asset.

The second term c oncerns the effect of ageing, which is measured by the price difference of two, otherwise identical, assets that differ precisely one period in age. This is called Hicksian or cross-sectional depreciation. The arithmetic mean is taken of cross-sectional depreciation at beginning and end of the period, and, hence, may be said to hold at the midpoint of period t.

The ratio-type counterpart of expression (3.79) is a decomposition into two Fisher-type indices,

$$\displaystyle \begin{aligned} \frac{P_{i,j+0.5}^{t^+}}{P_{i,j-0.5}^{t^-}} = \left[\frac{P_{i,j-0.5}^{t^+}}{P_{i,j-0.5}^{t^-}}\frac{P_{i,j+0.5}^{t^+}}{P_{i,j+0.5}^{t^-}}\right]^{1/2} \left[\frac{P_{i,j+0.5}^{t^+}}{P_{i,j-0.5}^{t^+}}\frac{P_{i,j+0.5}^{t^-}}{P_{i,j-0.5}^{t^-}}\right]^{1/2}. \end{aligned} $$
(3.80)

The first term at the right-hand side of the equality sign measures revaluation. The second ter m measures cross-sectional depreciation. As one sees, revaluation depends on age, and cross-sectional depreciation depends on time. In the usual model, these two dependencies are assumed away. Revaluation is approximated by \(P_{i}^{t^+}/P_{i}^{t^-}\), the price change of a new asset of type i from beginning to end of period t. Cross-sectional depreciation is approximated by 1 − δ ij, where δ ij is the (positive) percentage of annual depreciation applying to an asset of type i and age j. The specific formulation highlights the fact that ageing usually diminishes the value of an asset.

Under these two assumptions, the basic time-series depreciation model for an asset of type i and age j, over period t, is given by

$$\displaystyle \begin{aligned} \frac{P_{i,j+0.5}^{t^+}}{P_{i,j-0.5}^{t^-}} = \frac{P_{i}^{t^+}}{P_{i}^{t^-}}(1 - \delta_{ij}) \: \:(j=1,\ldots ,J_i). \end{aligned} $$
(3.81)

For assets that are acquired at the midpoint of period t it is useful to distinguish between new and used assets. Over the second half of period t, the model reads

$$\displaystyle \begin{aligned} \begin{array}{rcl}{} \frac{P_{i,0.5}^{t^+}}{P_{i,0}^{t}} & = &\displaystyle \frac{P_{i}^{t^+}}{P_{i}^{t}}(1 - \delta_{i0}) \\ \frac{P_{i,j+0.5}^{t^+}}{P_{i,j}^{t}} & = &\displaystyle \frac{P_{i}^{t^+}}{P_{i}^{t}}(1 - \delta_{ij}/2) \: \:(j=1,\ldots ,J_i), \end{array} \end{aligned} $$
(3.82)

where (1 − δ ij∕2) serves as an approximation to (1 − δ ij)1∕2. The percentage of annual depreciation, δ ij, ideally comes from an empirically estimated age-price profile for asset-type i. Under a geometric profile one specifies δ i0 = δ i∕2 and δ ij = δ i (j = 1, …, J i).

Replacing \(P_{i,j+0.5}^{t^+}\) by \(\mathcal {E}^{t^-}P_{i,j+0.5}^{t^+}\), \(P_{i,0.5}^{t^+}\) by \(\mathcal {E}^{t}P_{i,0.5}^{t^+}\), and \(P_{i,j+0.5}^{t^+}\) by \(\mathcal {E}^{t}P_{i,j+0.5}^{t^+}\) delivers expressions for anticipated time-series depreciation.

Appendix B: Geometric Profiles

Consider the fundamental asset price (equilibrium) equation in expression (3.29). To avoid notational clutter it is assumed that there is no tax. It is further assumed that the future interest rate is constant; that is, r t+j = r (j = 0, …, J i). Then

$$\displaystyle \begin{aligned} P_{i,j-0.5}^{t^-} = \frac{u_{ij}^{t}}{1+r} + \frac{u_{i,j+1}^{t+1}}{(1+r)^{2}} + \ldots + \frac{u_{i,J_{i}}^{t+J_{i}-j}}{(1+r)^{J_{i}-j+1}}. \end{aligned} $$
(3.83)

A geometric age-efficiency profileFootnote 19 means that \(u_{ij}^{t}/u_{i1}^{t} = \phi _{ij} = (1 - \delta _i)^{j-1}\) (i = 1, …, I;j = 1, …, J i). Equivalently, \(u_{i,j+1}^{t}/u_{ij}^{t} = 1 - \delta _i\). By substituting this into expression (3.83) one obtains

$$\displaystyle \begin{aligned} P_{i,j-0.5}^{t^-} = \frac{u_{ij}^{t}}{1+r} + \frac{u_{ij}^{t+1}(1-\delta_i)}{(1+r)^{2}} + \ldots + \frac{u_{ij}^{t+J_{i}-j}(1-\delta_i)^{J_{i}-j}}{(1+r)^{J_{i}-j+1}}. \end{aligned} $$
(3.84)

For a one-period older asset of the same type one then has, similarly but with one term less,

$$\displaystyle \begin{aligned} P_{i,j+0.5}^{t^-} = \frac{u_{i,j+1}^{t}}{1+r} + \frac{u_{i,j+1}^{t+1}(1-\delta_i)}{(1+r)^{2}} + \ldots + \frac{u_{i,j+1}^{t+J_{i}-j-1}(1-\delta_i)^{J_{i}-j-1}}{(1+r)^{J_{i}-j}}. \end{aligned} $$
(3.85)

Using again the geometric-profile assumption, expression (3.85) can be written as

$$\displaystyle \begin{aligned} P_{i,j+0.5}^{t^-} = \frac{u_{ij}^{t}(1-\delta_i)}{1+r} + \frac{u_{ij}^{t+1}(1-\delta_i)^2}{(1+r)^{2}} + \ldots + \frac{u_{ij}^{t+J_{i}-j-1}(1-\delta_i)^{J_{i}-j}}{(1+r)^{J_{i}-j}}, \end{aligned} $$
(3.86)

or

$$\displaystyle \begin{aligned} \begin{array}{rcl}{} P_{i,j+0.5}^{t^-} {=} (1-\delta_i)\left( \frac{u_{ij}^{t}}{1+r} + \frac{u_{ij}^{t+1}(1-\delta_i)}{(1+r)^{2}} + \ldots + \frac{u_{ij}^{t+J_{i}-j-1}(1-\delta_i)^{J_{i}-j-1}}{(1+r)^{J_{i}-j}}\right).\\ \end{array} \end{aligned} $$
(3.87)

The part between brackets appears to be identical to \(P_{i,j-0.5}^{t^-}\) minus its last term, so that

$$\displaystyle \begin{aligned} P_{i,j+0.5}^{t^-} = (1-\delta_i)\left(P_{i,j-0.5}^{t^-} - \frac{u_{i,J_{i}}^{t+J_{i}-j}}{(1+r)^{J_{i}-j+1}}\right). \end{aligned} $$
(3.88)

Now \(u_{i,J_{i}}^{t+J_{i}-j}\) is the user cost of an asset, of type i that is j years old in period t, in the final year of its (economic) existence. If J i → then this user cost will expectedly tend to 0. The denominator \((1+r)^{J_{i}-j+1}\) will then tend to . Together it is reasonable to expect that

$$\displaystyle \begin{aligned} P_{i,j+0.5}^{t^-} \approx (1-\delta_i)P_{i,j-0.5}^{t^-}, \end{aligned} $$
(3.89)

which means that the age-price profile of the asset is approximately geometric.

Reversely, consider expression (3.25) without the tax component:

$$\displaystyle \begin{aligned} u_{ij}^{t} = (1+r^{t})P_{i,j-0.5}^{t^-} - P_{i,j+0.5}^{t^+} \: (j=1,\ldots ,J_{i}). \end{aligned} $$
(3.90)

For a one-period older asset of the same type then

$$\displaystyle \begin{aligned} u_{i,j+1}^{t} = (1+r^{t})P_{i,j+0.5}^{t^-} - P_{i,j+1.5}^{t^+}. \end{aligned} $$
(3.91)

Assume now a geometric age-price profile; that is, \(P_{ij}^{t}/P_{i0}^{t} = \varphi _{ij} = (1 - \delta _i)^{j}\) (i = 1, …, I;j = 1, …, J i). Equivalently, \(P_{i,j+1}^{t}/P_{ij}^{t} = 1 - \delta _i\). Substituting this into expression (3.91) delivers

$$\displaystyle \begin{aligned} u_{i,j+1}^{t} = (1+r^{t})(1 - \delta_i)P_{i,j-0.5}^{t^-} - (1 - \delta_i)P_{i,j+0.5}^{t^+} = (1 - \delta_i)u_{ij}^{t}, \end{aligned} $$
(3.92)

which means that the age-efficiency profile of the asset is geometric.

Rights and permissions

Reprints and permissions

Copyright information

© 2021 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this chapter

Check for updates. Verify currency and authenticity via CrossMark

Cite this chapter

Balk, B.M. (2021). Capital Input Cost. In: Productivity. Contributions to Economics. Springer, Cham. https://doi.org/10.1007/978-3-030-75448-8_3

Download citation

Publish with us

Policies and ethics