Knowledge and innovation: An empirical analysis of firm’s organization in the Baltic countries

The aim of this paper is to investigate the relationship between firm’s innovation and knowledge sources in Baltic countries. We focus on Estonia, Latvia and Lithuania, since for these countries, the relationship between innovation and knowledge is unexplored from an empirical point of view. There is a theoretical background on the different types of knowledge that influence a firm’s innovation level. By using firm-level data of the Community Innovation Survey (CIS 2014), we observe this relationship in the specific sub-category of the manufacturing sector in Baltic countries, and consider a range of control variables. Our findings highlight a positive and consistent relationship between knowledge sources and innovation.

Knowledge and Innovation: an Empirical Analysis of Firm’s Organization in the Baltic Countries Annunziata de Felice¹, Antonella Biscione² and Isabella Martucci¹ ¹Department of Law University of Bari Aldo Moro Bari Italy ²Department of Bioeconomic Strategies in the European Union and in the Balkans, Catholic University “Our Lady of Good Counsel” annunziata.defelice@uniba.it isabella.martucci@uniba.it a.biscione@unizkm.al Abstract: The aim of this paper is to investigate the relationship between firm’s innovation and knowledge sources in Baltic countries We focus on Estonia, Latvia and Lithuania, since for these countries, the relationship between innovation and knowledge is unexplored from an empirical point of view There is a theoretical background on the different types of knowledge that influence a firm’s innovation level By using firm-level data of the Community Innovation Survey (CIS 2014), we observe this relationship in the specific sub-category of the manufacturing sector in Baltic countries, and consider a range of control variables Our findings highlight a positive and consistent relationship between knowledge sources and innovation Keywords: knowledge, innovation, Baltic Countries, manufacturing sector Introduction In recent years, there has been increased attention, among scholars and policy-makers, to the role played by different knowledge sources in the creation of innovation Several studies have highlighted how knowledge is associated with innovation (see among others Nonaka and Takeuchi, 1995; Van den Bergh, 2008; Schoenmakers and Duysters, 2010) In order to remain competitive, a firm must be innovative otherwise it is destined to disappear (Abbing, 2010; Cho and Pucik, 2005; Zawislak et al., 2012) In other words, the firm requires constant innovative actions to increase its competitiveness and maintain its market share Moreover, the competitiveness is closely linked to the process of knowledge accumulation (de Felice et al 2012) The capacity of accumulation of knowledge that produces innovation is strictly connected to the acquired competences and, above all, to those reached through the research and networks These relationships are stressed by knowledge-based literature (Camuffo and Grandinetti, 2006, 2011; Maskell, 2001; Grandinetti and Tabacco, 2003; Rullani, 2003) and relational literature (Wasserman and Faust, 1994; Powell, 1990; Giuliani and Bell, 2005; Ter Wal, 2013 e.g.) In particular, one part of the literature emphasizes the importance of “open innovation” (Chesbrough, 2003, 2006) or of the integration between external and internal knowledge in a firm to promote innovation The external relationship with suppliers, competitors, clients and university becomes essential to develop new products or new processes (Powell and Grodal, 2005; Huggins, 2010; Athaide and Zhang, 2011) Knowledge management plays a key role, using new methods of organizing external relations with other firms or public institutions (i.e use of alliances, partnerships, outsourcing or sub-contracting), establishing new business practices for organizing procedures (i.e supply chain management, business reengineering, lean production, quality management, etc.) and/or putting in place new methods of organizing work responsibilities and decision making In the light of these considerations, in this paper we provide an empirical contribution on the relationship between knowledge and innovation using firm-level data provided by the Community Innovation Survey (CIS 2014) based on the Oslo Manual (OECD, 2005) in the specific sub-categories of the manufacturing sector in the Baltic countries First of all, we examine the companies of three countries considering them as one economy ISSN 1479-4411 34 ©ACPIL Reference this paper: de Felice, A., Biscione, A., and Martucci, I., 2019 Knowledge and Innovation: an Empirical Analysis of Firm’s Organization in the Baltic Countries The Electronic Journal of Knowledge Management, 17(1), pp 34-48, available online at www.ejkm.com Annunziata de Felice, Antonella Biscione and Isabella Martucci Then, we observe the relationship between firm’s innovation and knowledge in the three countries in order to underline their differences We exploit a baseline regression and eventually we present some robustness tests For sake of robustness, we employ two different innovations: product and process In sum, results show that knowledge sources are positively and significantly associated with innovation In other words, knowledge appears to increase innovation This result appears to be robust across different specifications However, other contradictory results are also worth noting For example, knowledge management appears to have no effects when considering the three countries as one economy The paper is structured as follows In the next section, we focus on the literature review and conceptual background of innovation and its determinants The third section describes the data and the methodology; the results are discussed in the fourth section, followed by the conclusions in the fifth Literature and Conceptual Background In this section we try to expound the conceptual background of this work by surveying the existing literature on the linkage between innovation and knowledge sources The literature on innovation proposes a plethora of definitions, classifications, types and determinants of innovation According to Schumpeter (1971, 1977) innovation is a process of creative destruction or “new combinations” of existing resources, involving the introduction of new goods and/or new production processes to create a new organization, a new trade, or a new form of marketing which results in access to a new supply source of raw or semi-finished materials in a new industrial organization This technological change concerns not only the firms and the users, as technology sellers and buyers, respectively, but also public institutions; each of them, with their experience, contributing to the technological changes A part of the literature (Van Dijk et al., 1997; Schumpeter, 1971) highlights that large firms could be more innovative compared to small ones: the large ones having the market power and large profits necessary to finance R&D Other studies (Cohen and Levinthal, 1990; Romer, 1990; Van Dijk et al., 1997) observe the innovation of small firms, which benefit from public funding and from local markets Cohen and Levinthal, 1990; Romer, 1990; Van Dijk et al., 1997; Kang and Park, 2012 emphasize the importance of public funds for small firms in order to be more innovative It is important to point out that even if firms are characterized by size, they not innovate at the same time nor they make the same type of innovation A great number of empirical analyses provide evidence that knowledge is the driving force of innovation, particularly, the tacit dimension as well as the external and internal relationship within the organization (Nelson and Winter, 1982; Nonaka and Takeuchi, 1995; de Felice et al., 2012; Tsai, 2001) These relations allow innovation to develop within a firm’s organizational structure Significant are the channels of knowledge transfer such as local institutions, associations and university or research centers, to create innovation (Rigby et al., 2002; Camuffo and Grandinetti, 2006, 2011; Landry et al., 2002; Barrutia and Echebarria, 2010) Firms should combine external relationships and networks with their internal knowledge to develop innovation (Chesbrough, 2003, 2006; Chiaroni et al., 2011; Martin-de Castro, 2015) So, the firm’s knowledge management is crucial to complement external and internal knowledge in its organization (Probst et al., 2002) The firm’s internal resources are important in order to benefit from external knowledge This idea is based on absorptive capacity theory that show the important role of technological and organizational dimensions (Cohen and Levinthal, 1990; Zhara and George, 2002) in order to acquire, integrate and develop the external knowledge (Granero et al., 2014) In addition, to internalize technical knowledge and technology, foreign direct investment (FDI) is crucial, since this is a source of external technological knowledge (Buckley and Kafouros, 2008, Erdal and Gocer, 2015; Crescenzi et al, 2015) The effect of FDI on local firms is not always positive (Nam et al.,2012), in fact, the www.ejkm.com 35 ISSN 1479-4411 The Electronic Journal of Knowledge Management Volume 17 Issue 2019 positive effects of FDI in the host country depends on the firms’ absorptive capacity of the external knowledge available by FDI inflows (Banri et al., 2012) As in Baltic Countries there are few descriptive analyses, this paper can be considered as an empirical contribution that takes into account only these three countries Data and Methodology The aim of this study is to explore how the channels of knowledge transfer – external and/or internal – created innovation over the period covered by the CIS survey (2012-2014) for the Baltic countries We focus on product and process in the manufacturing sector We observe if this relationship is different and/or similar for different manufacturing sub-sectors (from C10 to C33) classified according to the taxonomies of the standard Eurostat classifications (NACE Rev.2) (Table 1) Table 1: NACE sample classification Divisions 10 - 12 Description Manufacture of food products, beverages and tobacco products Acronym Foodbev 13 - 15 Manufacture of textiles, wearing apparel and leather and related products Textiles 16 - 18 Manufacture of wood and of products of wood and cork, except furniture; manufacture of articles of straw, plaiting materials and manufacture of paper and paper products, and printing and reproduction of recorded media Woodpap 19 - 23 Manufacture of coke and refined petroleum products, of chemicals and chemical products, of basic pharmaceutical products and pharmaceutical preparations, of rubber and plastic products and of other non-metallic mineral products Cochem 24 - 25 Manufacture of basic metals and of fabricated metal products, except machinery and equipment Metals 26 - 30 31 - 33 Manufacture of computer, electronic and optical products, of electrical equipment, of machinery and equipment n.e.c., of motor vehicles, trailers and semi-trailers and of other transport equipment Manufacture of furniture, Other manufacturing and Repair and installation of machinery and equipment Elecmot Foroth We study this relationship in the Baltic area considering it as one economy, although in fact the three countries each have a unique history Together they became independent, were invaded by the Soviets, later by the Nazis and then again by the Soviets Together they regained independence and entered the EU and NATO They have a similar territorial and demographic weight too Baltic countries are characterized by common features such as similar history and economic structure, which gradually converges to the standards of the European Western countries and they are specialized in low-tech productions, but, each country has its peculiar economic characteristics In fact, Estonia is the most developed with a GDP per capita of 76% of the EU28 average in purchasing power parity; Latvia has a GDP per capita of 64% of the EU28 average being less influenced by regional spillovers; Lithuania, although presenting a larger territorial extension than the other two countries, has a GDP per capita the same as Estonia (Poissonier, 2017) The great recession beginning in 2007 in the USA influenced all three Countries, although Latvia was the most affected The Baltic Countries all specialize in wood and paper products as well as furniture and textiles, with In Lithuania, manufacturing sector employed for 27.6% the FDI which, in 2014, mainly derived from Sweden, followed by the Netherlands and Poland (Lithuanian Statistical Office) 84% of FDI in Estonia (Estonia Statistical Office) also come from Sweden, followed by Finland (22.3%) and the Netherlands (8%), while Estonian IDEs are headed to Lithuania (23.3%, Cyprus and Latvia (18%) Latvia, on the other hand, has known an increase in FDI of 3.7% compared to the previous year (Latvia Statistical Office) The distinction between a product and a process innovation is crucial in order to identify the different firms’ strategy Product innovation is associated with more radical strategies; process innovation prevails in small and traditional firms www.ejkm.com 36 ©ACPIL Annunziata de Felice, Antonella Biscione and Isabella Martucci other areas of specialization including the food sector and pharmaceutical industry in Latvia; the food sector and chemicals in Lithuania; and oil, electronics and electrical equipment in Estonia (Poissonier, 2017) Due to its proximity, Estonia has important relationships with Finland and has become a country where the tech sector is important Estonia also adopted the Euro in 2011 Lithuania maintains a stronger relationship with Poland and Central Europe and adopted the Euro in 2015 Latvia occupies a strategic position and acts as a transit route, connecting the Western and Scandinavian countries with Russia and the other ex-Soviet republics Latvia adopted the Euro in 2014 For this reason, we also analyze this relationship country by country For this purpose, we use data from the Community Innovation Survey (CIS), a survey about innovation activities in enterprises covering European Countries and we refer to CIS14 (2012-2014) In order to compare the evolution of firms in the manufacturing sector, Eurostat develops a standard questionnaire accompanied by a set of definitions and methodological recommendations also based on the Oslo Manual (2005) The surveys provide information about a sample of 2449 firms The sample for each Baltic Country is divided in the following way: (i) in Estonia there are 941 firms; (ii) in Latvia 532 firms and, finally, (iii) in Lithuania 976 firms (Table 2) These firms have 10 employees or more Table 2: Observations number of the firm for NACE divisions in manufacturing sector Latvia Lithuania Estonia 2014 2014 2014 divisions Freq Obs Freq Obs Freq Obs 10 - 12 329 82 473 94 185 112 13 - 15 214 56 397 95 200 129 16 - 18 577 110 626 155 300 150 19 - 23 216 99 367 204 161 134 24 - 25 210 55 289 117 268 86 26 - 30 147 81 231 173 190 166 31 - 33 243 49 542 138 233 164 Total 1936 532 2925 976 1538 941 Source: our elaboration on data CIS 2014 Table shows the distribution of firms in Baltic Countries by size based on CIS data In all three countries most of them (57.17% in Estonia; 47.78% in Latvia; 49.69% in Lithuania) are small-sized firms, the others are medium-and large-sized firms (42.82% in Estonia; 52.22% in Latvia; 50.31% in Lithuania) Table 3: Distribution of firms by their size and Countries EE Size Small Medium Large Total Frequency 538 353 50 941 LV Percentage 57.17 37.51 5.31 100.00 Frequency 258 226 56 540 LT Percentage 47.78 41.85 10.37 100.00 Frequency 485 360 131 976 Percentage 49.69 36.89 13.42 100.00 Source: our elaboration on data CIS 2014 Table reveals the incidence of innovation by firm type Product and process innovation is contextually diffused among medium firms in all three countries (8.18% and 7.65% in Estonia; 13.14 and 12.22% in Latvia; 14.45% and 19.67 in Lithuania, respectively) www.ejkm.com 37 ISSN 1479-4411 The Electronic Journal of Knowledge Management Volume 17 Issue 2019 Table 4: Incidence of Innovation by firm dimension (%) EE Size Small Medium Large Total Product 6.81 8.18 1.80 15.72 LV Process 5.53 7.65 2.34 15.51 Product 7.60 13.14 4.26 25.00 LT Process 5.18 12.22 4.63 22.03 Product 11.47 14.45 8.71 34.63 Process 13.62 19.67 9.84 43.14 Source: our elaboration on data CIS 2014 In addition, all three Baltic countries and all divisions of the manufacturing sector (Table 5) present a good performance in innovation activities both in product and process Particularly, Lithuania has seen the most important value in product and process innovation in 2014 Table 5: Innovation in manufacturing sub-sector Divisions 10 - 12 13 - 15 16 - 18 19 - 23 24 - 25 26 - 30 31 - 33 Total LV % 17.88 5.70 4.30 16.94 11.02 28.12 9.84 11.40 Product Innovation LT EE % % 39.51 30.92 30.96 6.17 11.91 10.73 22.57 14.03 29.88 11.17 42.67 18.77 19.82 10.44 25.97 13.94 Process Innovation LV LT EE % % % 14.27 42.41 22.52 5.23 27.10 8.88 11.01 24.74 18.54 17.59 34.81 13.62 9.71 39.13 8.85 23.20 47.10 15.24 5.54 31.06 11.76 11.75 33.54 14.12 Source: our elaboration on data CIS 2014 Table shows the variables employed in this analysis Table 4: Variables used in the econometrics analysis Variables Type Innovation Dummy Product innovation (product innovation) Process innovation (process innovation) Dummy Dummy External knowledge relation (external knowledge) Dummy Research and Development (R&D) Dummy Public Founding (public founding) Dummy Knowledge Management (knowledge management) Dummy FDI Orientation in foreign market Profit Performance: Dummy Dummy Dummy Turnover (Turn1) Turnover (Turn2) Turnover (Turn3) Firm Size: Dummy Small (small) Medium (medium) Large (large) Sectors: Foodbev Dummy Textiles Woodpap Cochem Metals Elecmot Foroth www.ejkm.com 38 ©ACPIL Annunziata de Felice, Antonella Biscione and Isabella Martucci The dataset provides detailed firm-level information not only on innovation behaviors by type of enterprise and by sector, but also on various aspects of the development of an innovation, such as objectives, sources of knowledge and information, public funding or expenditure It is important to point out that the dataset used has some limitations, therefore we use dichotomous variables in order to bypass these restrictions Innovation is the dependent variable In our analysis we considered the Schumpeterian definition of innovation described in the theoretical framework It includes the technological innovation or the introduction of new products (product innovation) and new processes (process innovation) by a firm of the sample For these variables we use a binary indicator because we distinguish between innovative firms (1) and non-innovative firms (0) When we refer to product innovative firms, we consider firms which introduced a new or significantly improved good or service For process innovative firms we include firms which have carried out one of the follow strategies: Introduction of new or significantly improved methods of manufacturing or producing goods or services; introduction of new or significantly improved logistics, delivery or distribution methods for firm’s inputs, goods or services; New or significantly improved supporting activities for firm’s processes, such as maintenance According to the theoretical part, the most important independent variable that we have selected is the knowledge Knowledge is distinguished as external relations knowledge (external knowledge) and knowledge management (KM) We define external relations knowledge as the tacit/explicit dimension of knowledge that are sourced or derived from cooperation in innovation activities through external and distance relationships We include in this variable cooperation arrangements with other enterprises, suppliers, clients, competitors, consultants, institutions, universities and research institutes located in the same country and/or in Europe, in EFTA Countries, in EE-CC, in the United States, in China, India, and in other countries These relationships are external and geographically distant We use a dummy variable which is equal to if a firm has cooperated with at least one different external partner for each category We use knowledge management as a firm’s innovation capabilities depend on the intellectual assets and knowledge that managers have (Subramaniam and Youndt, 2005) Knowledge management also has the role of sharing the tacit or codified dimension of knowledge in the firm or in the enterprise group through face to face interactions (Koskinen et al., 2003) and though social interactions (Nonaka and Takeuchi, 1995) Using the CIS14 questionnaire, we construct the knowledge management variable It is equal to if a firm, over the period 2012-2014, organized external relations with stakeholders, or/and it established new business practices for organizing procedures and/or it used new methods for organizing work responsibilities and decision making In order to innovate, another important variable is expenditure in R&D or the innovative effort This represents the commitment to innovation (OECD, 2005; Malerba, 2005) and catches firms’ absorptive capacity (Cohen and Levinthal, 1990; Aghion and Jaravel, 2015) Relying on the CIS, we have considered the research activity in internal or in-house activities that create new knowledge or solve scientific or technical problems and external R&D that the firm has contracted out to other enterprises We use a dummy which is equal to if the firm invests in R&D In the Baltic Countries an important role is played by public funding to support the innovation activity This support comes from local or regional authorities; central government; the European Union or the EU 7th Framework Programme It is a dummy variable which is equal to if the firm has received at last one amount of public funding among these institutions We include other important regressors including firm size, profit performance, orientation in foreign markets, FDI, and firm sectors, because these are some of the innovation www.ejkm.com 39 ISSN 1479-4411 The Electronic Journal of Knowledge Management Volume 17 Issue 2019 determinants generally used in empirical studies (i.e Acs and Audretsch, 1988; Kleinknecht, 1989; Shefer and Frenkel, 2005; Martinez-Ros and Labeaga, 2010) With reference to the firm size, CIS does not provide the exact number of employees, but rather a range and firms with less than 10 employees are excluded from the sample The size classes used with the correspondent value is the following: if the firm has 10-49 employees (small); if the firm has 50-249 employees (medium); if the firm has more than 250 employees (large) The profit performance of the firm is measured by turnover growth rate We expect that high levels of performance can facilitate growth and subsequent profit performance (Price et al 2013) and innovation We distinguish three categorical variables: turnover if the firms have a turnover growth rate negative or equal to 0, turnover if the firms have a turnover growth rate between 0.1 and and turnover if the firms have a turnover growth rate more than Another important variable is the orientation in the foreign markets that we define as the export propensity of the firm Recent empirical studies have studied the relationship between exports and innovation comparing the export performance of innovative and non-innovative firms (Castellani and Zanfei, 2007; Cassiman and Golovko, 2011; Hwang and Dong 2015; Lopez-Bazo and Motellòn, 2013) According to Choi (2015) exporting firms tend to invest more in product and process innovation We construct this dimension as a dummy variable of if the firm sells goods/services in local, National, EU and extra EU market In our analysis, we consider the sector variable as having a significant influence on innovation activities For this reason, we have distinguished the sectors according to the NACE classification (see Table 1) Also the foreign direct investment (FDI) is considered as an important determinant for innovation, and is defined “as a form of inter-firm cooperation that involves a significant equity stake in, or effective management control of foreign enterprise” (Erdal et al., 2002) It is a dummy and its value is if in the three Baltic countries there are branches and contextually there are patents Table presents the set of variables used, Table shows the descriptive statistics Table 7: Descriptive Statistics Variables Innovation Process Innovation Product Innovation External Knowledge relation Public Funding Research & Development Orientation in Foreign Market Knowledge Management Firm Size Small Large Medium Sectors Foodbev Textiles Woodpap Cochem Metals Elecmot Foroth Profit Performance Turnover Turnover Turnover Foreign Direct Investment Countries Estonia Lithuania Latvia www.ejkm.com Obs 2457 2457 2457 2457 2457 2457 2457 2457 Mean 0.355 0.279 0.253 0.128 0.132 0.208 0.832 0.189 Std Dev 0.479 0.449 0.435 0.334 0.338 0.406 0.374 0.391 Min 0 0 0 0 Max 1 1 1 1 2457 2457 2457 0.521 0.382 0.096 0.500 0.486 0.295 0 1 2457 2457 2457 2457 2457 2457 2457 0.118 0.115 0.170 0.178 0.106 0.171 0.143 0.322 0.318 0.375 0.383 0.308 0.377 0.350 0 0 0 1 1 1 2457 2457 2457 435 0.240 0.495 0.094 0.115 0.427 0.500 0.291 0.319 0 0 1 1 2457 2457 2457 0.383 0.397 0.220 0.486 0.489 0.414 0 1 40 ©ACPIL Annunziata de Felice, Antonella Biscione and Isabella Martucci The Empirical Model And Results Given the binary nature of the dependent variable, we use the following Probit model: Pr (𝑌𝑖 = ) = 𝜙(𝑋, 𝛽) 𝑋 where 𝜙 is the cumulative distribution function of the standard normal distribution and 𝑌𝑖 is the dummy variable that takes value if the firm i introduces innovation and otherwise X is a set of covariates described in Table The following tables present the results of a set of Probit regressions estimating the impact of external knowledge and management knowledge on the probability of introducing an innovation and the probability of introducing a product and a process innovation Table reports the main results Table 8: The impact of knowledge on the propensity to innovate External Knowledge Public Founding R&D Foreign Market Knowledge Management Small( as reference) medium large Foroth( as reference) Foodbev Textiles Woodpap Cochem Metals Elecmot Turn1 ( as reference) Turn2 Turn3 FDI Estonia ( as reference) Lithuania Latvia Constant N obs Baltic countries Dependent variable: innovation dy/dx 1.237*** 0.401*** [0.312] [0.069] 1.649*** 0.456*** [0.501] [0.063] 2.266*** 0.649*** [0.297] [0.051] 0.163 0.064 [0.244] [0.096] Estonia Dependent variable: innovation dy/dx 1.259*** 0.364*** [0.307] [0.118] 2.013*** 0.650*** [0.624] [0.203] 2.490*** 0.740*** [0.209] [0.054] 0.421** 0.061*** [0.193] [0.026] Lithuania Dependent variable: innovation dy/dx 1.862*** 0.558*** [0.423] [0.055] 3.027*** 0.693*** [0.327] [0.026] 1.732*** 0.549*** [0.274] [0.044] 0.413** 0.159** [0.177] [0.066] Latvia Dependent variable: innovation dy/dx 2.885*** 0.850*** [1.123] [0.172] 2.078*** 0.644*** [0.598] [0.205] 2.741*** 0.810*** [0.362] [0.076] -0.082 -0.012 [0.278] [0.041] 0.135 [0.220] 0.052 [0.083] 0.108 [0.292] 0.020 [0.057] 1.105*** [0.217] 0.409*** [0.065] 1.050*** [0.327] 0.241** [0.108] 0.638*** [0.221] 1.149*** [0.350] 0.239*** [0.082] 0.361*** [0.081] 0.389** [0.158] 0.523** [0.270] 0.076** [0.031] 0.122 [0.075] 0.277** [0.142] 0.798*** [0.256] 0.110** [0.056] 0.304*** [0.087] 0.455* [0.270] 1.100*** [0.351] 0.077 [0.052] 0.287** [0.123] 0.934*** [0.335] 0.314*** [0.088] 1.021*** [0.304] 0.101*** [0.101] 0.295 [0.299] 0.117 [0.119] -0.274 [0.507] 0.709* [0.384] 0.975*** [0.300] 0.721* [0.418] 1.259*** [0.407] 1.391*** [0.380] 0.239** [0.103] 0.326*** [0.079] 0.252** [0.120] 0.372*** [0.073] 0.393*** [0.063] 0.471 [0.333] 0.890*** [0.312] 0.544 [0.345] 0.655 [0.430] 0.752** [0.340] 0.101 [0.083] 0.212*** [0.092] 0.123 [0.094] 0.147 [0.121] 0.182* [0.102] 0.291 [0.276] 0.027 [0.235] -0.030 [0.239] 0.280 [0.254] 0.067 [0.260] 0.115 [0.110] 0.011 [0.092] -0.012 [0.094] 0.111 [0.101] 0.026 [0.103] -0.913 [0.584] -0.019 [0.445] 0.458 [0.514] -0.390 [0.457] -0.167 [0.498] -0.034 [0.056] 0.079*** [0.030] -0.003 [0.063] 0.084 [0.115] -0.045 [0.044] -0.021 [0.058] 0.101 [0.207 0.267 [0.381 0.039 [0.080] 0.100 [0.140] -0.100 [0.178] -0.063 [0.332] 0.099 [0.147] 0.323 [0.238] 0.039 [0.058] 0.128 [0.094] -0.230 [0.260] 0.363 [0.302] -0.031 [0.035] 0.059 [0.059] -0.103 [0.409] -0.040 [0.159] -0.879 [0.620] -0.017 [0.031] -0.011 [0.054] 0.087*** [0.027] -0.640 [0.716] -0.224 [0.208] 2.218*** [0.639] 0.704*** [0.191] 1.280*** [0.247 0.853*** [0.269 0.452*** [0.076] 0.287*** [0.080] -3.151*** [0.396] 443 2.958*** [0.334] 881 1.557*** [0.266] 886 -1.908*** [0.490] 468 Standard errors in brackets; statistical significance *** p < 0.01, ** p < 0.05, * p