Notes

^1.

Similar results have been reported for the United States (Doms and Jensen 1998), the United Kingdom (Conyon et al. 2002) and Indonesia (Takii 2004).

^2.

During the period, the file was essentially a census of all plants—with the smallest plants being covered with administrative tax files. It is only post-2003 that the file has become a sample survey.

^3.

For further descriptions of the file, see Baldwin (1995) and Baldwin, Beckstead and Girrard (2002).

^4.

Ibid.

^5.

Because of the care taken to maintain both plant and firm identifiers and the annual coverage of the survey, this file offers much better coverage than some others that have been used to study the impact of mergers. Other studies—(McGuckin and Nguyen 1995; Lichtenberg and Siegel 1992)—use the U.S. Longitudinal Research Database (LRD), the disadvantage being that neither the frame nor the register was kept up to date on an annual basis. Because of the difficulty in creating longitudinal firm identifiers, many of the LRD studies have been restricted to individual industries like food or meat-packing.

^6.

International comparisons need to be cognizant of the fact that firm identifiers can be defined at different levels—a first-level legal entity or a higher-level consolidated entity. Differences exist across countries in this practice that are often linked to differences in legal systems and practice.

^7.

Baldwin and Gu (2005) point out that Canadian multinationals tend to be larger, more capital intensive, pay higher wages, and be more likely to be more innovative.

^8.

Jarrad (2005) finds that regulatory and technological shocks drive industry merger waves.

^9.

If a foreign firm takes over a Canadian-owned plant to enter the Canadian market, it is often referred to in the literature as acquisition foreign direct investment.

^10.

We make use of our variables in relative form to deal with the lack of special price data for each plant. A similar methodology is employed in Christensen, Cummings and Jorgenson (1981), Olley and Pakes (1992), Bartlesman and Dhrymes (1994), Baily, Campbell and Hulten (1992), Baldwin and Gorecki (1991), and McGuckin and Nguyen (1995).

^11.

The average wage rate was calculated using both production and non-production workers. Use of production workers alone produced qualitatively similar results.

^12.

Value added in the Census of Manufactures (now the Annual Survey of Manufactures and Logging) is the value of sales minus the value of purchases of intermediate goods and energy.

^13.

See Baldwin, Beckstead and Caves (2002) for a study of plant level diversification and a discussion of the Herfindahl measure of product diversification.

^14.

See Baldwin and Rafiquzzaman (1994) for a discussion of the methodology used to create these groupings.

^15.

It is noteworthy that a considerable number of plants report zero non-production workers and that a good portion of the magnitude of the coefficient is due to the difference in the probability of takeover between the group with no non-production workers and the group with some non-production workers.

^16.

The time dummies reveal that there is a pattern of reduced merger activity during the major recessions in the early 1980s and 1990s.

^17.

During this period, foreign controlled firms were withdrawing from labour-intensive sectors as competition from developing nations became particularly intense in the non-durable sector. For a more detailed discussion, see Baldwin and Gellatly (2005)