Skip to main content

Advertisement

SpringerLink
Log in

Windfall Resource Income, Productivity Growth, and Manufacturing Employment

  • Research Article
  • Published:
Open Economies Review Aims and scope Submit manuscript

Abstract

I study the independent impacts of windfall income from non-renewable resource exports and productivity growth on the changing share of employment in manufacturing in an open economy. The framework includes non-unitary income and substitution elasticities, so that both windfall income and productivity growth lead to sectoral reallocation of labor. I use the model to account for the declining share of manufacturing employment in Canada. I find that the relative importance of these factors has varied over time. While productivity growth is responsible for a substantial fraction of the decline in the share of manufacturing employment since 1960, the windfall income from the booming resource sector contributed to this decline significantly during the 2000s, when the Canadian terms of trade improved rather dramatically.

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

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3

Similar content being viewed by others

Notes

  1. While the declining employment levels in manufacturing in industrial countries is a relatively new theme (e.g., Pierce and Schott (2012)), the declining share of employment in manufacturing has attracted attention at least since the early 1980s (e.g., Baumol et al. 1989).

  2. Newfoundland and Saskatchewan are the two other Canadian provinces with large extractive industries. The boom in Alberta oil sands has also generated a spirited debate, because of its negative environmental and health effects. Oil-sands in Alberta, if extraction continues as planned, will irreversibly transform boreal landscape equivalent in surface area to contemporary Greece.

  3. While, in this paper, I focus on the interaction between productivity growth and windfall resource income, the employment in manufacturing has been an ongoing phenomenon in many industrial economies without any natural resource booms. So, it is natural to think that there are factors common to both types of countries underlying these broader trends. In fact, the productivity-growth channel studied in this paper is no doubt more important in economies without a relatively large resource sector (e.g., İşcan (2010)).

  4. Matsuyama (2009) also argues for thinking about structural change in an open economy context.

  5. One issue I do not directly address is the response of international trade to income. Sauré (2012) argues that if marginal utility of individual consumption categories is bounded, rising productivity or declining transportation costs can result in a higher share of international trade in income. Ceglowski (2014) argues that although there has been a sharp increase in the income elasticity of international trade, growing vertical specialization may account for this trend.

  6. Modelling only agriculture, manufacturing, and market services is standard in the literature (e.g., Echevarria 1997; İşcan 2010; Uy et al. 2013).

  7. Ricardian production structure is standard in models that emphasize sectoral allocation of labor; see, e.g., Kuralbayeva and Stefanki (2013), Matsuyama (2009), and Uy et al. (2013). In particular, physical capital accumulation is unlikely to matter for labor reallocation across sectors and only differential (across sectors) capital deepening matters (Acemoglu and Guerrieri 2008). However, even if sectors have varied rates of capital deepening, this effect tends to have an economically negligible impact on the rate of labor reallocation across sectors (Dennis and İşcan 2009). An additional and a more practical motivation for the Ricardian model is the difficulty associated with the available sectoral total factor productivity (TFP) estimates for Canada. In the official data, TFP growth in personal services is significantly negative and large throughout the sample; see the online supplementary material, Fig. S3. This sustained negative TFP growth is difficult to explain.

  8. Figure 1c below shows that the employment share of mining in Canada has been small and stable over time. Therefore, it is reasonable to ignore employment in this sector in a model of sectoral labor reallocation. For similar reasons, while natural resource output is likely to respond to changes in its price, by modelling this industry as an enclave, I keep such responses in the background.

  9. The major consumption categories I use are primarily distinguished by the differences in their income elasticity of demand. Demand studies have consistently shown that the income elasticity of demand for food is significantly less than those for services (e.g., Taylor and Houthakker 2010, Chp. 15). Even seemingly overlapping consumption categories across sectors exhibit these differences. For instance, the income elasticity of demand that is most relevant for agriculture (food at home) is significantly lower than that for services (food away from home) (Taylor and Houthakker 2010), providing empirical support for the view that industries can be distinguished based on their income elasticities of (derived) demand. Moreover, income elasticity of demand for different commodities and goods originating from sub-sectors within agriculture and manufacturing tend to be similar. This is consistent with the traditional industry classifications used in the literature on structural change (e.g., Clark 1957; Kuznets 1966)

  10. These assumptions about international trade focus the impact of mining on manufacturing employment, an issue that has attracted the most attention. While unitary elasticity of substitution between home and foreign manufactured goods is possibly special, it leads to an analytic solution involving the terms of trade.

  11. Consistent with the original booming natural resource sector literature, the spending and resource reallocation effects are both driven by changes in real income from mining and q, which is the counterpart of the “real exchange rate” in the model. Of course, resource reallocation effects in the data may also be driven by changes in the exchange rate that are unrelated to those effects considered here. While modelling such short-run deviations from the purchasing power parity exchange rates is beyond the scope of this paper, non-homothetic preferences (Bergstrand 1991), non-traded goods (Balassa-Samuelson effect), and the terms-of trade effects modelled here help explain these deviations.

  12. The income effect Δ t contains the term (P o t Y o t /L t )/P m t z m t , which is the ratio of mining value added divided by total employment to manufacturing value added divided by manufacturing employment (P o t Y o t /L t )/(P m t Y m t /L m t ).

  13. Kuralbayeva and Stefanki (2013) make a similar point in a related model.

  14. Canada trades extensively with the United States (about 80% of its foreign trade). Over the sample period Canada had a (small) international trade surplus in agricultural goods. This surplus shrank over time. In 1961, the export share of agricultural goods was about 25 percent, and their import share was 14 percent. By 2010, the export share had declined to 10 percent, and the import share stood at 8 percent. These data are from Cansim Table 383-0028.

  15. From a resource allocation standpoint, it is equally interesting to consider total hours worked by industry. In the online supplementary material, I document that within these three industries indexes of employment and total hours worked are almost identical, except for agriculture at the beginning of the sample period.

  16. In the online supplementary material, I extend the employment data back into the nineteenth century and find that the level of agricultural employment peaks sometime by the WWII, the share of employment in manufacturing exhibits a hump-shape peaking by the 1950s, and a secular increase in the share of employment in tertiary industries throughout.

  17. Figure 1c also shows that finance is the only other industry providing market services that has exhibited significant labor reallocation (absorption, in this case) over the sample period.

  18. In the context of mining, the windfall income concept is also appropriate from another perspective: small open economies typically do not have any control over the international price of the resource, though they may have non-negligible control over its production.

  19. The sample correlation coefficient between the manufacturing terms of trade and the terms of trade (ToT4 defined in Appendix A) is −.52. Notice also that export price indexes for Canada include (and are heavily influenced by) commodity and energy prices.

  20. Since not all mining output is exported to the rest of the world, an alternative approach would be to deduct domestic consumption from domestic output. The difficulty here is reminiscent of the problem that arises in the context of value added and final demand. For instance, Canada produces and exports crude oil, but consumes gasoline and heating oil.

  21. In particular, let \(l^{\mathrm {s}}_{it}(\nu ,\gamma _{i})\) be the model-based series corresponding to a particular choice of ν and γ i ’s, and calculate

    $$\text{RMSE}(\nu,\gamma_{i}) = {\sum}_{t}{\sum}_{i} \left( l^{\mathrm{s}}_{it}(\nu,\gamma_{i}) - l_{it}\right)^{2}. $$

    I report the simulation results corresponding to those parameter combinations that deliver the lowest RMSE for a given model.

  22. Since both the actual and model-based data are in terms of share of employment by sector, restricting attention to manufacturing is still informative (these shares must add up to one). Also, much of the booming resource sector literature has used manufacturing employment to assess resource reallocation effects of commodity booms and busts. Note that total employment in Canadian manufacturing increased until 2000s.

  23. The changes in employment reported in the text are calculated as follows: Let the gross rate of growth in the model-based share of employment in manufacturing given by

    $$\frac{{l^{s}}_{mt}}{l^{s}_{m,t-1}} = \frac{L^{s}_{mt}/L_{t}}{L_{m,t-1}/L_{t-1}}. $$

    The online supplementary material, Table S1 reports the changes in manufacturing employment by decade and for alternative models.

  24. In Appendix A, I discuss the construction of \(P^{*}_{m}\) using fragmented data on manufacturing import prices. These data suggest that using the import price deflator most likely leads the model to overstate the rise in \(P^{*}_{m}\) from 1971 to 2000 (thus overstate the decline in q), and to understate the fall in \(P^{*}_{m}\) from 2000 to 2008 (thus overstate the decline in q). So, while it is not possible to construct a consistent time series from these fragmented data, it is possible that these biases are in part responsible for the gap between the actual and model-based data, especially those in the 1970s and 2000s (Fig. 3b).

  25. Recall that the baseline parameter value is μ=0.569, and \(\mu \rightarrow 1\) amounts to turning off international trade.

  26. For such theoretical analyses, see, for instance, Matsuyama (1992), and Matsen and Torvik (2005) in different contexts.

  27. A comprehensive treatment of trade agreements requires a multi-country, multi-sector approach, as there are both trade creation and trade diversion effects associated with these agreements, and highly heterogeneous responses by sectors and countries.

  28. Another possibility is the comparative advantage effects identified by Matsuyama (2009), where faster productivity growth in the Canadian automotive industry relative to that of the United States would lead to an increase in Canadian automotive employment and a decrease in U.S. automotive employment. There is some suggestive evidence for this channel: in Canada, the share of manufacturing employment in transport equipment rose significantly during the 1960s (data not reported here). To be precise about the magnitude of such effects, one would require a multi-country general equilibrium model.

  29. Transportation is another important component of trade costs. As in the cases of rising productivity and falling tariffs, a decline in transportation costs would tend to increase international trade, and may even increase the share of foreign trade in total income (e.g., Sauré 2012). It is an empirical question whether declining transportation costs had differential effects across countries on the sectoral allocation of labor.

  30. I do not pursue a full quantitative analysis of the Australian data for a number of reasons. First, there is substantially less time-series data available on the sectoral allocation of labor in Australia and than in Canada. Second, since the early 1980s, the natural resource sector in Canada has been dominated by oil, which provides a closer match between the model (or the booming sector literature) and the data, whereas the Australian resource sector has been considerably more diversified.

  31. Pierce and Schott (2012) argue that granting of permanent normal trade relations status to China in late 2000 had a substantial (one-time) impact on the pace of employment losses in U.S. manufacturing coinciding with the 2001 recession—though, they do not discuss the sources of long-term decline in manufacturing. At the same time, Bernard and Fort (2013) discuss the recent rise of the “factoryless goods producers,” which are classified in the wholesale trade, and argue that the reclassification of employment over time from manufacturing to wholesale trade may be partly responsible for the declining share of employment in U.S. manufacturing.

  32. The online supplementary material also shows that the share of employment in agriculture in Canada has at times been significantly larger than that of the United States (Fig. S6C). So, the aggregate data suggests that Canada has a relatively more resource-based (including agriculture) economy, and allocates a larger share of its labor force to resource-based industries. Almon and Tang (2011) compare the share of hours worked by industry in Canada and the United States since 2000 with trends similar those of employment.

References

  • Acemoglu D, Guerrieri V (2008) Capital deepening and nonbalanced economic growth. J Polit Econ 116:467–498

    Article  Google Scholar 

  • Almon M-J, Tang J (2011) Industrial structural change and the post-2000 output and productivity growth slowdown: A Canada–U.S. comparison. Int Product Monit 22(Fall):44–81

    Google Scholar 

  • Baldwin JR, Macdonald R (2009) The Canadian manufacturing sector: Adapting to challenges. Technical Report 11F0027M – No. 057, Statistics Canada, Economic Analysis (EA) Research Paper Series, Ottawa

  • Baumol WJ, Blackman SAB, Wolff EN (1989) Productivity and American leadership: the long view. MIT Press, Cambridge

    Google Scholar 

  • Beine M, Bos CS, Coulombe S (2012) Does the Canadian economy suffer from Dutch Disease? Unpublished manuscript, University of Ottawa

  • Bergstrand JH (1991) Structural determinants of real exchange rates and national price levels: Some empirical evidence. Am Econ Rev 81(1):325–334

    Google Scholar 

  • Bernard AB, Fort TC (2013) Factoryless goods producters in the US. Working Paper No. 19396, National Bureau of Economic Research

  • Bevan D, Collier P, Gunning JW (1992) ICS Press, San Francisco

  • Bodman PM, Crosby M (2000) Phases of the Canadian business cycle. Can J Econ 33(3):618–633

    Article  Google Scholar 

  • Ceglowski J (2014) Has trade become more responsive to income? Assesing the evidence for US imports. Open Econ Rev 25(2):225–241

    Article  Google Scholar 

  • Charnavoki V, Dolado JJ (2014) The effects of global shocks on small commodity-exporting economies. Am Econ J Macroecon 6(2):207–237

    Article  Google Scholar 

  • Clark C (1957) The conditions of economic progress, 3rd edn. Macmillan, London

  • Corden WM (1984) Booming sector and Dutch disease economics: Survey and consolidation. Oxf Econ Pap 36:359–380

    Google Scholar 

  • Dennis BN, İşcan TB (2006) Terms of trade risk with partial labor mobility. J Int Econ 68:92–114

    Article  Google Scholar 

  • Dennis BN, İşcan TB (2009) Engel versus Baumol: Accounting for U.S. structural change using two centuries of data. Explor Econ Hist 46:186–202

    Article  Google Scholar 

  • Diewert WE, Yu E (2012a) New estimates of real income and multifactor productivity growth for the Canadian Business Sector, 1961–2011. Int Product Monit 24:27–48

    Google Scholar 

  • Diewert WE, Yu E (2012b) New estimates of real income and multifactor productivity growth for the Canadian Business Sector, 1961–2011. Online Appendix 2. Int Product Monit 24

  • Echevarria C (1997) Changes in sectoral composition associated with economic growth. Int Econ Rev 38(2):431–452

    Article  Google Scholar 

  • Egert B, Leonard CS (2008) Dutch disease scare in Kazakhstan: Is it real?. Open Econ Rev 19:147–165

    Article  Google Scholar 

  • EU KLEMS (2012) EU KLEMS growth and productivity accounts

  • Firestone OJ (1958). Bowes & Bowes, London

  • Fuss MA, Waverman L (1992) Costs and productivity in automobile production: the challenge of japanese efficiency. Cambridge University Press, Cambridge

    Book  Google Scholar 

  • Gaston N, Trefler D (1997) The labour market consequences of the Canada–U.S. Free Trade Agreement. Can J Econ 30:18–41

    Article  Google Scholar 

  • Gavin M (1990) Structural adjustment to a terms of trade disturbance: The role of relative prices. J Int Econ 28:217–243

    Article  Google Scholar 

  • Gu W (2012) Estimating capital input for measuring business sector multifactor productivity growth in Canada: Response to Diewert and Yu. Int Product Monit 24:49–61

    Google Scholar 

  • Gu W, Ho MS (2000) A comparison of industrial productivity growth in Canada and the United States. Am Econ Rev Pap Proc 90 (2):172–175

    Article  Google Scholar 

  • İşcan TB (2010) How much can Engel’s law and Baumol’s disease explain the rise of service employment in the United States? B.E. J Macroecon (Contributions) 10(article 26)

  • Kehoe TJ, Ruhl KJ, Steinberg JB (2013) Global imbalances and structural change in the United States. Working Paper No. 19339, National Bureau of Economic Research

  • Kollmeyer C (2009) Explaining deindustrialization: How affluence, productivity growth, and globalization diminish manufacturing employment. Am J Soc 114(6):1644–1674

    Article  Google Scholar 

  • Kongsamut P, Rebelo S, Xie D (2001) Beyond balanced growth. Rev Econ Stud 68:869–882

    Article  Google Scholar 

  • Kuralbayeva K, Stefanki R (2013) Windfalls, structural transformation and specialization. J Int Econ 90:273–301

    Article  Google Scholar 

  • Kuznets S (1966) Yale University Press, New Haven

  • Matsen E, Torvik R (2005) Optimal Dutch disease. J Dev Econ 78:494–515

    Article  Google Scholar 

  • Matsuyama K (1992) Agricultural productivity, comparative advantage, and economic growth. J Econ Theory 58:317–334

    Article  Google Scholar 

  • Matsuyama K (2009) Structural change in an interdependent world: A global view of manufacturing decline. J Eur Econ Assoc 7:478–486

    Article  Google Scholar 

  • Ngai RL, Pissarides CA (2007) Structural change in a multisector model of growth. Am Econ Rev 97:429–443

    Article  Google Scholar 

  • Nordhaus WD (2008) Baumol’s diseases: A macroeconomic perspective. B.E. J Macroecon (Contributions), 8(article 9)

  • Obstfeld M, Rogoff K (1996) Foundations of international macroeconomics. MIT Press, Cambridge

    Google Scholar 

  • OECD (1963) Manpower Statistics, 1950–1962. Paris

  • OECD (1965) Manpower Statistics, 1954–1964. Paris

  • O’Mahony M, Timmer MP (2009) Output, input and productivity measures at the industry level: The EU KLEMS Database. Econ J 119:F374–F403

    Article  Google Scholar 

  • Pierce JR, Schott PK (2012) The surprisingly swift decline of U.S. manufacturing employment. Working Paper No. 18655, National Bureau of Economic Research

  • Sauré P (2012) Bounded love of variety and patterns of trade. Open Econ Rev 23(4):645–674

    Article  Google Scholar 

  • Sturgeon TJ, Memedovic O (2011) Mapping global value chains: Intermediate goods trade and structural change in the world economy. Technical Report 05/2010, UNIDO, Development Policy and Strategic Research Branch, Vienna

  • Taylor LD, Houthakker HS (2010) Consumer demand in the united states: prices, income, and consumption behavior, 3 edn. Springer, Dordrecht

    Book  Google Scholar 

  • Triplett JE, Bosworth BP (2004) Brookings Institution, Washington

  • Uy T, Yi K-M, Zhang J (2013) Structural change in an open economy. J Monet Econ 60:667–682

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Economics, Dalhousie University, 6214 University Avenue, P.O. Box 15000, Halifax, NS, B3H 4R2, Canada

    Talan B. İşcan

Authors
  1. Talan B. İşcan
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Talan B. İşcan.

Additional information

I thank two anonymous referees, the editor, Mario Crucini, Andrea Giusto, Ryan McDonald, and Lars Osberg for detailed comments, Ryan McDonald and Wulong Gu for guiding me to data sources, Natàlia Díaz-Insensé for editorial suggestions, and Universidad Carlos III de Madrid for its hospitality where the first draft of this paper was completed. This research was not externally funded. The author declares no competing financial interests.

Electronic supplementary material

Below is the link to the electronic supplementary material.

(PDF 462 KB)

Appendices

Appendix A: Data Sources

Statistics Canada KLEMS Productivity Accounts Database. For reasons unknown to me, employment (“number of persons engaged”) by industry data are available until 2010, whereas value added by industry end in 2008. Table 3 shows the detail of aggregation at the major industry level.

Table 3 Industry classification in EU KLEMS
Full size table

Unfortunately, the empirical counterpart of P m /P ∗ is not easy to construct. The methodology pursued to construct this series involves assumptions and is based on several data sources. In particular, the index of import prices \(P_{m}^{*}\) is the import deflator from Cansim table 383-0027. For domestic manufacturing prices P m , I constructed an implicit “export” price deflator based on value-added price indexes from KLEMS database and export shares from Cansim table 380-0027. The price index contains a composite machinery and equipment price index from the price indexes of machinery, and electrical and optical equipment, using equal weights. The export deflator is a weighted average of price indexes of machinery and equipment, and transport equipment with weights given by their export shares. However, export share data start in 1971, and export shares for 1961–1970 are assumed to be equal to those in 1971. Also, since price indexes are missing in the KLEMS database for 2009 and 2010, year-to-year changes in export prices for machinery and transport equipment (SITC 7) from Cansim table 228-0052 (Paasche current weighted) are used to extend the P m series until 2010. The constructed series are labelled as q in Fig. 2.

Also, for comparison purposes the following series have been constructed.

  • 1961–1975 (ToT1): The index of export prices divided by the index of import prices (1971=100) from the Historical Statistics of Canada, Series G388.

  • 1971–2000 (ToT2): The export price index for other manufactured goods (Paasche, 1992=100) divided by the price index for total merchandise imports (Paasche, 1992=100) from Cansim tables 176-0006 and 176-0007, respectively. This manufacturing price index excludes motor vehicles and parts, fertilizers, pulp and paper, and sawmill products.

  • 2002–2010 (ToT3): The export price index for machinery and transport equipment (Paasche, 2002=100) divided by the price index for merchandise imports (Paasche, 2002=100) from Cansim tables 228-0052 and 380-0027, respectively.

  • 1961–2010 (ToT4): The index of export prices divided by the index of import prices from Cansim table 383-0027. This export price index includes crude oil and mineral ores.

Figure 4 presents the growth rate of these alternative terms of trade series. The coverage of ToT1 is in principle identical to that of ToT4, and ToT2 and ToT3 are closest in spirit to manufacturing terms of trade q in the model. However, it is difficult to reconcile and parse ToT2 and ToT3 into a coherent series, and extend back to 1961 without more detailed data. Hence, the analysis uses manufacturing terms of trade as described above. Note that as in the case of constructed manufacturing terms of trade, for overlapping years, there are significant differences in the growth rates of ToT2 and ToT3 relative to those of ToT4.

Fig. 4
figure 4

Annual growth rate of the Canadian terms of trade, 1961–2010. ToT1 is the export price index divided by the import price index. ToT2 is the export price index for other manufactured goods divided by price index for total merchandise imports. ToT3 is the export price index for manufactured goods divided by the price index for merchandise imports. ToT4 is the index of export prices divided by the index of import prices. gToT is the average annualized growth rate for the corresponding measure of ToT. The figures in parentheses indicate the years covered by each series.Sources: Historical Statistics of Canada, Series G388, Cansim tables 176-0006, 176-0007, 228-0055, 380-0027, and 383-0027

Full size image

Finally, Diewert and Yu (2012a) use parsed import and export price indices for 8 export and 7 import categories essentially using the above mentioned series. I constructed a manufacturing export (Fisher) price index and a manufacturing import price index using three of their product categories (machinery and equipment, automotive products, consumer goods). The correlation between q defined above and the manufacturing terms of trade based on Diewert and Yu (2012b) data is −0.623. Given this substantial difference, I use these estimates in the sensitivity analysis.

In all the measures of terms of trade reported above and in constructing the manufacturing terms of trade q, due to lack of data, I use the price deflator for merchandise imports. However, conceptually, the manufacturing terms of trade requires a price deflator for manufacturing imports. The closest empirical counterpart of this series I can construct is based on two non-overlapping series; (1) a geometric average of import price indexes for “motor vehicles and parts” and “machinery and equipment” from 1971 to 2000 from Cansim table 176-0007 (Paasche price index, 1992=100), and (2) a geometric average of import price indexes for “machinery and transport equipment” from Cansim table 228-0052 (Paasche current weighted, 2002=100) and “other consumer goods” from Cansim table Table 228-0055 (Paasche current weighted, 2002=100) from 2002 to 2010. (Unfortunately, I was unable to find comparable data prior to 1971.) A comparison of this import price deflator for “manufactured” goods with the broader import deflator indicates a significantly slower import price inflation from 1971 until 2000, and a significantly faster import price deflation from 2002 until 2008. These suggest that the overall decline in the manufacturing terms of trade q from 1971 to 2000 was probably less than what is indicated in Fig. 2, and the manufacturing terms of trade from 2002 to 2008 did not exhibit a significant deterioration—a pattern that would be consistent with the terms of trade in Fig. 2a. For the open economy model, this would imply lower manufacturing employment during the 1971–2000 period and higher manufacturing employment during the 2002–2008 period.

The empirical counterpart of this variable is constructed using value added in mining and quarrying in the EU KLEMS database. While this variable has the advantage of being consistent with the rest of the value added concepts in the analysis, it requires an adjustment using the balance of international trade on relevant items to arrive at a concept that is net of domestic consumption. Unfortunately, available data do not cover the entire period.

  • 1961–1975: The difference between exports and imports of crude materials (inedible) from the Historical Statistics of Canada, Series G419, G433.

  • 1971–2010: The difference between exports and imports of energy products from Cansim table 380-0027. Alternative estimates include the difference between the exports and imports of industrial goods and materials, which includes metals and metal ores.

In the absence of more detailed data, I constructed a single time series by backcasting the series on energy products using the series on crude materials.

Imports of goods (current prices) excluding “special transactions” and “other balance of payments adjustments” all from Cansim table 380-0027 are used. Since a non-negligible fraction of these imports is intermediate inputs, an adjustment is made to arrive at the imports of final goods. The adjustment for manufactured intermediate inputs trade is based on the estimates of Sturgeon and Memedovic (2011, Table 2), which estimates this trade at about 300 billion US dollars for Canada in 2006. In converting this to a total fraction of total trade in goods, average exchange rate of 1.15 CAN$/US$ and total foreign trade of 857 billion dollars (exports and imports of goods) from Cansim Table 380-0027 were used. This suggests that roughly 40 percent of total trade in goods is in manufactured intermediate goods. The remaining are labeled as final goods.

Balance on trade, goods (current prices) from Cansim table 380-0027. In the baseline model, this variable is restricted to zero. This variable is only used to check the sensitivity of μ to nonzero trade balance (see Appendix B).

B: Calibration

Previous research (e.g., İşcan 2010) finds that low complementarity across consumption categories provides the empirically most plausible estimate for the United States. Based on these estimates, a grid search for values of ν∈[0.025−−0.975] is conducted. The value of ν that yields the lowest root mean squared error between the realized labor shares by sector and the labor shares from the numerical solution of the model is used.

These are calibrated to match the long-run (in 2010) shares of personal expenditures, and are calculated as explained in Appendix A (income and expenditure accounts).

Calibrated to match the share of employment in each sector in 1961.

This is calibrated to match the sample average of the following series

$$ 1 - \frac{P^{*}_{m}M^{*}}{P_{o}Y_{o}+P_{m}Y_{m}-NX}, $$
(B.1)

where \(P^{*}_{m}M^{*}\) is imports of goods (non-food), P o Y o is mining value added, P m Y m is non-food value added, and NX is net exports, which according to the model is zero. This calibration target follows from the balanced trade identity and the first-order condition to the utility maximization problem

$$\begin{array}{@{}rcl@{}} P_{o}Y_{o}+P_{m}Y_{m}-P_{m}M - P^{*}_{m}M^{*} = 0, \end{array} $$
(B.2)
$$\begin{array}{@{}rcl@{}} \frac{P_{m}M}{P^{*}_{m}M^{*}} = \frac{\mu}{1-\mu}. \end{array} $$
(B.3)

The values of μ for different configurations of data are as follows; see also Appendix A. For nonzero NX and \(P^{*}_{m}M^{*}\) including “balance of payments adjustments” μ 1=0.474, and \(P^{*}_{m}M^{*}\) excluding “balance of payments adjustments μ 2=0.495. For N X=0 and \(P^{*}_{m}M^{*}\) including “balance of payments adjustments” μ 3=0.553, and \(P^{*}_{m}M^{*}\) excluding “balance of payments adjustments” μ 4=0.569. The baseline parametrization uses μ 4.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

İşcan, T.B. Windfall Resource Income, Productivity Growth, and Manufacturing Employment. Open Econ Rev 26, 279–311 (2015). https://doi.org/10.1007/s11079-014-9330-z

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11079-014-9330-z

Keywords

JEL Classification

Search

Navigation