Political connections, network centrality and firm innovation

Political connections, network centrality and firm innovation

Accepted Manuscript Political connections, network centrality and firm innovation Li-Chuan Tsai, Ruhui Zhang, Cuifang Zhao PII: DOI: Reference: S154...

591KB Sizes 0 Downloads 3 Views

Accepted Manuscript

Political connections, network centrality and firm innovation Li-Chuan Tsai, Ruhui Zhang, Cuifang Zhao PII: DOI: Reference:

S1544-6123(18)30229-0 10.1016/j.frl.2018.04.016 FRL 926

To appear in:

Finance Research Letters

Received date: Accepted date:

2 April 2018 27 April 2018

Please cite this article as: Li-Chuan Tsai, Ruhui Zhang, Cuifang Zhao, Political connections, network centrality and firm innovation, Finance Research Letters (2018), doi: 10.1016/j.frl.2018.04.016

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Political connections, network centrality and firm innovation✩ Li-Chuan Tsaia,∗, Ruhui Zhangb , Cuifang Zhaoc Jiangxi University of Finance and Economics, Nanchang, China 330013 b National Tsing Hua University, Hsinchu, Taiwan 30013 c National Taiwan University, Taipei, Taiwan 10617

CR IP T

a

Abstract

This paper examines how the strength of political connections affects innovation. Considering complex relationships between firms and governments as well as different powers of bureaucracy

AN US

ranks, we model political connections of all Chinese listed firms as weighted social networks and measure their strength by network centralities. The empirical results show that firms with stronger political connections innovate more. This paper further identifies two possible mechanisms of this finding: political connections help firms get more government subsidies and intellectual capital, both of which are verified to promote innovation with statistical significance.

M

Keywords: social network; political connections; firm innovation

ED

1. Introduction

The presence of political connections in listed firms has been found globally (Faccio, 2006; Fan

PT

et al., 2007; Goldman et al., 2009). Pervasive literature finds that political connections provide various benefits for firms. Politically connected firms can get favorable resources such as bank loaning

CE

(Khwaja and Mian, 2005) and government subsidies (Wu and Cheng, 2011), a lower interest rate of loan and tax rate (Faccio, 2006). Firms with political connections also enjoy supportive government

AC

policies and regulations (Al-Hadi et al., 2016; Agrawal and Knoeber, 2001). Extensive evidence supports that political connections do influence firm performance (Niessen and Ruenzi, 2010; Li et al., 2008; Wu et al., 2012) and firms’ strategic choices such as corporate philanthropies (Li et al., Declarations of interest: none. Corresponding author. Address: Room H219, School of Software and Communication Engineering, Jiangxi University of Finance and Economics, Nanchang, China 330013 Email addresses: [email protected] (Li-Chuan Tsai), [email protected] (Ruhui Zhang), [email protected] (Cuifang Zhao) ✩ ∗

Preprint submitted to Elsevier

April 30, 2018

ACCEPTED MANUSCRIPT

2015), corporate bailouts (Faccio, 2006), and corporate transparency (Leuz and Oberholzer-Gee, 2006). However, few studies investigate how political connections affect firm innovation, which has been widely recognized as a key driver for national- and firm-level competitiveness (Porter, 1992). Benefits of political connections on innovation are mainly reflected in two aspects. First, new knowledge and information are crucial to innovation (Drucker, 1994; Hall et al., 2005). Politically

CR IP T

connected firms may be more informed of government policies and information relevant to innovation. Second, as Porter (1992) states, “Innovation and upgrading come from sustained investment in physical as well as intangible assets.” Political connections may boost innovation by facilitating the access to more tangible resources like government subsidies. Regarding the channel of intangible resources, social capital literature (Rothstein and Stolle, 2002; Brehm and Rahn, 1997; Hall,

AN US

1999; Kumlin and Rothstein, 2005) claims that government institutions and policies can create, channel and affect social capital such as generalized trust. Nahapiet and Ghoshal (2000) point out that firms with more dense social capital have advantages in creating intellectual capital. The

capital, such as human capital.

M

generalized trust induced by political connections, may in turn help firms acquire more intellectual

Furthermore, it’s noted that resources brought by political connections may vary with the

ED

strength of political connections, and researchers attempt to capture this strength with different measures. For example, one of the proxies for the strength of political connections in Francis et al.

PT

(2009) is the number of board directors who work or worked in government bureaucracies. On the other hand, Ling et al. (2016) take into account the political ranks of chief officers and construct a

CE

composite score to measure the strength of political connections for firms. Though aforementioned measures give some profound thoughts, they coincide with measuring the connections merely from

AC

firms’ perspective. Standing in the shoes of government institutions may provide a different way of seeing things, resulting in insights otherwise unobservable. For instance, leaders from different firms serving in the same government institution may share information and exchange/compete resources in one way or another. To reconcile above reflections while maintaining focus on firms, this paper models the political connections of firms as a social network where the vertices represent board members, senior exec-

2

ACCEPTED MANUSCRIPT

utives or supervisors (henceforth top leaders) and the edges are linked when two leaders serve in the same government institution. Besides, edge weights are added based on the bureaucracy rank. The political connection strength for each firm is the aggregation on network centralities of all its top leaders. Utilizing social network analysis, this paper aims to investigate how the strength of political

CR IP T

connections of Chinese listed firms affects innovation. To dig into the effects of political connections on innovation, two channels are further identified. In all, our paper contributes to the existing literature in several ways.

1. The influences of social network characteristics are increasingly studied in economic and fi-

AN US

nancial fields (Chuluun et al., 2017), such as portfolio choice (Massa and Simonov, 2004), stock market participation (Hong et al., 2004) and mutual fund portfolio decisions (Fu and Gupta-Mukherjee, 2014). This paper broadens previous studies of social networks by examining the effects of network centrality on firm innovation.

M

2. The relationship between political connections and firm innovation is first studied in Hou et al. (2017), which uses a dummy variable to measure political connections of non-state-

ED

owned companies and finds that political connections impede innovation. Instead of using the notion of connection existence, we evaluate the connection strength among these politically

PT

connected firms by network centralities and find a positive relationship between the strength of political connections and innovation.

CE

3. This paper inspects two possible channels that explain the positive effects of political connections on innovation. The first observed channel includes two positive relationships: political

AC

connections on government subsidies, as well as government subsidies on innovation. Besides, this paper also shows a pioneering channel that political connections help firm obtain greater intellectual capital that in turn facilitates innovation.

3

ACCEPTED MANUSCRIPT

2. Data sample and variable description 2.1. Data sample The data sample covers all Chinese listed companies on Shanghai Stock Exchange (SHSE) and Shenzhen Stock Exchange (SZSE) from 2008 to 2014. Most of the data are obtained from the China Stock Market & Accounting Research Database (CSMAR), and the number of employees

Economic Journal (TEJ) database. 2.2. Network structures and political connections measurement

CR IP T

held Master’s degree or above, used for measuring intellectual capital, is obtained from the Taiwan

We construct annual weighted social networks to model the political relationships of top leaders

AN US

in the listed firms. In such networks, vertices represent top leaders and edges demonstrate their co-attendance in the same government institution, where a higher level bureaucracy is assigned with a greater weight. Table 1 shows the weights corresponding to different bureaucracy levels. Next we compute the strength of political connections for a firm by the following steps. (1)

M

For any individual top leader of the firm in the networks, calculate three centralities, including degree centrality, betweenness centrality and closeness centrality. (2) For each centrality measure,

ED

sum over centrality values of all the top leaders in the firm. (3) Conduct a principal component analysis and choose the component with the largest variance to represent the strength of political

PT

connections. In particular, this selected component accounts for 99% variance of the original data.

AC

CE

Table 1: Weights of different bureaucracy levels

Level of bureaucracy Weight Central government 5 Provincial level 4 Prefectural level 3 County level 2 Township level and below 1

2.3. Innovation variable Following FANG et al. (2014), Tian and Wang (2014) and Lerner et al. (2011), innovation output, denoted by Innovation, is measured by the applications of patents, which in this paper are calculated in five ways for robustness. The first measure (Apply) is the number of patent 4

ACCEPTED MANUSCRIPT

applications a firm registers in a year. The second (ApplyGrant) is the number of registered patent applications in a year that are finally granted until this study is conducted. The rest measures are calculated according to three categories of Chinese patent applications: invention, utility model and external design, denoted by IApply, U Apply and DApply, respectively. Afterwards these

right skewness. 3. Empirical results 3.1. The effects of political connections on innovation

CR IP T

variables are transformed to natural logarithm of the application number plus one to alleviate

AN US

At given time t, the influence of political connections (P Ci,t ) on firm i’s innovation output (Innovationi,t+1 ) is estimated by the following equation:

= a + b ∗ P Ci,t + c0 ∗ ControlV ariablesi,t + errori,t ,

M

Innovationi,t+1

where Innovationi,t+1 is measured by Applyi,t+1 , ApplyGranti,t+1 , IApplyi,t+1 , U Applyi,t+1 and

ED

DApplyi,t+1 , respectively. The control variables, denoted as ControlV ariablesi,t , are listed in the Table 2. Also, we fix the industry and year effects.

Definition The amount of money invested in the R&D process divided by the logarithm of total asset in a year of a firm. The logarithm of asset value of a firm. The ratio of market price to book price. The ratio of total debt to total asset. The ratio of net profit to total net asset. The ratio of cash value to total asset. The annual growth rate of income. Dummy variable, which equals 1 when a firm is state-owned. The ratio of market value to total asset . The number of years since a firm was established.

CE

Control Variable InnovSpend

PT

Table 2: The definition of control variables

AC

Asset M/B Leverage ROA CashAssetRatio SaleGrowRate SOE T obinQ F irmAge

The results of the ordinary least squares (OLS) regression are shown in Table 3. We observe 5

ACCEPTED MANUSCRIPT

that all five measures of innovation are positively correlated to political connection strength with statistical significance. Moreover, the values of adjusted R2 in most of the models are above 0.4. Table 3: The strength of political connections and innovation output

AC

CE

PT

ED

M

AN US

CR IP T

Applyt+1 ApplyGrantt+1 IApplyt+1 U Applyt+1 DApplyt+1 P Ct 3.6157*** 3.0720*** 3.8978*** 1.8590*** 1.2998*** (4.9950) (4.4354) (6.2068) (2.9529) (2.8182) InnovSpendt 0.5521*** 0.4858*** 0.5025*** 0.3580*** 0.2280*** (12.5594) (11.5507) (13.1757) (9.3646) (8.1406) Assett 0.4252*** 0.3889*** 0.4208*** 0.3410*** 0.1712*** (13.2719) (12.6873) (15.1411) (12.2394) (8.3892) M/Bt -0.024 -0.0334 -0.0095 -0.0524*** -0.0173 (-1.1403) (-1.6594) (-0.5229) (-2.8603) (-1.2920) Leveraget 0.0712 0.1309 0.1597 0.3759*** 0.135 (0.5750) (1.1047) (1.4868) (3.4905) (1.7112) ROAT 3.1526*** 3.0692*** 2.9692*** 2.5440*** 1.9972*** (7.3084) (7.4359) (7.9340) (6.7811) (7.2663) SaleGrowRatet -0.0699 -0.0567 -0.0447 -0.0522 -0.0481 (-1.0574) (-0.8966) (-0.7797) (-0.9080) (-1.1434) SOEt -0.0172 -0.0067 0.0577** 0.0404 -0.0294 (-0.5206) (-0.2136) (2.0081) (1.4023) (-1.3964) T obinQt 0.06 0.0524 0.0569 0.0754** 0.047 (1.4351) (1.3106) (1.5684) (2.0738) (1.7631) F irmAget -0.0413*** -0.0403*** -0.0303*** -0.0348*** -0.0077*** (-13.0616) (-13.2989) (-11.0401) (-12.6383) (-3.8363) Constant -6.8021*** -6.1935*** -7.2588*** -5.6296*** -3.1044*** (-11.3914) (-10.8401) (-14.0119) (-10.8402) (-8.1594) Industry Yes Yes Yes Yes Yes and year fixed effects Observations 9434 9434 9434 9434 9434 Adjusted 0.4892 0.469 0.4202 0.4508 0.2088 R2 F-statistic 96.11 88.71 72.99 82.51 27.21

Notes: The t-values are reported in the parentheses. *, ** and *** indicate statistical significance at the 10%, 5% and 1% levels, respectively.

Next, a two-stage least square (2SLS) regression is conducted to deal with possible endogeneity problems. Similar to (Ling et al., 2016; HOUSTON et al., 2014), our instrument variable is defined as the distance between the register office of a firm and Beijing. Table 4 shows the regression results 6

ACCEPTED MANUSCRIPT

in the second stage, which are consistent with our previous findings. Table 4: The results of 2SLS with instrument variable

Constant Control variables Industry and year fixed effects Observations 2nd-stage Adjusted R2 1st-stage F-statistic 2nd-stage F-statistic

Applyt+1 ApplyGrantt+1 IApplyt+1 U Applyt+1 DApplyt+1 80.5758** 94.0335** 27.6305 69.4679** 68.6416*** (2.0276) (2.4791) (0.8019) (2.0200) (2.7448) 1.0981*** 1.1054*** 0.3917** 0.7777*** -0.0116 (6.1692) (6.5065) (2.5383) (5.0490) (-0.1036) Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes 7474 7474 7474 7474 7474 0.4885 0.4689 0.4236 0.4523 0.2196 33.11 33.11 33.11 33.11 33.11 76.93 71.21 59.43 66.65 23.37

CR IP T

P Ct

AN US

Notes: The t-values are reported in the parentheses. *, ** and *** indicate statistical significance at the 10%, 5% and 1% levels, respectively.

4. Possible mechanisms 4.1. Government subsidies

M

The following work demonstrates two possible mechanisms through which political connections benefit innovation: government subsidies and intellectual capital. As mentioned earlier, Porter

ED

(1992) suggests both tangible and intangible assets are important input factors for innovation. It has been observed that managerial political connections can help firms get more government

PT

subsidies (Wu and Cheng, 2011), which we consider here as a tangible support for innovation. To test the relationships between political connections, government subsidies and firm innovation, we

CE

employ the following two linear models:

AC

GovSubi,t = a + b ∗ P Ci,t + c0 ∗ ControlV ariablesi,t + errori,t

Innovationi,t+1 = d + e ∗ P Ci,t + f ∗ GovSubi,t + g 0 ∗ ControlV ariablesi,t + errori,t ,

where GovSubi,t represents the amount of government subsidies that firm i receives in year t. From the results reported in Table 5, we find a possible channel that firms with stronger political connections obtain more government subsidies and thereupon the government subsidies improve innovation output. 7

ACCEPTED MANUSCRIPT

Table 5: The relationship between political connections, government subsidies and innovation

GovSubt Constant Control variables Industry and year fixed effects Observations Adjusted R2 F-statistic

-18.1339*** (-10.9213) Yes

Applyt+1 ApplyGrantt+1 IApplyt+1 U Applyt+1 DApplyt+1 1.3734** 1.0278 1.7241*** 0.1901 0.2377 (1.9875) (1.5545) (2.8758) (0.3164) (0.5409) 0.2822*** 0.2570*** 0.2754*** 0.2160*** 0.1182*** (17.5058) (16.6639) (19.6889) (15.4090) (11.5309) -1.7436*** -1.5883*** -2.3069*** -1.7033*** -1.1221*** (-4.3032) (-4.0966) (-6.5624) (-4.8330) (-4.3540) Yes Yes Yes Yes Yes

Yes

Yes

Yes

9434 0.2814 39.88

9434 0.4893 95.14

9434 0.469 87.82

CR IP T

GovSubt 7.7552*** (3.8577)

P Ct

Yes

Yes

Yes

9434 0.4203 72.24

9434 0.4507 81.64

9434 0.2117 27.38

AN US

Notes: The t-values are reported in the parentheses. *, ** and *** indicate statistical significance at the 10%, 5% and 1% levels, respectively.

4.2. Intellectual capital

Intellectual capital, as one type of intangible asset, plays an essential role in creating new

M

technology and productions, and is measured by the number of firm employees obtaining master

ED

degree or higher (Gogan and Draghici, 2013). We characterize the second mechanism through the following models.

PT

ICi,t = a + b ∗ P Ci,t + c0 ∗ ControlV ariablesi,t + errori,t

CE

Innovationi,t+1 = d + e ∗ P Ci,t + f ∗ IC i,t + g 0 ∗ ControlV ariablesi,t + errori,t , where the variable ICi,t stands for the intellectual capital of firm i in year t. From Table 6, we see

AC

this mechanism works as expected with statistical significance: the intellectual capital is positively correlated to the strength of political connections, and the coefficients of intellectual capital on innovation are all positive as well.

8

ACCEPTED MANUSCRIPT

Table 6: The relationship between political connections, intellectual capital and innovation

ICt Constant Control variables Industry and year fixed effects Observations Adjusted R2 F-statistic

Applyt+1 ApplyGrantt+1 1.9930** 1.7682 (2.1114) (1.9586) 4.4496*** 4.0212*** (13.2362) (12.5066) -1.3670*** -3.0998*** -2.8343*** (-13.6917) (-7.8013) (-7.4580) Yes Yes Yes Yes

Yes

Yes

5141 0.1893 14.19

5141 0.5131 59.89

5141 0.49 54.69

IApplyt+1 U Applyt+1 DApplyt+1 1.8570** 1.4294 0.8981 (2.2173) (1.7147) (1.5002) 4.5194*** 3.3412*** 2.0739*** (15.1522) (11.2543) (9.7277) -3.1511*** -2.3547*** -1.6681*** (-8.9382) (-6.7103) (-6.6195) Yes Yes Yes

CR IP T

ICt 0.2777** (2.3923)

P Ct

Yes

Yes

Yes

5141 0.4505 46.81

5141 0.4709 50.73

5141 0.2393 18.57

AN US

Notes: The t-values are reported in the parentheses. *, ** and *** indicate statistical significance at the 10%, 5% and 1% levels, respectively.

5. Conclusion

M

This paper identifies the positive and significant impact of social-network based political connections on firm innovation. In the framework of social network analysis, we employ three network

ED

centrality measures to calculate the strength of firms’ political connections and find that political connections benefit innovation. This positive relationship between political connections and

PT

innovation is further explained by two possible channels: government subsidies and intellectual capital. We provide support that political connections help firms get more government subsidies

CE

and intellectual capital, which both contribute to innovation.

AC

References

Agrawal, A. and Knoeber, C. R. (2001). Do some outside directors play a political role? Journal of Law and Economics, 44(1):179–98.

Al-Hadi, A., Taylor, G., and Al-Yahyaee, K. (2016). Ruling family political connections and risk reporting: Evidence from the gcc. The International Journal of Accounting, 51(4):504–524.

9

ACCEPTED MANUSCRIPT

Brehm, J. and Rahn, W. (1997). Individual-level evidence for the causes and consequences of social capital. American Journal of Political Science, 41(3):999–1023. Chuluun, T., Prevost, A., and Upadhyay, A. (2017). Firm network structure and innovation. Journal of Corporate Finance, 44:193 – 214.

CR IP T

Drucker, P. F. (1994). Post-capitalist society. Routledge. Faccio, M. (2006). Politically connected firms. American Economic Review, 96(1):369–386. Fan, J. P., Wong, T., and Zhang, T. (2007). Politically connected ceos, corporate governance, and post-ipo performance of china’s newly partially privatized firms. Journal of Financial Economics,

AN US

84(2):330 – 357.

FANG, V. W., TIAN, X., and TICE, S. (2014). Does stock liquidity enhance or impede firm innovation? The Journal of Finance, 69(5):2085–2125.

M

Francis, B. B., Hasan, I., and Sun, X. (2009). Political connections and the process of going public: Evidence from china. Journal of International Money and Finance, 28(4):696 – 719.

ED

Fu, R. and Gupta-Mukherjee, S. (2014). Geography, informal information flows and mutual fund

PT

portfolios. Financial Management, 43(1):181–214. Gogan, L.-M. and Draghici, A. (2013). A model to evaluate the intellectual capital. Procedia

CE

Technology, 9:867 – 875.

Goldman, E., Rocholl, J., and So, J. (2009). Do politically connected boards affect firm value?

AC

The Review of Financial Studies, 22(6):2331–2360.

Hall, B. H., Jaffe, A., and Trajtenberg, M. (2005). Market value and patent citations. The RAND Journal of Economics, 36(1):16–38.

Hall, P. A. (1999). Social capital in britain. British Journal of Political Science, 29(3):417–461. Hong, H., Kubik, J. D., and Stein, J. C. (2004). Social interaction and stock-market participation. The Journal of Finance, 59(1):137–163. 10

ACCEPTED MANUSCRIPT

Hou, Q., Hu, M., and Yuan, Y. (2017). Corporate innovation and political connections in chinese listed firms. Pacific-Basin Finance Journal, 46:158 – 176. HOUSTON, J. F., JIANG, L., LIN, C., and MA, Y. (2014). Political connections and the cost of bank loans. Journal of Accounting Research, 52(1):193–243.

CR IP T

Khwaja, A. I. and Mian, A. (2005). Do lenders favor politically connected firms? rent provision in an emerging financial market*. The Quarterly Journal of Economics, 120(4):1371–1411. Kumlin, S. and Rothstein, B. (2005). Making and breaking social capital: The impact of welfarestate institutions. Comparative Political Studies, 38(4):339–365.

AN US

Lerner, J., Sorensen, M., and Strömberg, P. (2011). Private equity and long-run investment: The case of innovation. The Journal of Finance, 66(2):445–477.

Leuz, C. and Oberholzer-Gee, F. (2006). Political relationships, global financing, and corporate

M

transparency: Evidence from indonesia. Journal of Financial Economics, 81(2):411 – 439. Li, H., Meng, L., Wang, Q., and Zhou, L.-A. (2008). Political connections, financing and firm

ED

performance: Evidence from chinese private firms. Journal of Development Economics, 87(2):283 – 299.

PT

Li, S., Song, X., and Wu, H. (2015). Political connection, ownership structure, and corporate

411.

CE

philanthropy in china: A strategic-political perspective. Journal of Business Ethics, 129(2):399–

AC

Ling, L., Zhou, X., Liang, Q., Song, P., and Zeng, H. (2016). Political connections, overinvestments and firm performance: Evidence from chinese listed real estate firms. Finance Research Letters, 18:328 – 333.

Massa, M. and Simonov, A. (2004). History versus geography : The role of college interaction in portfolio choice and stock market prices. CEPR Discussion Paper.

11

ACCEPTED MANUSCRIPT

Nahapiet, J. and Ghoshal, S. (2000). Chapter 6 - social capital, intellectual capital, and the organizational advantage*. In Lesser, E. L., editor, Knowledge and Social Capital, pages 119 – 157. Butterworth-Heinemann. Niessen, A. and Ruenzi, S. (2010). Political connectedness and firm performance: Evidence from

CR IP T

germany. German Economic Review, 11(4):441–464. Porter, M. E. (1992). Capital disadvantage: America’s failing capital investment system. Harvard Business Review.

Rothstein, B. and Stolle, D. (2002). How political institutions create and destroy social capital:

AN US

An institutional theory of generalized trust.

Tian, X. and Wang, T. Y. (2014). Tolerance for failure and corporate innovation. The Review of Financial Studies, 27(1):211–255.

Wu, J. and Cheng, M. L. (2011). The impact of managerial political connections and quality

M

on government subsidies: Evidence from chinese listed firms. Chinese Management Studies,

ED

5(2):207–226.

Wu, W., Wu, C., Zhou, C., and Wu, J. (2012). Political connections, tax benefits and firm

AC

CE

PT

performance: Evidence from china. Journal of Accounting and Public Policy, 31(3):277 – 300.

12