Inter-Builder Ties, Clientalized Relationships and Organizational Failure in Clyde River Shipbuilding, 1711 -1990 Ingram & Lifschitz
Networks and Organizational Performance • Influence of a given social form depends on others • Horizontal and vertical ties make up the social structure of industries – Position in these networks is likely correlated – Overlap in the mechanisms of their influence, e. g. reputation and knowledge sharing – No analysis of performance considers them simultaneously (few enough consider them separately)
Our Argument: 1) Horizontal ties to competitors reduce failure; 2) Vertical ties to customers reduce failure; 3) Horizontal ties contribute to the establishment of vertical ties.
Clyde Shipbuilding Setting • “Shipbuilding capital of the world” • Technology leader, first steam, first iron, first professorship in Marine Engineering • Part of an industrial complex that made Glasgow the 2 nd city of the empire • Industry peaks in 1900, in crisis from the 1930 s onward
Social Ties in Clyde Shipbuilding • Three centuries of ties • Horizontal and vertical ties • Multiple measures of each form
Robert Napier’s ties 1833 -1963 1852 -1896 1886 -1970 1834 -1873 1870 -1886 1879 -1897 1850 -1858 -1870 1899 -1906 1861 -1931 1821 -1836 -1900 1848 -1861 1844 -1892 1844 -1876 1843 -1851
W Denny & Bros’ ties 1845 -1853 1838 -1865 1853 -1866 1833 -1838 1821 -1906 1844 -1849 1821 -1931 1818 -1833 1849 -1963 1868 -1950 1877 -1947 1871 -1954 1907 -1957 1877 -1951 1874 -1949 1894 -1951
Horizontal Ties • Inter-builder learning and knowledge transfer • Collusion • Access to extra-industry resources • Customer Sharing
Knowledge Transfer • Ties between competitors promote knowledge transfer through: – Opportunity, motivation, and capacity • E. g. – Steam/Iron technologies, from D. Napier, to R. Napier, to a host of tied others – William Denny’s administrative innovations
Customer Sharing • Capacityconstrained industry (as hotels) • More than a favor to a friend, a way of maintaining good relationships w/customers Likelihood of Customer Sharing Extent of Customer Sharing Formal Tie 0. 342 0. 215 Informal Tie 0. 570** 0. 253* Proximity to Industry Leaders 0. 711** 0. 222** Same Region 0. 420** 0. 011**
Sharing Knowledge and Sharing Customers H 1: Shipbuilders will have a lower failure rate to the extent they have more ties to other shipbuilders.
Access to extra-industry resources • Transcending the provincial bounds of the industry for: – Capital – Political Clout – Prestige in the wider social system
Cunard’s Founding • The extra ship and the increase in size raised Cunard’s costs and there was little evidence of support from English investors. Napier, with a group of Glasgow friends and business partners, was able to float the British and North American Royal Mail Steam Packet Coy. (which later became the Cunard Steam Ship Coy. Ltd. ) with Samuel Cunard as the main shareholder, Napier himself investing £ 6100 of the initial capital of £ 270, 000. Establishing this great shipping company was no easy task…. It was largely Napier’s reputation and his proven ability to provide reliable and powerful engines which persuaded his fellow Glaswegians to invest in Cunard’s high-risk venture (Osborne, 1991).
Industry Leaders • Napier, Denny, Beardmore, Scott • H 2: Shipbuilders will have a lower failure rate to the extent they are more closely connected to industry leaders.
Collusion • Some evidence this was attempted • Tough here: – Customers few, deals are large – Competition from other locations • But, collusion, or cohesion wrt workers is another matter…
Builders vs. Labour • “Red Clydeside” • Builders’ Association acts like a capitalists’ union • Espionage and other efforts to hold down sedition
Collusion/Cohesion • Cohesion in networks between competitors: – Shut customer’s structural holes – Provide trust necessary for the transfer of sensitive information H 3: Shipbuilders will have lower failure rates when their horizontal networks are more dense.
Clientalized Relationships • Repeat interactions between seller and buyer lead to trust and problem solving (Larson, 1992; Uzzi, 1996, 1997, 2001; Zaheer, Mc. Evily and Perrone, 1998) • William Denny: “[w]e are bound, if we render ourselves worthy of it, to occupy the position of professional advisers, instead of that of mere manufacturers”
Clientalization E. G. , Cunard, John Brown & Co. and the Lusitania H 4: Shipbuilders with more clientalized relationships to customers will have lower failure rates.
Over-clientalization • At some point, clientalized relationships may imply: – Overdependence (Uzzi, 1996) – Insufficient diversity (Regans & Zuckerman, 2001) H 5: At high levels, the clientalized relationship’s of shipbuilders to customers will come to increase failure rates.
Links between Horizontal and Vertical Ties • Ties between competitors may lead to ties directly to customers through: – Referrals – Builder Reputation H 6: Builders with more horizontal ties will have more clientalized relationships with customers.
Indirect Links between Horizontal Ties and Vertical Ties • It’s a complicated world…. • Organizations look to competitors to choose how to meet the market (White, 1981) • Organizations whose CEOs have ties to others in industry follow industry’s mean tendencies regarding strategy
Indirect Effect of Horizontal Ties on Vertical Ties H 7: Builders will tend to have patterns of clientalized relations that reflect those of their horizontal ties.
Data & Method • 299 Builders, 1711 to 1990 • 205 Fail • 76 moved or merged + 18 alive at end of period (right censored) • No left censoring • Piecewise Exponential models of failure
Measures of Horizontal Ties • H 1 tested with total horizontal ties, and the total Broken into its two components: • Informal Ties – Family ties – Founders had worked together • Formal Ties – Financial Association – Sub-contracting arrangements • H 2: Proximity to Napier, Denny, Beardmore, Scott • H 3: Density among ties (ties between ties)
Measures of Vertical Ties • Repeat Customers • Non-repeat customers • Herfindahl Index of Customer Concentration (first & second order) • % of sales to best customer • Current and past sales to Admiralty
Results
Results for Horizontal Ties (Table 2) • H 1 Supported: – Total Horizontal Ties Reduce Failure (20% per) – Formal ties reduce failure (11% per) – Informal ties reduce failure (31% per) • H 2 Supported – Reach to the dominant clique reduces failure (up to 74%) • H 3 Rejected: – Density among horizontal ties increases failure
Examining the Knowledge Sharing Angle (Table 3) • Own Experience, and Experience of Network Contacts • Four discount rates to allow for experiential decay • Own Experience Reduces Failure • Experience of Formal Ties Reduces Failure
Results for Vertical Ties (Table 5) • H 5 supported: – Repeat customers reduce failure – Customer concentration reduces failure • H 6 supported: – One-shot customers reduce failure – Reliance on best customer raises failure • Selling to Admiralty doesn’t influence failure
Production Volatility (Table 6) • Repeat Customers have a large stabilizing effect, one-shot customers a smaller one – Customer portfolio effect, with additional steady business from clientalized ties • Proximity to leaders and network density also reduce volatility
Influence of Horizontal on Vertical Ties (Table 7) • Direct effect is minor (weak support for H 6) – Proximity to Leaders leads to Repeat Customers – Formal Ties leads to Customer Concentration • Indirect effect is major (strong support for H 7) – Ties’ average affects focal builder’s level on each of Building for Admiralty, Repeat Customers, One-Shot Customers, Customer Concentration
Summary • Horizontal Ties Reduce Failure – Knowledge sharing through formal ties – Customer sharing and ? ? Through informal ties – Extra-industry resources through industry leaders • Vertical ties Reduce Failure – Clientalized relations are good – One-time customers are good – Overdependence is bad • Builders mirror patterns of vertical relations of their horizontal ties
Informal/Formal Distinction • Informal and family ties and the heart of an important industry • Tie types depend on firm history
Hierarchy in the Industry • Why doesn’t cohesion reduce failure? – Difficulty of collusion • Industry is cliquey – Correlation of horiz. ties and network density – Clustering coefficient (Watts, 1999) – Visual Inspection of graphs
Figure 1 Horizontal Network of Clyde Shipbuilders, 1900 Scotts Denny Napier Scotts Napier Denny
Industry Leaders as “Good Fathers” • Leaders at the centers of the cliques • Direct effect to reduce failure • Customer sharing and smooth productin within their shadow • Horizontal analog to Keiretsu?
That was then, this is now… 1863 1990 Size of the largest cluster 20 2 Avg. Horizontal Ties 1. 3 0. 33 Avg. Repeat Customers 1. 48 0. 12
Next Steps • Does a given tie mean the same thing at across history? – Networks and institutions as compliments and substitutes • Effect of Provincial Status in a Declining Empire • Network Externalities: Implications at the Industry Level • Active comparison of ecological and network arguments – Analysis of Founding
Figure 1 Influence of Horizontal Ties Over Time
Скачать презентацию Inter-Builder Ties Clientalized Relationships and Organizational Failure in
69943df978a8e2c5c783939d6cd02746.ppt