Embed Size (px)
Citation preview
COMPARATIVE COUNTERFACTUAL IMPACT EVALUATION OF REPAYABLE AND NON-REPAYABLE FINANCIAL ASSISTANCES TO SMES IN HUNGARY
ATTILA BÉRES & GERGŐ ZÁVECZ
[email protected] I www.equinoxconsulting.hu
INTRODUCTION
Financial assistance to small and medium sized enterprises (SMEs)provides a significant part of the EU Cohesion Policy, and there isa long debate on the effectiveness of these instruments.
Our evaluation focused on the effectiveness of two types ofassistance: grants and financial instruments (FI) to SMEs, co-financed by the European Regional Development Fund (ERDF). Wecompared two very similar interventions, with similar targetgroups, objectives, eligible costs and size of assistance (EUR20 000 – 40 000 in average). The only significant difference wasthat one of the schemes was a repayable (financial instrument),the other was a non-repayable (grant) one. Both types of schemeswere available in the more developed Central Hungarian Regionand also in the less developed (‘convergence’) regions, thus wewere able to compare the effectiveness of the instruments in twodifferent types of regions.
SMALL AND MEDIUM SIZED ENTERPRISES IN
HUNGARY
SMEs play a significant role in the Hungarian economy. More than600 thousand SMEs
Employ ~70% of total employees Create ~50% of the gross value added Provide ~30% of total exports
The main difficulties of SMEs in Hungary are:
The low investment rates (in many cases below amortization) Difficulties in accessing financial resources from the market Low productivity
A policy debate is on-going on whether investments of SMEsshould be subsidized through repayable or non-repayablefinancial assistance to achieve better results.
Types of Assistance
Less Developed Regions Central Hungarian Region
Grant FI Grant FI
Time period 2007-2011 2007-2012 2007-2011 2007-2012
Number of calls 6 calls continuous 6 calls continuous
Type Investment grant Soft loan Investment grant Soft loan
Repayable /Non-repayable
Non-repayable Repayable Non-repayable Repayable
RegionsLess developed (convergence)
Less developed (convergence)
More developed
(Central Hungary)
More developed (Central Hungary)
Average amount of the subsidy
~40 000 EUR ~17 000 EUR ~28 000 EUR ~18 000 EUR
Average aid intensity
40% n.a. 30% n.a.
Number of closed projects
~7300 ~1000 ~1800 ~300
Selection processSimplified (only
eligibility criteria)Financial
intermediariesSimplified (only
eligibility criteria)Financial
intermediaries
Table 1. Four forms of assistance in Hungary used in the analysis
For further information on the instruments, see: https://www.palyazat.gov.hu/download.php?objectId=1781 andhttps://www.palyazat.gov.hu/download.php?objectId=25843
WHAT DO WE KNOW ABOUT THE EFFECTS OF
FINANCIAL ASSISTANCE?
GENERAL IDEA OF THE PAPER
MAIN FINDINGS IN THE LITERATURE
Heterogeneous findings based on country specificities.Some examples:
The evaluation by Bondonio and Martini (2012) found that revolvingfunds can be as effective as grants, under certain circumstances.
Pellegrini and Centra (2006) found that turnover, employment andfixed assets grew more rapidly in subsidized firms. On thecontrary, labor productivity grew slower in these enterprises, whichcan lead to efficiency problems in the long run.
In Northern Ireland, Harris and Trainor (2005) found that capitalsubsidies increased the total factor productivity of subsidizedfirms.
Harris and Robinson (2004) found that financial assistance had noeffect on productivity in the UK.
In Sweden, Bergström (2000) found no evidence on the positiveeffects of capital subsidies on total factor productivity at firm level.
EXPECTATIONS
Evaluation question: “What effects the four financial assistance had oncertain economic indicators of the participating enterprises?”
RESEARCH QUESTION
As it has been stated in the Introduction, our evaluation focuses onthe effectiveness of two types of assistance. We compare two verysimilar interventions, targeting the same geographical regions, wheretarget groups, objectives, eligible costs and the size of assistancewere also similar. The only significant difference is that one of theschemes was repayable, the other was non-repayable. Both types ofschemes were available in the more developed Central HungaryRegion and also in the less developed convergence regions.
This makes a fourfold comparison of the effects possible (the fourcategories are based on the repayable/non-repayable and regionalcategorizations, as seen in Figure 1) that fills a lacuna in the literaturesince other studies compared financial assistance forms which wererather different in many other details as well. (These four are similarexcept for the two differences depicted above in Figure 1.)
The question is the following: “What effects the four selected types offinancial assistance have on certain economic indicators of theparticipating enterprises?”
Financial
assistance
Economic
indicators
Repayable /
Non-repayable
Region
Figure 1. Main research question
POPULATION
Hungarian small and medium sized enterprises having successfully applied for one (and only one) of the four financial assistance schemes in the 2007-2013 period served as treated group. All other small and medium sized enterprises that did not receive any financial resources from EU Funds served as a pool for possible control cases.
RESEARCH DESIGN
DATA Monitoring data of non-repayable assistance Monitoring data of repayable assistance Physical and financial data of Hungarian
enterprises (industry, location, operating time, liabilities, tangible assets, sales)
The database used for the analysis has beenconstructed by combining data on the basic characteristics of the small and medium sized enterprises in Hungary and adding some information about those who successfully applied. (The only additional information was whether they received a subsidy or not, and if yes, which type.)
First, enterprises receiving financial assistance in 2013 were eliminated since data on their characteristics were not available for 2015.
Second, the years were pooled; year ‘t’ was used as the year of assistance, thus not the calendar years but rather the relative years as compared to the time of subsidy were included in the analysis.
Step 1Data Sources
Step 2Combination of
Datasets
Step 3Two Additional Modifications
METHODS
To define the treated and the control groups, we used a nearest neighbor (one-to-one) matching method.
Matching was based on location, industry, time of operation and the lagged (t-1) economic indicators (for a list, see step 3).
Even after matching, differences between subsidized and non-subsidized companies in year t-1 remained.
Change over time in economic indicators of subsidized and non-subsidized companies were compared (see adjusted control).
T-tests were used to compare economic indicators of control and treated companies in t+2.
The economic indicators analysed were: tangible assets, sales and liabilities.
Step 1Propensity Score
Matching
Step 2Difference-in Difference
Step 3Two-sample t-tests
DESCRIPTIVES
Figure 2. Sales (in thousand HUF)
Figure 3. Liabilities(in thousand HUF)
Figure 4. Tangible assets (in thousand HUF)
Our results support the idea that repayableassistance is more effective in subsidising SMEs,and SMEs operating in the economically lessdeveloped regions need public funds to operatemore effectively. However, it needs to be added thatthese results are valid only for a small segment ofregimes supporting SMEs (small scale investmentsubsidies). We need more evidence on the effects ofSME subsidies to have a clearer view on this topic.
Economicindicators
Non-subsidized (difference int+2 and t-1)
Subsidized (difference in t+2 and t-1)
Double difference
p-value
Non-repayable
Sales (in thousandHUF)
-3 910 25 215 29 125 0,21
Tangible assets (in thousand HUF)
452 15 515 15 063 <0,001***
Liabilities (inthousand HUF)
-3 172 12 102 1 5275 <0,01***
Repayable
Sales (in thousandHUF)
1 444 18 642 17 198 <0,1*
Tangible assets (in thousand HUF)
194 4992 4 798 <0,05**
Liabilities (inthousand HUF)
1 655 7 556 5 901 <0,01***
Economic indicators
Non-subsidized (difference in t+2 and t-1)
Subsidized (difference in t+2 and t-1)
Double difference
p-values
Non-repayable
Sales (in thousandHUF)
12 118 -5 547 -17 666 0,729
Tangible assets (in thousand HUF)
5 299 13 728 8 429 0,616
Liabilities (inthousand HUF)
13 144 12 451 -693 0,971
Repayable
Sales (in thousandHUF)
-120 15 492 15 612 0,528
Tangible assets (in thousand HUF)
-1 397 4 164 5 560 <0.1*
Liabilities (inthousand HUF)
2 112 9 089 6 977 0,328
RESULTS
ACKNOWLEDGMENTS
The study was conducted in the framework of theex-ante evaluation of the Economic Developmentand Innovation Operational Programme (Hungary,2014-20 period). The evaluation was sponsored bythe Prime Minister’s Office in Hungary, in 2014. Theanalysis was prepared under the overallmanagement of Attila BÉRES and Gergő ZÁVECZ. Wewould like to thank Szilvia HAJDU and Gyula SZŐKE,former colleagues of the Prime Minister’s Office, forhelping the preparation of the analysis with highlyuseful advice and comments.
KEY LITERATUREBERGSTRÖM, Fredrik (2000): Capital Subsidies and the Performance of Firms, SmallBusiness Economics, 14, 183-93.
BONDONIO, Daniel, Alberto MARTINI (sponsored by the European Commission, 2012):Counterfactual Impact Evaluation of Cohesion Policy: Impact and Cost-Effectiveness ofInvestment Subsidies in Italy, Associazione per lo Sviluppo della Valutazione e l’Analisi dellePolitiche Pubbliche
GERTLER, Paul J. et al. (2011): Impact evaluation in practice. The World Bank.
HARRIS, Richard and Catherine ROBINSON (2004): Industrial Policy in Great Britain and itsEffect on Total Factor Productivity in Manufacturing Plants, 1990-1998”, Scottish Journal ofPolitical Economy, 51, 528-43.
HARRIS, Richard and Mary TRAINOR (2005): Capital Subsidies and Their Impact on TotalFactor Productivity: Firm-level Evidence from Northern Ireland, Journal of Regional Science,31, 49-64.
PELLEGRINI, Guido and Marco CENTRA (2006): Growth and efficiency in subsidized firms,Paper prepared for the Workshop "The Evaluation of Labour Market, Welfare and FirmsIncentives Programmes", May 11th - 13th 2006, Istituto Veneto di Scienze
* significant at 5%, ** significant at 1%, *** significant at 0.1%
More developed region (Central Hungary)
Less developed (’convergence’) regions
