Lecture 6 - Productivity and Growth (cont.)

Productivity slowdown and real wage slowdown in 1970s and 1980s

Compared to 1950-1972, all OECD nations experienced a slowdown after 1973 until 1994. See table 3, page 9.

Key questions:
Why did this happen?
Why did the US pick up after 1995 and other European nations didn't?

Labor Productivity Model

We graphed real wages (W/P) against labor supply and demand, L. L_S is assumed constant against wages and is therefore vertical. Labor demand increases with decreasing wages. There's an equilibrium point between the L_S vertical and the L_D line.

Labor demand is represented by MPL - the marginal productivity of labor. MPL=ΔY/ΔL. Average labor productivity is Y/L. MPL moves, in general, with Y/L. If Y/L is down, as it was in the 70s and 80s, MPL will shift down too.

What happens to the equilibrium when MPL shifts downward? There are two possibilities:

If the labor market keeps the labor supply at the same size, wages will go down. This is the case in the US.

However, if the labor market is more interested in wages staying the same, then the labor supply will shrink, i.e. unemployment will increase. This is the case in Europe. In Europe, the strong labor unions demand constant high wages. This comes at the expense of increased unemployment in Europe, as the data in the graph shows. High hiring costs tend to keep the unemployment rate somewhat permanent.

Look at the annual turnover of firms in manufacturing. From 1989-1994, the rate in France and the UK was 22-23%. Very dynamic. US was 18.5%. Italy and Germany were lower than the US. US is not an outlier.

In net employment gain in manufacturing after 2 yrs, US is at 134%; while European countries are around only 5-23%. This is due to the barriers to hiring in Europe. These hiring barriers may also impact the willingness of European firms to adopt new technology since they would need to hire more and more skilled workers.

Why was there a worldwide slowdown in productivity?

1. Composition of labor force has been changing. Babyboomers entered the workforce. They are less experienced and therefore less productive.
2. Increasing government regulations. For example, environmental protection and workplace safety. These both impact productivity, even though they are good causes.
3. Oil price shock. Sharp increases in oil prices may have made some of the capital stock permanently obsolete. However, since 1985 until recently, there have been large oil price decreases, yet the productivity growth revived only in manufacturing sector. (So this is not a satisfactory answer on its own.)
4. Could it be that the world has run out of new ideas about how to produce? Although computers and IT technology are significant innovations, it seems that they are only beginning to yield significant productivity gains, mainly in the manufacturing sector. Computer/IT technology only show up in the data after around 1995. This lag between invention and higher productivity is not unusual. (See the stages of technological revolution below.)
5. Mis-measurement of output growth and productivity. Perhaps we should measure the quality of products instead of the number of products. Health care and financial services have improved, but how do we capture that?
6. Lower saving? Less investment in new innovations.

Information Technology and the New Economy in the 1990s and 2000s

Labor productivity growth in the non-farm business sector increased from about 1.5% in 1973-95 to 2.5% in 1995-2000. Since 2000 (through 2004), it averaged about 3.4% per year. Perhaps as much as half of the acceleration came through increasing IT capital per worker - capital deepening - and about a quarter of it through improvements in the efficiency with which IT goods were produced.

Nominal IT investment increased from 1987-95 to 1995-1999 (9.3 to 16.6). Some connect this investment to the growth in labor productivity and conclude that IT caused an increase in labor productivity throughout the economy. Examples: Dell, Amazon. It changed retail purchasing and delivery model. See graphs on page 12 of packet, based on McKinsey (2002) "How IT Enables Productivity Growth".

But it's not so simple. See "How IT Changes US Productivity" by McKinsey (2002) (I could not locate this publication online) and the graph page 13 in packet. They found only 6 "jumping" industries that benefit greatly from the IT investment. But others, such as agriculture, may actually see negative growth despite increased IT investment.

Historical Perspective

How does Information and Communication Technology (ICT) compare to the greatest inventions of the 20th century: steam power, railway, electricity, etc.?

The recent boom and collapse in ICT stock prices and in spending on goods embodying new technology is typical of technological revolutions.

Example: the railway system in London in 1840s. The expansion was fueled by stock investment boom, followed by stock market crash, but the late 19th century continued to benefit from this innovation.

3 typical stages of technological revolution:
Stage 1: Productivity growth in innovating sector (e.g. computer manufacturing industry)
Stage 2: A fall in price of innovation that encourages its wide use by business or consumers (also accompanied by wage increases since workers are more productive)
Stage 3: Production in all sectors reorganize around the innovation that embody new technology
leading to broader-based surge in productivity. (this may be where we are today with IT)

It seems that the US is now just entering the 3rd stage.

Comparing Countries

See table 4 on page 16 of packet, comparing GDP per capita (PPP) of selected countries in 1950 and 2000. Look at Ireland and Japan. Rapid growth rate enabled them to close the gap. The beauty of compound interest!

Notes on these data: Comparing GDP based on exchange rates can be misleading because the CPI in one country may be different than another. Therefore the PPP takes this into account to calculate a "purchasing power parity" statistic.

Absolute Convergence Hypothesis

Poor nations have lower (K/L), but higher MPL.
Two reasons for absolute convergence:
1. Law of diminishing marginal product of capital stock
2. don't need to reinvent wheeel: advantages to second comers. take advantage of foreign advanced tech (copy, adopt and assimilate...): improve A

conversely, rich countries would eventually

Plotting GDP per capita growth vs. GDP per capita, we would expect a negative correlation, but the scatter plot doesn't support this theory. There's no correlation. We need to reexamine the hypothesis.

BTW, among OECD countries, Korea, Ireland and Portugal have highest real per capita GDP growth.

Conditional Convergence Hypothesis

Whether poorer countries can grow faster and hence catch up with richer countries or not: It turned out to be conditional on having good policies and institutions in place such as:
- investment in education, see panel 3
- investment in physical capital stock, see panel 4
- trade openness, see panel 5 (there are other geographical factors)
- stable macroeconomic management (low inflation, low budget deficits, stable exchange rates)
- quality of public institution, see panel 6, related to expropriation risk

We actually find poor nations lending to richer nations (like China to US). Why not just invest it internally? Because of all the risks involved. Even though returns in US may be less, the risk is less too. The risk-adjusted rate of return in poorer nations is actually pretty low.

Global imbalances and capital flows graph shows that the US and G7 nations (except Japan) are borrowing heavily and the loans are coming from emerging and developing nations. The e&d nations don't have enough capital flows that would help them grow because of all the risks they have.