![Sample our Geography journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5937/TOC-Subject-Sample-2021-Geography.jpg"})
Abstract
Michalek J., Ciaian P. and Kancs d’A. Investment crowding out: firm-level evidence from Northern Germany, Regional Studies. The main objective of this paper is to estimate the extent to which firm investment is substituted (crowded out) by investment support policies granted under the European Union Rural Development Programme (RDP). The empirical analyses employed the difference-in-difference propensity score-matching approach, which allows several important sources of bias to be addressed, such as selection bias, simultaneity bias and functional form misspecification, from which many previous studies suffer. Using panel data of 1333 firms from the Schleswig-Holstein region in Germany, it is found that the crowding-out effect of the RDP is close to 100%, implying that firms use public support to substitute for private investments. Furthermore, no evidence was found that, due to the RDP programme support, firms would have brought forward their investments planned originally for a later period, rejecting the inter-temporal substitution of investments.
Michalek J., Ciaian P. and Kancs d’A. 投资排挤：德国北部的企业层级证据，区域研究。本文的主要目的在于评估企业投资，被欧盟乡村发展计画（RDP）授与的投资支持政策所取代（排挤效应）的程度。本经验分析运用「差异中之差异」的倾向分数配对方法，让部分重要的偏误来源—例如选择偏差、同期性偏误与函数形式误设—得以被处理，而这些是过往诸多研究所遭受的问题。本研究运用德国石勒苏益格—荷尔斯泰因（Schleswig-Holstein）区域中1,333 个企业的面板数据，发现 RDP 的排挤效应接近百分之百，意味着企业利用公共支持来取代私人投资。此外，本研究未发现任何证据，支持厂商会因为 RDP 计画的支持，而在日后实现他们原本规划的投资，因此否决了投资的跨期替代。
Michalek J., Ciaian P. et Kancs d’A. L’éviction de l'investissement: des preuves au niveau de l'entreprise provenant du nord de l'Allemagne, Regional Studies. Cet article cherche principalement à estimer dans quelle mesure l'investissement des entreprises est remplacé (évincé) par des politiques de soutien financier en faveur de l'investissement mises en oeuvre sous l’égide du Programme de développement rural (PDR) de l'Union européenne. Les analyses empiriques ont employé la méthode de différence en différence avec appariement par score de propension, qui permet d'aborder plusieurs sources de biais importantes, telles le biais de sélection, le biais de simultanéité et l'erreur de spécification de la forme fonctionnelle, dont souffrent de nombreuses études antérieures. À partir des données de panel auprès de 1.333 entreprises situées dans le Schleswig-Holstein en Allemagne, il s'avère que l'effet d’éviction du PDR s'approche de 100%, ce qui laisse supposer que les entreprises se servent du soutien financier public pour remplacer l'investissement privé. En outre, aucune preuve n'a été trouvée, à cause du soutien financier en faveur du PDR, que les entreprises auraient donné la priorité aux investissements initialement prévus pour une période ultérieure, rejetant ainsi la substitution intertemporelle de l'investissement.
Michalek J., Ciaian P. und Kancs d’A. Verdrängungseffekte bei Investitionen: Belege auf Firmenebene von Norddeutschland, Regional Studies. Wichtigstes Ziel dieses Beitrags ist eine Schätzung des Umfangs, in dem Firmeninvestitionen im Rahmen des Programms für ländliche Entwicklung der Europäischen Union durch Subventionen zur Investitionsförderung ersetzt (bzw. verdrängt) werden. Bei den empirischen Analysen kam der Ansatz des Difference-in-Differences-Propensity-Score-Matching zum Einsatz, mit dem sich mehrere wichtige Verzerrungsursachen beheben lassen, unter denen zahlreiche frühere Studien leiden, wie zum Beispiel eine Auswahlverzerrung, eine Gleichzeitigkeitsverzerrung und eine Fehlspezifikation der Funktionsform. Anhand der Paneldaten von 1333 Firmen im deutschen Bundesland Schleswig-Holstein wird festgestellt, dass der Verdrängungseffekt des Programms für ländliche Entwicklung bei nahezu 100% liegt, was darauf schließen lässt, dass Firmen die öffentlichen Subventionen als Ersatz für private Investitionen nutzen. Darüber hinaus fanden sich keine Belege dafür, dass die Firmen Investitionen, die ursprünglich für einen späteren Zeitpunkt geplant waren, aufgrund der Subventionen des Programms für ländliche Entwicklung vorziehen, was gegen eine intertemporale Substitution der Investitionen spricht.
Michalek J., Ciaian P. y Kancs d’A. Efecto de exclusión de inversiones: evidencia de empresas en el norte de Alemania, Regional Studies. El principal objetivo de este artículo es calcular en qué medida las inversiones de empresas se sustituyen (excluyen) por las políticas que apoyan la inversión concedidas en el marco del Programa de Desarrollo Rural (PDR) de la Unión Europea. En los análisis empíricos empleamos el enfoque de comparación de resultados tendenciales junto con el método de diferencias en diferencias, lo que permite abordar varias fuentes importantes de parcialidad evidentes en muchos estudios previos, tales como el sesgo de selección, el sesgo de simultaneidad y la mala especificación de formas funcionales. A partir de datos de panel de 1333 empresas de la región alemana de Schleswig-Holstein, observamos que el efecto de exclusión del PDR es casi del 100%, lo que implica que las empresas utilizan apoyo público para sustituir las inversiones privadas. Además, no se observan datos que indiquen que las inversiones de empresas, que en un principio estaban planificadas para un periodo posterior, se realizasen antes debido al PDR, lo que no corresponde con una sustitución intertemporal de inversiones.
Notes
1 Additionality is one of the key principles of EU funding. Three types of potential additionality can be identified: project level, programme level and at member state level. Although additionally is a more general concept, this paper considers financial additionality at a firm level; that is, whether it stimulates investment expenditure at a firm level relative to the situation without the support.
2 For empirical estimations it is important to identify a control group as similar as possible to treated farms. Yet, some farms (control farms), irrespective of whether support is provided, may not be willing to invest due to a number of reasons, e.g. lack of a farm successor. As the latter factor is usually an unobservable, i.e. cannot be derived from micro-data, it would be inappropriate to compare farms that received investment support with all those others which did not invest. In order to circumvent this problem, selected into the potential control group were only those farms that were ‘willing to invest', i.e. those that in a given period undertook analogous investment (i.e. modernization of buildings), yet at various intensity levels.
3 The range 2000–06 refers to the years 2000–06, whereas the period of analysis is 2001–07 (see further below).
4 Given that the RDP support is project based, the start of granting actual support usually does not correspond with the actual start of the financial period (i.e. the year 2000), and often is delayed because of the time needed to set-up the granting system and to implement the actual selection of the submitted projects.
5 FADN is a European system of farm surveys that take place every year and which collects structural and accountancy information on farms. Farms are selected to take part in the survey based on stratified sampling of farms.
6 For specification test results see Appendix A1 in the Supplemental data online.
7 The literature does not have a specific rule with respect to which covariates should be selected in the estimation of the PSM function. In general, besides applying economic theory and using empirical evidence, there are three possible strategies for the selection of covariates (Caliendo and Kopeinig, Citation2008): (1) a hit or miss method (Heckman et al., Citation1997); (2) a statistical significance method; and (3) a leave-one-out cross-validation (Black and Smith, Citation2004). The present paper has followed the empirical evidence on the implementation of AFP to select the covariates. More precisely, the selection of covariates was based on expertise of branch specialists (beef and milk production) by taking into consideration a prerequisite that selected covariates simultaneously have to affect a farm's decision to participate in a programme as well as outcome variables.
8 The expression ‘corrected' is specific to the variable ‘profit per farm'. Corrected/adjusted profit per farm means current profits corrected for specific revenues and expenses linked to other periods (current profits minus specific revenues generated in other periods plus specific costs related to other periods). Corrected profits include also adjustment for taxes and other payments prepaid or received in relation to other periods.
9 The perishability of milk, the need to maintain strict sanitary control and the relatively high transportation costs of fluid milk lead to regional differentiated adjustments of milk prices.
10 Due to the dropping of programme non-treated farms located in regions with the highest programme intensity from the database, i.e. regions Nordfriesland and Schleswig-Flensburg, and those farms where the modernization of farm buildings did not take place, only 161 non-treated (control) farms were left to re-estimate crowding-out effects.
11 For specification test results, see Appendix A2 in the Supplemental data online.
12 For specification test results see Appendix A3 in the Supplemental data online.
13 To check the robustness of the results further, this paper estimated the leverage effects for the same group of farms that were included in the computation of crowding-out effects (i.e. the subsample of only those farms that were willing to invest and excluding programme-affected non-treated farms, as reported in ). The results show that (1) additional money transfers from farm to farm households for living expenses increased by €2575 compared with control farms; (2) the building of private non-farm assets decreases by –€6253; and (3) the total leverage effect increased by €5291. Apparently treated farms in comparison with similar control farms that were also willing to invest did not have a preference for transfers from farm aiming at increasing private non-farm assets (negative value), yet transfers to farm household for living expenses were positive (€2575) and also the total leverage effect was strongly positive (€5291).
