Editor's note: Michael S. Bernstam, a research fellow at Stanford's Hoover Institution and a former economic advisor to the Russian government, has kindly provided us with his short analysis of the meaning of Polish recovery.

1. Recovery

The data available from international sources (the IMF, the World Bank, UN's statistical publications, London's Economist Intelligence Unit) and the data from Główny Urząd Statystyczny (Rocznik Statystyczny 1995, pp. XXXV, 528) indicate the following:

The gross domestic product (GDP) of Poland, the broad measure of goods and services produced in the country, a concept almost identical to national income, grew by 1.7% per year on the average in 1980-90 and fell by 1.5% per year on the average in 1990-94. These yearly averages do not tell us much. There had been a depression in 1990-91 and a recovery in 1992-95. The fall in 1990-91 yields cumulatively -17.8%. The recovery in 1992-95 yields cumulatively +20.8%. The recovery seems to have been virtually complete by 1996. Consider the calculation:

(1) 100% - 17.8% = 82.2%


(2) 82.2% x (100% + 20.8%) = 99.3%.

How reliable are these numbers? The reliability is in the eye of the beholder. In Jeffrey Sachs' assessment, the fall of 1990-91 had been exaggerated and in reality was negligible. The IMF and the World Bank offered similar assessments. They can be summarized as follows:

a) not all new private businesses in services reported their sales receipts in order to avoid taxes

b) even if the fall of the GDP had been roughly as reported, there were unmeasured gains in consumer satisfaction due to the end of shortages and queues.

This leads to the claim that not only was the depression exaggerated, but the recovery was also underestimated for the same reason: there occurred a growth of the unreported shadow economy. New businesses have underreported sales and old businesses underreported additional output for tax reasons.

In economic literature, similar reasoning has been in evidence concerning Russia and other Eastern European and Asian countries including Mongolia and the Central Asian republics. Remarkably, however, this reasoning has not been applied to Chinese growth. Instead, it has been claimed that Chinese growth is not underestimated but rather overestimated. Chinese growth did, of course. occur outside the parameters of advice given by the IMF and related institutions. These arguments can be challenged on technical grounds.

Consider a). Undoubtedly, many businesses in post-Soviet countries evade taxes and underreport sales receipts. But there is also an opposite force. State enterprises continue to be subsidized. The subsidy represents the difference between market prices and costs per unit of output. This stimulates overreporting, rather than underreporting. Under central planning, overreporting was profitable for reasons of rewards for quantities of goods; in postcommunist conditions, it is profitable now for reasons of subsidies. Suppose I produce coal at the cost of $50 per ton. The market price is $30 per ton. I get $20 per ton in subsidy. If I reported 1,000 tons produced and sold, I get $20,000 in subsidies. I am interested in reporting 1,500 tons produced and sold so I can get $30,000 in subsidies. By cooking up the books, I get an additional $10,000. There is less control and more corruption than before, and overreporting and ripping-off the public purse is easier than before. Consider that about 50% of the GDP in Poland is produced in the subsidized state sector. It is easier for that sector to increase overreporting than for the other 50% of the GDP, or the private sector, to increase underreporting. The penalties for tax evasion exist. The penalties for subsidy appropriation do not exist.

Consider b). Suppose every person stood in line two hours per day to buy goods, and now he has two extra hours per day to ride a bicycle, pat the kids, and watch the birds. This sounds like an increase in intangible consumption, according to standard opinion. This standard opinion is technically wrong (and its authors probably know it). The reduction in hours spent on unpleasant activities is not an increase in consumption but an increase in leisure and a reduction in labor hours. Nothing less and nothing more. Staying in queues was a part of the labor life in socialist countries. Privileged workers could purchase goods at enterprise shops. Now, instead of a 10-hour working day they have an 8-hour working day. They work 40 hours per week instead of 50. They had to work an extra 10 hours per week as a price for buying goods at prices lower than the wages they were paid. Thus they got extra wages for the extra 10 hours per week. Now both the extra wages and extra hours were cut. It is more efficient but one cannot say that consumption increased. They did not get additional consumption because bicycle trails, birds, and kids were there before to enjoy; they got extra leisure to enjoy them. Leisure does not count as consumption.

All these considerations call for a recalculation of both depression and recovery in Poland and elsewhere. Such recalculations require much more technical means and data than are at my disposal. I can only use a rule of thumb. I accept the standard data on the grounds that underreporting of services in the private sector and overreporting of manufacturing in the state sector yield a wash if the sectors constitute 50% each of the GDP. Thus I tentatively accept the numbers for Poland. I do not accept them for Russia because the new business private sector (not the subsidized privatized enterprises) is too small to tip the scale even if this sector is twice as large as officially counted.

The growth data had been a highly contested ideological matter with respect to the old Soviet Union. Some people disputed high growth numbers, others defended them. The situation has not become more quiet nor more honest and objective after the fall of the Soviet empire. Numbers remain ideological weapons.

2. Recovery and growth

Recovery is not equal to growth. If you made $100,000 in 1989, fell to $80,000 in 1990-91, grew to $85,000 in 1992, to $90,000 in 1993, to $95,000 in 1994, and to $100,000 in 1995, you have recovered your income. You did not recover your losses! You are still short of $70,000. Thus you have lost cumulatively 70%, not 20% of your income. A mere recovery is no achievement. There was no reason to change economic policy if the result was a recovery, not an improvement. Only if your income grows to, say, $114,000, and stays that way for five years, you recover the $70,000 you had lost.

However, it is much easier to recover the level of 1989 than grow beyond that level. This is because the productive capacity is there. One does not need additional investment, additional education of the labor force, additional management, etc. Recovering within the old capacity is usually fast and easy.

By international standards for after-recessions, Poland's recovery has been anemic, while the recovery of other Central and Eastern European countries has been embarrassing. Germany grew at 10% per year for twenty years after the 1948 reform. Japan grew likewise. China has grown at 10% per year for 20 years since 1976. Anything less than that indicates severe institutional problems, unresolved vestiges of the past, and newly emerged distortions and aberrations.

The critical test will be whether the Polish growth rate accelerates or decelerates in 1996-98 and beyond. If it decelerates óand there are signs in 1996 that it will ó or if it continues at an anemic rate, we would know that the economic strategy has been deficient not only on the depression grounds, but also on the post-depression growth grounds. In addition, unemployment at 15% of the labor force in Poland indicates that the economy is not healthy despite the recovery.

3. Indirect evidence on depression

Since post-reform contractions in Poland, Russia, and other transition economies are disputed on grounds of statistical uncertainty, independent evidence is in order. To test whether a depression had occurred, we need evidence that is simultaneously

a) independent of the measurement of gross national product (to be free of the doubts that mar the latter measurement at a time of transition); and

b) related to the nature of economic well-being.

One measure clearly satisfies these two juxtaposing conditions: the length of life of the population in a given year under the existing conditions. Although technically called `life expectancy at birth,' this measure indicates not how many years a person born this year would live in the future (this would depend on future conditions), but rather how many years this person would live if the probability of death at each age, from 0 to 100, were the same as under the current conditions during this given year.

Thus "life expectancy" gives us the snapshot of health, social conditions, and general well-being over all age groups in a given year. At the same time, it is calculated from demographic data and thus is free from all underreporting and overreporting related to changing incentives in the production sector.

The conditions for the test are simple: If we think that economic depression occurred in 1990-91 and recovery started in 1992, life expectancy should fall in 1990 relative to the previous years, further fall in 1991 (because the income falls further), then rise in 1992, and continue to rise in 1993 and 1994. Such coinciding of economic and demographic ups and downs cannot be a coincidence (no pun intended). Also, similar relationships should take place elsewhere, both in other transition economies and in the U.S. during the Great Depression. Let us observe what had happened in Poland according to Rocznik Statystyczny 1995, p. 67.

Life expectancy at birth

The declines and increases of life expectancy for males fully coincide with the dynamics of contraction and recovery. Among females, we observe a lagged response within the same pattern of a decline and a recovery. There was a decline of female life expectancy in 1991 relative to 1990 and a slower recovery in 1992-93. The first year of contraction, 1990, was actually not bad for females; it was bad for males. A similar difference between male and female responses to depression can be observed in Russia ó in fact, more saliently. One can suggest that women gained more from the end of shortages than men: fewer hours in the queues, more leisure, more relaxation, less stress. This one-time gain in 1990 in Poland and in 1992 in Russia could have temporarily outweighed the negative influences of the declining income, worsening health care, deteriorating capital maintenance at the work place (leading to more industrial accidents), etc.

To check this pattern of life expectancy changes with respect to depressions and to see that the Polish relationship is not a fluke, one can compare international experience.

(U.S. Bureau of the Census. Historical Statistics of the United States, Part 1. Washington, D.C.: USGPO, 1975, p. 55.)

In the United States, we see a strong fall in life expectancy in response to the Great Depression and a recovery later on. The lag of demographic response to economic contraction is in this case one year and to recovery, three (the trough of the Great Depression was in 1993). These lags were probably due to the fact that in the 1930s the principal factor was not the health care system but nutrition: it was before the antibiotics.

(Russian State Committee on Statistics, annual publications).
The fall of life expectancy in Russia in response to a very large contraction (over 40%) is about six years for males (more than in the U.S. in the 1930s), and half of that for females. Like in Poland, females must have gained from the end of shortages. They acquired more opportunities to sleep seven or eight hours per day, a seemingly trivial but important health factor.

One can observe that Russian life expectancy did not recover owing to the fact that the Russian economy continued its contraction.

To sum up, one has a consistent set of strong indirect evidence to indicate that a depression did indeed take place in Poland around 1990-91, with a subsequent recovery.

Characteristically, the World Bank in its 1996 annual report included a section entitled “Is transition a killer?î (World Bank. World Development Report 1996. From Plan to Market. New York: Oxford University Press, 1996, p. 128). It points out that Russian adult mortality is now 10% higher than that in India. It also provides regional evidence from Hungary that links economic hardship with declining health.