Basic growth theory and the basic growth accounting methodology leave a large part of aggregate growth and productivity growth in the Solow Residual and unexplained. Growth in employment and capital stock, and the relationship between them, play important roles in aggregate growth and labour productivity growth, but the effects of changes in the characteristics of labour and the composition of capital stock reside in the total factor productivity estimates of the Solow residual. This residual captures changes in technology along with other undefined factors.
Recent work on productivity growth still uses the growth accounting methodology, but focused on particular sectors of the economy and with extensions designed to unpack some things previously left in the residual. These include, in particular, investment in different components of the capital stock and changes in the composition and quality of labour.
Increases in capital per worker or per hour of work—capital deepening—continues to be important, but increases in capital stock are disaggregated into investment into several categories such as:
- Information and communications technology;
- Machinery and equipment; and
- Physical structures.
Changes in educational qualifications, gender structure, and age structure are used as indicators of changes in the composition and quality of the labour force. To the extent that these measures of change in characteristics and structure affect productivity less remains in the residual. More importantly, these are areas in which policies to support education, training, and labour force participation could affect productivity growth. Table 13.6 gives an example of some recent results based on this approach.
The estimates in the table show the effect of increases in the information, communications, and technology components of the capital stock. Particularly over the 1997 to 2000 period, the increase in ICT capital per unit of labour input was the largest part of capital deepening. This is also the period of strongest growth in labour productivity and the largest increase in total factor productivity. The latter captures the effects of other improvements in technology.
Table 13.6 Sources of Canadian business sector labour productivity growth, 1974–2005 (%/yr)
| |
1974–1996 |
1997–2000 |
2000–2005 |
| Labour productivity |
1.4 |
3.0 |
1.0 |
| Capital deepening |
1.1 |
1.0 |
0.7 |
|
Info & communications technology (ICT) |
0.4 |
0.7 |
0.3 |
|
Non-ICT |
0.7 |
0.4 |
0.4 |
| Labour quality |
0.4 |
0.4 |
0.4 |
| Total factor productivity |
0.0 |
1.6 |
–0.1 |
Source: R. Dion, "Interpreting Canada's Productivity Performance in the Past Decade: Lessons from Recent Research." Bank of Canada Review, Summer 2007.
The estimated contributions of changes in labour quality to labour productivity growth provide an interesting refinement of the growth accounting methodology. Ignoring this change, as the simple accounting process did in earlier examples, leaves the contribution of changes in the structure and characteristics of the labour force in the residual measure of total factor productivity. The results in the table extract and quantify this important source of productivity growth.
Changes in the structure of investment and in the quality of labour are both areas in which public policy plays a role. Government tax policy can be designed to encourage producers to direct investment to areas like communications and information technology. Some of this investment has effects confined to specific industries. Other parts create and strengthen national capacity and efficiency that is widely available and used by business and households, for example high-speed internet service and expanded wireless service and capacity. Educational policy and other human resource policies contribute to the quality, adaptability, and mobility of the labour force.
Endogenous growth theory
Another interesting strand of work makes more fundamental changes in growth theory. The simple neoclassical growth theory made economic growth depend on exogenous variables, the rate of population growth, the saving rate, and the rate of capital accumulation, whose values are determined outside the growth model. The subsequent work on catch-up and convergence makes technical progress respond to economic and political factors. But it would be nice to have a stronger link between economic behaviour and the rate of economic growth. We want to make growth endogenous, or determined within our model. Endogenous growth implies that the steady-state growth rate is affected by economic behaviour and economic policy.
Exogenous variable: a variable with a value determined outside the model.
Endogenous growth: growth determined economic behaviour and policy within the model.
Professor Paul Romer of the University of Chicago pioneered endogenous growth theory. Growth theories are built on the saving that drives investment and capital accumulation. While the production function used the basic growth model, increasing the stock of capital and the capital labour ratio leads to a diminishing marginal product of capital. Output and output per worker increase at a decreasing rate but in the steady state output grows at the rate of growth of population and labour force. Output per worker cannot grow indefinitely.
Endogenous growth theory is based on a different view of the roles capital plays in the growth process. It retains the assumption that, for the individual firm, the marginal product of capital decreases as capital stock increases. But it adds the insight that in the aggregate economy there may be significant positive externalities to the increase in capital stock. In other words increases in capital stock by one firm may result in improved productivity and output in other firms. Improved internet speed and access, or faster wireless voice and data transmission comes from investment by internet service providers and wireless companies. Using the new capacity provided improves the efficiency and productivity of production in other companies. Aggregate productivity and output grows as a result of increased investment by individual producers.
This line of argument is illustrated by the shapes of the production functions. For the individual firm the production function still looks like that in Figure 13.1, concave to the horizontal axis. Increases in the capital/labour ratio increase output per worker at a decreasing rate. For the aggregate economy, according to endogenous growth theory, increases in capital per worker do not face diminishing returns because of the externalities that come with increases in the capital stock. As a result, the aggregate production function has constant returns to capital. It is a straight line rising from the origin with a constant slope. Growth in the capital/labour ratio produces a constant rate of growth in output per worker and aggregate output. A 10 percent increase in capital per worker gives a 10 percent increase in output per worker.
Constant returns to raising the aggregate capital to labour ratio in the economy allows an escape from the key growth limitation in the neoclassical theory. It makes growth endogenous and dependent upon parameters that could be influenced by private behaviour or public policy. Any policy that succeeded in raising the rate of investment would permanently raise the growth rate. Similarly, any policy achieving a one-time improvement in technology (for example, greater workplace efficiency) would permanently raise the growth rate of capital per worker. This would mean permanently faster output growth.
Not only can government policy affect growth in this framework, government intervention may also increase efficiency. In the simple endogenous model described here, there are externalities to capital accumulation. Individual producers may not realize that by investing to increase their capital stock they may also improve productivity in other firms. Public policy that recognizes this economy wide effect can subsidize investment to increase investment and aggregate economic growth. By the same argument, externalities to investment in human capital support government subsidies to education and training.
However, endogenous growth theory faces criticism based on the assumption that there are exactly constant returns in the aggregate from accumulating one factor of production. The diminishing returns in the basic model make long-run growth exogenous. An economy with increasing returns experiences continuously increasing capital stock and output. This sort of explosive growth does not correspond to any empirical observation.
New growth theory emphasizes the roles of research and development, innovation, education and "learning by doing" as sources of improved technology and productivity. Research, development, and innovation come from decisions to invest in new knowledge and to apply it to production processes. Education comes from decisions to invest in human capital. Learning by doing is a natural outcome of employment experience. All these are ongoing processes, although they may be pursued unevenly over time as economic conditions and economic policies change.
As a result, increases in the stock of capital and the level of employment always embody new technology and knowledge. There is no separation between increases in the capital to labour ratio and the state of technology, as in the basic neoclassical model. But if new capital stock and new employees bring new technology to the production process, the per worker production function shifts up as capital per worker increases. As capital per worker increases, the economy moves up and along a new per worker production function because the embodied technological improvements offset otherwise diminishing returns.
Recent studies of the sources of productivity growth based on growth accounting are consistent with this approach to reconciling neoclassical and new growth theory. The findings reported in Table 13.6 above show the contributions to productivity growth made by different types of capital equipment and changes in labour force structure. These changes in "technology" are made integral parts of the growth in capital stock and employment rather than left as exogenous residuals.
The costs of growth
Can the benefits of economic growth be outweighed by its costs? Pollution, congestion, and a hectic lifestyle are a high price to pay for more cars and trucks, washing machines, video games, smart phones and tablets.
Since GDP is an imperfect measure of the true economic value of goods and services produced by the economy, there is no presumption we should want to maximize the growth of measured GDP. Without government intervention, a free market economy produces too much pollution. But the elimination of all pollution is also wasteful. Society should undertake activities accompanied by pollution up to the point at which the marginal net benefit of the goods produced equals the marginal pollution cost imposed on society. Government intervention, through pollution taxes or regulation of environmental standards, can move the economy towards a more efficient allocation of resources and a higher standard of living, broadly defined.
The full implementation of such a policy would (optimally) reduce growth of measured GDP to below the rate when there is no restriction on pollution and congestion. This is the most sensible way in which to approach the problem. It tackles the issues directly.
In contrast, the "zero growth" solution is a blunt instrument. It does not differentiate measured outputs that have social costs, from outputs without new social costs. As a result there are no new incentives to minimize the externalities already caused by pollution, congestion and environmental degradation. These call for taxes or incentives to reduce current social costs. 'No growth' does not go far enough.
A more extensive and inclusive measure of GDP would be a step toward recognizing and dealing with the costs of growth. Such a measure might include both positive aspects of production and consumption that contribute to welfare, such as environmental quality and low levels of congestion. It might also adjust to give a more complete measure of both private and social costs of different patterns of production, consumption and leisure. We have seen one way to approach this in the United Nations Human Development Index in Chapter 4, but a broader based index is needed.
