At the end of 1992, the single market programme (SMP) came into force in Europe. It was aimed at eliminating the remaining barriers to trade among member countries. The expected consequences were increases in competition, industrial restructuring and reallocation of economic activities. In turn, these consequences were intended to induce three categories of gain. Allocative efficiency gain: when producers have market power, prices deviate substantially and persistently from marginal costs. Thus the structure of consumption is distorted and total output is kept below its socially optimal level. Productive efficiency gain: while firms produce at lowest cost under conditions of competition, they begin to operate inefficiently (through overstaffing, higher wages, lack of response to new opportunities, poor management) in situations of poor competition. Dynamic efficiency gain: fostering product and process innovations and, hence, speeding up the move to the modern technology frontier, which is a major source of growth. Recent empirical studies have tried to assess the allocative and productive efficiency gains from trade liberalization focusing on industries. Although similar gains are expected in services, no study provides a quantitative assessment.
A recent study by Salgado (2002) investigates the impact of trade, domestic product market and domestic labor market reforms on productivity performance. The analysis is based on panel data for 20 OECD countries over the period 1965-1998. The results suggest that trade and domestic product market reforms explain the trend in productivity growth. Their impact on total factor productivity growth is weak in the short run but substantial in the long run (i.e. between 0.2 and 0.3 percentage points a year). This is, however, an average estimate over OECD countries. Given the difference in the coverage and speed of reforms across these countries, the estimate may not be a reliable measure of the impact of the SMP. Notaro (2002) focused on the impact of the SMP on industrial productivity in a panel of 6 European countries and 30 industries. Using Buigues et al (1990) methodology, these industries are classified as non-sensitive, moderately sensitive and highly sensitive to the SMP. The econometric evidence provided strong support to the positive impact of the SMP on industrial productivity in the last category of industries. In 1992 and 1993 the productivity in the high and medium sensitive sector increased by around 2%.
Allen et al. (1998) studied the impact of the SMP distinguishing among its effect on patterns of production and trade and its effects on price-cost margins and industrial restructuring. They also classified the industries according to their sensitivity to the SMP. The results showed that the SMP was mainly trade creating: the domestic production share of demand has fallen by 5.4 percentage point on average while the shares of both intra-European trade and extra-European trade have increased by 2.95 and 2.45 percentage point respectively. With respect to price, the result suggested that price competition has increased: on the average, price-cost margins have fallen by 3.6 percentage points in the high and medium sensitive industries. Bottasso and Sembenelli (2001) examined the impact of the SMP on market power of a large sample on Italian firms using a similar industry split. They found that in the most sensitive industries, firms' market power decreased by around 10 percentage points during the implementation of the SMP.
The results of these studies point to a consistent positive (negative) impact of the SMP on productivity (mark-up) in the high and medium sensitive industries. However, no clear pattern emerged in the other industries. The high and medium sensitive industries represent around 25% of the EUGDP. To get an estimate of the impact of the SMP on mark-up and productivity at the EU economy as a whole, we take the estimates by Notaro (2002) and Allen et al. (1998) and weight them by the share of the high and medium sensitive industries in the EUGDP. This implies an average decrease of mark-up by around 0.9 percentage points and an increase of productivity by around 0.5%. The evidence on services is not clear yet and there exists no study which provides a quantitative assessment of the Internal Market effect on services. There are, however some results from network industries (telecom, electricity and transportation) which have been used for our macro assessment. Nevertheless, our results should be considered as a "lower bound" to the overall impact of the SMP on the EU economy.
Ranges of these estimates are also computed using the standard deviation provided in the respective papers. These ranges give a lower bound of 0.45 and an upper bound of 1.35 percentage points for mark-up decrease. The bounds for productivity increase are respectively 0.25% and 0.75%.
The Commission's QUEST II model is used to assess the macroeconomic effect of these two shocks. The following question is asked in this exercise: What would have been the level of GDP and employment in 2002, 2012 and 2022 if the single market programme would not have been implemented? Various simulations are performed depending on whether average, upper or lower bounds of the SMP impact on productivity and mark-up are considered. Moreover, three scenarios are envisaged concerning the time it takes for the full realization of the TFP increase and the mark-up reduction: 3, 5 and 7 years. The results emanating from the manufacturing sector and the network industries separately have been calculated separately. A table at the end of the note gives the total results.
The simulation results suggest that real GDP would have been 1.4% lower (with a lower and upper of .76 and 2.05) in 2002 without the internal market programme. Small additional gains are to be expected in this and the next decade, with an additional GDP effect of .4% until 2012 and .5 in 2022. Similarly the level of employment would have been .86% lower (with a lower bound of .43 and an upper bound of 1.3%) compared to its actual current level in 2002 without the single market programme. However, according to the simulation results no further employment gains should be expected. Both the increase in efficiency of production and increased competition contribute about equally to the GDP gain, while about 80% of the increase in employment is due to the removal of restrictions impeding competition (1).
Not much empirical work has so far been undertaken in assessing the effects of liberalization in network industries. Some preliminary evidence for the electricity sector (Roeger and Warzynski, 2002) suggest a decline in the price cost margin from 25% to 19%. Given that the price decline in the telecom sector (relative to the consumer price deflator) has been more significant, namely about 23% relative to 9% in the electricity sector since 1996, profit margins have most likely been reduced more strongly in that sector. This would also be consistent with the more advanced state of liberalization in telecommunication. Taking into account the relative GDP weights and making the cautious assumption that mark-ups in the telecom sector have only decreased by 50% more compared to electricity would yield an aggregate mark-up decline of about .5%
In the forthcoming European Economy Review 2002 (2), the GDP and employment effects from the more recent liberalization of network industries, in particular electricity and telecom markets, are estimated to be .4% and .6% respectively after 4 years already and GDP will increase by .6% after 10 years. These effects are somewhat stronger than the internal market effects because in these simulations it is assumed that deregulation also has an effect on rent sharing between workers and firms. Thus the decline in price cost mark-ups is associated with a decline in the mark-up of wages over the reservation wage.
The following table provides absolute GDP and employment figures from the SMP and the liberalization of network industries.
(1000 of persons) in
(Bio of Euros) in
(1) These results are similar, though slightly smaller compared to the results presented in European Economy, 1996, No 4, Economic Evaluation of the Internal Market, Chapter 6: Income and Employment Effects. In these simulations an aggregate mark-up reduction of .6 was assumed and the study was more optimistic concerning TFP growth where a permanent increase in TFP growth of .1% was assumed.
(2) European Economy Review 2002, Chapter 3, Structural Reforms in Labour and Product Markets and Macroeconomic Performance in the EU.