The value of active politicians on supervisory boards: evidence from the Berlin stock exchange and the parliament in interwar Germany

ABSTRACT

We provide the first overview over all political connections of firms via current Members of Parliament on supervisory boards and board of directors listed on the Berlin stock exchange in the 1920s. In contrast to anecdotal evidence, which suggest that political connections were expected to have a positive effect on firms’ performance, an event study based on the election in December 1924 and May 1928 shows only little evidence that a political connection via a newly elected or re-elected politician generated value. These results complement previous research emphasising that political connections might have mattered less in democracies.

KEYWORDS:

• Political connections
• interwar Germany
• roll-call votes
• stock market performance

JEL CODES:

• G14
• N14
• N24
• N84

Acknowledgements

We wish to thank the Editor and two anonymous referees for comments that substantially improved the paper. Furthermore, we are grateful for the comments of Carsten Burhop, Andreas Neumayer, Jochen Streb and Fabian Wahl. The usual disclaimer applies.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Further research in this area has looked at the existence and impact of bankers on supervisory boards as well as at the connections between different firms via supervisory board members (see for instance: Fohlin, 2007; Krenn & Marx, 2012; Windolf, 2006; Windolf & Beyer, 1997), but not on the value added by having politicians on the boards. Furthermore, all of the above focused on samples of firms. We study all firms listed on the Berlin stock exchange.

2 ‘Look at that, they are not sitting at the square of kings to take care of the needs of all German people. They just care about Breweries, the guilds of plumbers, great land owners or metal workers?’ The Königsplatz (today’s Platz der Republik) is the square were the parliament was located. Furthermore, he mentions that member bulletins of economic associations proudly reported the number of politicians that were members of its organisation or board members of a firm.

3 Klein (2014) has recently collected famous cases of corruption in Interwar Germany.

4 See also: Verhandlungen des Preussischen Landtags, 2. WP. Bd. 5, 95 Sitzung, 11. November 1925, 6267, 6326.

5 When the central bank intervened in order to prevent a bubble in 1927, the result was a medium-severe stock market crash. Net investment and GDP were likewise affected (see Voth, 2003).

6 Nominal GDP was taken from Burhop and Wolff (2005).

7 Rajan and Zingales (2003) estimate cover all stock exchanges in Germany and not just the central one in Berlin. However, Burhop and Lehmann-Hasemeyer (2016) have recently shown for that 1913 even if all stock exchanges are covered, the overall market capitalisation is slightly lower. A similar revision of the Rajan and Zingales (2003) data has been made for the US and the UK (see Musacchio & Turner, 2013, p. 528).

8 We used the sectors provided by Handbuch der deutschen Aktiengesellschaften and aggregated these categories into ten broader categories. Please see online appendix Table A1 for further information.

9 More precisely, the data are taken from Handbuch der Deutschen Aktiengesellschaften (1925, 1929), each referring to the end of the preceding year.

10 We are grateful to Carsten Burhop for providing us with an unpublished dataset of all firms listed on the Berlin stock exchange in December 1924, including name of the firm, size of the issue year of incorporation, location of headquarters and sector. This significantly reduced the amount of data that we had to collect.

11 This is quite a common problem as Radandt (1966) points out. Furthermore, to avoid such mistakes we checked the bibliographic information carefully.

12 Since 1884, all persons, regardless of whether they held shares or not, could be elected (see Burhop, 2006, p. 14).

13 Altogether, the 65 MPs held 269 board memberships in 1924. In 1928, 72 MPs held 276 board memberships (See Die Weltbühne, 1925, pp. 29–31; Die Weltbühne, 1928, pp. 256–258). Please note that Morus also reports the board memberships of firm that were not listed in Berlin.

14 However, since we only list the firms that were already listed in 1924, it is possible that the overall share remained constant.

15 The DVP were principally liberal: in contrast to the DDP, the party was rather right-leaning (Treue, 1954, pp. 117–130).

16 The aims of the Social Democrats were defined in their Erfurt Programme of 1891, which contained several practical demands in favour of the working class, such as the improvement of labour conditions, the eight-hour working day and free health care (see Treue, 1954, p. 76).

17 See previous section, Die Weltbühne, 1925, pp. 29–31, and Klein (2014, pp. 229–228).

18 Adressbuch der Direktoren und Aufsichtsräte, Band I & II, 1925/1926. Die Weltbühne, 1925, pp. 29–31.

19 For similar approaches see, for instance, Ferguson and Voth (2008), Turner and Zhan (2012), Lehmann-Hasemeyer, Hauber, and Opitz (2014) or Grossman and Imai (2016). MacKinlay (1997) gives an overview of this method.

20 Previous research yields more predictions how stocks may respond to elections. An overview is given by Sattler (2013). In general, it is possible that electoral outcomes are at least partly anticipated in advance. In that case investors form clear expectations before the election takes place. However, anticipation is lowest for close elections, when either parties or blocs have roughly equal chances of winning. However, even if one side is more popular, elections are rarely perfectly predictable, and both sides retain a nontrivial chance of winning.

21 They only issued ‘vinkulierte Namensaktien’, that is, registered shares with restricted transferability (For details Gelman & Burhop, 2008, p. 4).

22 We collected prices for 100 firms and 11 weeks per event (200 * 11). Prices were taken from the Berliner Börsenzeitung. Prices are listed as a percentage of a security’s nominal value. The data are available at the Staatsbibliothek zu Berlin or online (http://zefys.staatsbibliothek-berlin.de/list/title/zdb/2436020X/?no_cache=1).

23 Formally, the index is built according to the following equation: index_i = connection_i+ CEO_i + minister_i + mandates_i+ years as politician_i, where connection_i is equal to one, if the firm had some sort of a political connection. CEO_i is equal to one if the CEO was a current of former MP. minister_i is equal to 1 if the MP was a former minister. mandates_i covers the percentage share of the number of board memberships divided by the highest number a politician had in this year. Years as politician_i, is calculated as the number of years the board member was a member of the parliament divided by the largest number of years observed in this year. Thus the latter two values also rank between 0 and 1 and the index ranks between zero and five. If a firm had more than one politician on the supervisory board, we use the largest index.

24 For a more detailed overview of the advantages of the method, see King, Nielsen, Coberley, Pope, and Wells (2011).

25 In our example, we manually set breakpoints when there is plausible reasoning. Distance to Berlin is split into Berlin-based (<10 km) and based in another city. Firm age is classified into start-up (<5 years), founded after 1900 (<25 years) and before 1900 (>25 years).

26 This outcome is consistent with Deloof and Vermoesen (2016), who find no positive effect of board size in times of crisis.

27 See Ferguson and Voth (2008). The examination of political connections in Imperial Germany must be left to future research. Including this period would shed light on the effectiveness of connections under various regime types.

Additional information

Funding

This work was supported by Deutsche Forschungsgesellschaft as part of the Priority Programme 1859 (Grant number LE 280/4-1).