After robust growth in ICT investment in Canada and the United States during the 1980s and 1990s, growth in ICT investment started to slow after 2000 and its share in GDP entered a downward trajectory. To explain this trend, we focus on two facets of the changing ICT spending pattern driven by the rise in cloud computing starting in the mid-2000s. Instead of investing in ICT capital goods, organizations now purchase cloud services from cloud service providers that appear to be more efficient in producing computing services. Second, cloud service providers undertake substantial own-account investment in ICT equipment which is not counted in official statistics. We find that spending on cloud services in both Canada and the United States rose substantially from the mid-2000s. Also, adding own-account ICT equipment investment by cloud service providers results in additional 1.7 percentage-points in annual growth for nominal ICT investment in Canada and additional 1.1 percentage-points in the United States.

It is well known that there has been a secular decline in the manufacturing share of total employment in Canada, with the decline accelerating after 2000. Among the factors that contributed to that trend, this report focuses on rising Chinese import competition in Canada, which also accelerated after 2000. We find that the trade-induced job loss in manufacturing amounts to 113.5 thousand during the 2001-2011 period but the loss was not equally distributed across skill groups. We estimate that the loss was largely driven by: low-skilled occupations (89.8 thousand) when analyzed by skill level; and occupations in services (57.6 thousand), technical/paraprofessional (12.1 thousand), and production (51.8 thousand) when analyzed by skill type. A trade-induced job loss in manufacturing is also estimated by gender. We find that females had a larger loss than males did in relative terms in all skill groups. Lastly, the labour reallocation in response to a trade shock is found to be important in offsetting the negative employment effect, but the degree of reallocation varies across skill groups and genders.