3 Assessing overall impact induced by policy
By applying the multiregional I-O model developed, there were estimated employment and
labour income impacts produced in Romania by application of development policies related to EU
accession financial package for the period 2007-09.
Results from impact analysis reveal that labour income and employment variations in Romania
will be 2,425 million euro and about 1.4 million of labour units, respectively. Variation of income per
capita is estimated to be 108 €. In terms of income, services and industry are the sectors attracting
most part of impact: services absorb 50% of impact whereas industry attracts 45% of income variation.
Agricultural employees only receive 5% of income variation. As regards employment, most part of
impact is concentrated on agriculture (50% of employment variation) whereas the remained part is
distributed between industry (29%) and services (21%) (Tab. 7).
In comparison with 2000 data, income and employment are forecasted to increase by about 16%
and 17%, respectively. The bigger variation is registered by agriculture, followed by industry and,
finally, services. In terms of effectiveness, policy generates an increase in income by 32% of public
expenditure and in employment by 183 labour units for each one million euro. At a sector level, policy
demonstrates to be more effective in services, as for income, and in agriculture, as for employment.
To improve the analysis of effectiveness, it is interesting to verify if policy will contribute or less
towards a reduction of territorial and sector disparities. Through the analysis of income distribution,
there emerges that territorial variability12 among regions tends to diminish, passing from 24.4% to
22.4% (Tab. 8). Even the variability among sectors decreases going from 86.5% to 83.3%.
Considering all sectors and regions jointly, total variability decreases from 92.8% to 89.3%. As far as
employment distribution is concerned, variability tends to increase. Variability among regions passes
from 20.4% to 24.3%. That among sectors goes from 157.9% to 159.6%. Finally, total variability
shifts from 168.8% to 173.1%.
Results in terms of variability show that, at an aggregate level, policy helps to reduce both sector
and territorial disparities in terms of income, favouring a more uniform development, but sharpens the
differences among sectors and regions from an employment point of view.
Application of a multiregional I-O model permits to increase the level of detail, analysing impact
at a sub-national level.
The regions attracting bigger impacts are SR and NER whereas the regions registering lower
impacts are WR, at an income level, and BR, as for employment (Tab. 7). In terms of income, in all
the regions, services and industry attract a bigger share of regional impact. As far as employment is
concerned, agriculture absorbs most impacts in all the regions except for CR and BR where effects are
concentrated on extra-agricultural sectors.
Compared to 2000 data, following to policy application, SR and NER grow more in terms of both
income and employment. BR is the region growing less.
In terms of effectiveness, policy is by far more effective in generating income in BR (78% of
public expenditure transforms into income). In the other regions, the level of effectiveness is roughly
similar going from 23% (NER) to 35% (CR). Services are the sector where policy effectiveness is
bigger in all the regions except for WR, where the sector in which policy is more effective is industry.
As far as employment is concerned, SR demonstrates to be the region able to valorise policy funds
better: for each one million euro, policy generates about 272 labour units. The less competitive region
from the policy-use standpoint is SER with 137 labour units for each one million euro. At a sector
level, higher effectiveness can be noticed in agriculture in all the regions reaching in BR the level of
about 1,758 labour units for each one million euro.
An analysis of sector differences can be also extended at a regional level. In terms of income
distribution, there can be noted that sector variability decreases in all the regions. Bigger decreases in
sector disparities involve SR and NER. Also with regard to employment, sector differences tend to
decrease with the exceptions of SR, where variability increases by 8%, and NER, where variability
remains unaltered.
12 Variability is measured by variation coefficient, calculated before and after application of policy.
11