Content uploaded by Robin Pearson
Author content
All content in this area was uploaded by Robin Pearson on Mar 22, 2016
Content may be subject to copyright.
Business History Review 82 (Spring 2008): 59–86. © 2008 by The President
and Fellows of Harvard College.
T
Robin Pearson and Mikael Lönnborg
Regulatory Regimes and
Multinational Insurers before 1914
At the end of the twentieth century, the global diffusion of one
important fi nancial service, insurance, was encouraged by
deregulation, but it also encountered diffi culties where de-
regulation remained incomplete and where there were many
nonregulatory barriers to entry. International insurance was
already well developed before 1914. The growth in the global
insurance trade, however, occurred against a background of
increasing national regulation and fi scal burdens in many
countries, making international business affordable only for
the largest companies with the deepest reserves. This paper
offers some preliminary estimates of the extent of the interna-
tional insurance trade during the half-century before the First
World War, and assesses the impact of national regulatory re-
gimes and nonregulatory factors on the development of this
business. The analysis is placed within the framework of mod-
ern theories of regulation and multinational enterprise.
ROBIN PEARSON is professor of economic history in the Department of History at the
University of Hull in the United Kingdom.
MIKAEL LÖNNBORG is assistant professor in the Department of Social Sciences at the
University College of Södertörn and the Department of Economic History in the Uppsala
Centre for Business History, University of Uppsala, Sweden.
Earlier versions of this paper were presented at the Business History Conference in Low-
ell, at seminars at the University of Southern Mississippi and Queen’s University Belfast, and
at the Association of Business Historians Conference, London. Our thanks go to the partici-
pants for their comments, as well to three anonymous referees. Robin Pearson also wishes to
thank the Nuffi eld Foundation (award number SOC/181/1876) and the U.K. Economic and
Social Research Council (award number R000234859) for their support of some of the re-
search presented here, and also Lucy Ponting for her invaluable research assistance.
1 Worldwide insurance premiums amounted to US$3.4 trillion across 145 countries in
2005. Swiss Re, “World Insurance in 2005,” Sigma, no. 5 (2006).
2 Studies of international insurance have been hitherto largely confi ned to histories of in-
dividual companies or national markets. Exceptions include Mira Wilkins, The Emergence of
he modern insurance industry is a global multi-trillion dollar busi-
ness.1 Surprisingly, however, comparative international insurance
history remains in its infancy.2 We seek to advance this history by
Robin Pearson and Mikael Lönnborg / 60
e xamining the impact of regulatory regimes on the international diffu-
sion of property insurance before 1914. Our focus is principally on fi re and
marine insurance, as other forms of nonlife insurance, such as auto-
mobile and liability, remained underdeveloped in this period. Because the
historical trajectory of life insurance, and some of its regulatory history,
differs from that of property insurance, it deserves a separate study.3
The paper proceeds in three parts. First, we present an outline of
the growth of international property insurance during the half-century
before the First World War, using new data on several dozen national
and colonial insurance markets compiled from a few existing studies
and an array of contemporary sources, notably trade journals and offi -
cial publications. The numbers are fragmentary, refl ecting the paucity
of long-run time series for insurance before 1914 and the unevenness of
fi nancial reporting requirements in this period. Nonetheless, the data
cover North America, most of the important European markets, and
Japan, together representing a wide variety of political regimes and reg-
ulatory systems. Second, we identify multinational forms of property
insurance in this period and discuss fi rms’ motives for adopting differ-
ent organizational methods when entering foreign markets. Third, we
survey the regulatory regimes for insurance that emerged in this period.
In passing, we draw on modern regulatory and multinational theory for
insights that help explain some of the patterns that emerge from our
analysis. In our conclusion, we point to possible links between multi-
national forms of insurance and regulatory regimes and assess the impact
of those links on the growth of international insurance before 1914.
Cross-Border Trade in Insurance before 1914
At this stage of research, we cannot reconstruct precisely the global
diffusion of insurance, but we do have some idea of the major phases and
Multinational Enterprise: American Business Abroad from the Colonial Era to 1914 (Cam-
bridge, Mass., 1970); Robin Pearson, “The Development of Reinsurance Markets in Europe
during the Nineteenth Century,” Journal of European Economic History 24, no. 3 (1995):
557–71; Peter Borscheid, “Vertrauensgewinn und Vertrauensverlust: Das Auslandsgeschäft
der deutschen Versicherungswirtschaft 1870–1945,” Vierteljahrschrift für Sozial und Wirt-
schaftsgeschichte 88, no. 4 (2001): 311–45; Kai Umbach, “Die Position deutscher Transport
versicherer auf dem Weltmarkt 1880–1914,” Vierteljahrschrift für Sozial und Wirtschaftsge-
schichte 94, no. 4 (2006): 413–37. See also several of the contributions to Peter Borscheid
and Robin Pearson, eds., Internationalisation and Globalisation of the Insurance Industry
in the Nineteenth and Twentieth Centuries (Marburg, 2007).
3 Mira Wilkins has shown the way with her analysis of U.S. life-insurance companies
abroad before 1914, but there is much more to be done. Wilkins, Emergence of Multinational
Enterprise, 65. See her comments in “Multinational Enterprise in Insurance: An Historical
Overview,” in Internationalisation and Globalisation of the Insurance Industry, eds. Bor-
scheid and Pearson, 4–26.
Regulatory Regimes and Multinational Insurers / 61
patterns of growth. Marine insurance had been international in nature
since its origins in Mediterranean lands during the fourteenth century.
It expanded greatly with the growth of the Baltic, Asian, and Atlantic
trades between the sixteenth and the eighteenth centuries. By the mid-
nineteenth century, marine insurance was dominated by a few large
British insurance companies, by Lloyds of London, and by brokers op-
erating out of ports such as Hamburg and Bremen.4 Inland-freight and
river-transport insurance in continental Europe, however, rapidly be-
came the preserve of European companies. Swiss companies, for in-
stance, made inroads into Italian and Austrian transport insurance,
while their German counterparts were numerous in the Netherlands,
Austria, Italy, southeastern Europe, Turkey, and Scandinavia, although
large numbers did not always equate to a large aggregate market share
for foreign fi rms.5
By contrast, fi re insurance was fi rst sold overseas between 1780 and
1820 by a few British offi ces, led by the Phoenix of London. By 1815, the
Phoenix had made forty-two agency appointments in Europe, North
America, the West Indies, Buenos Aires, and the Cape. During the fol-
lowing decade, overseas insurance, which was generally very profi table,
accounted for half of the company’s premiums.6 A second phase of dif-
fusion commenced in the mid-1820s, when bilateral treaties began to
be struck between several German, French, Belgian, Russian, and Brit-
ish companies to reinsure property in the United Kingdom and Europe.7
From its early centers in Europe, North America, and the West Indies,
fi re insurance spread to British India, China, and the Malay Peninsula
during the 1830s and 1840s, largely through the agencies of British in-
surers, though local companies were also set up by colonial merchants
in Bombay and Canton.8 After 1850 it became truly global, diffusing to
Australia, Africa, Asia, Latin America, and the Middle East, and reach-
ing those parts of Europe hitherto largely untouched, such as Spain, the
Balkans, and Eastern Europe. This third phase of growth, commencing
in the early 1850s and ending with the 1873 fi nancial crisis in Europe,
was marked by the rise of powerful new Liverpool-based insurance ex-
porters, such as the Royal and the Liverpool & London, which used the
4 Robin Pearson, “Finance, Insurance,” in History of World Trade since 1450, 2 vols., ed.
John J. McCusker (Farmington Hills, Mich., 2006), vol.1: 290–91.
5 There were, for example, eleven German transport insurers in Sweden in 1893, fourteen
in Austria in 1913, seventeen in Italy in 1912, nineteen in Denmark, and ten in Finland in
1903. In all these countries, native offi ces accounted for two-thirds or more of premiums by
the decade before the First World War. Umbach, “Deutscher Transportversicherer.”
6 Clive Trebilcock, Phoenix Assurance, 2 vols. (Cambridge, U.K., 1985, 1998), vol.1:
248–67.
7 Pearson, “Development of Reinsurance Markets.”
8 P. G. M. Dickson, The Sun Insurance Offi ce, 1760–1960 (Oxford, 1960), 192.
Robin Pearson and Mikael Lönnborg / 62
mercantile connections of their founders to extend underwriting to the
Pacifi c coast of North America and to India and the Far East. Insurers
quickly followed Commodore Perry’s gunships into Japan. There were
thirty-seven foreign companies operating there by 1870, including Brit-
ish, Chinese, and Dutch fi rms, only nine years after the fi rst had arrived
and nine years before Japan’s fi rst native insurance offi ce was estab-
lished.9 The postbellum period in the United States also witnessed in-
creasing numbers of British, German, and French fi re insurers, espe-
cially after the failure of American companies in the wake of the giant
confl agrations in Chicago (1871) and Boston (1872). In the older centers
of underwriting in Europe, fi re insurance became fi ercely competitive,
aided in part by a boom in the foundation of specialist reinsurance com-
panies. By 1863, for instance, there were seventy-fi ve fi re-insurance
agents operating in Hamburg and fi fty in Copenhagen.10
In countries where they were not excluded by law, the numbers of
foreign insurers generally rose during the late nineteenth century, but
entry to markets continued to be cyclical. There were at least three fur-
ther peaks before the First World War: in the early 1880s, the late
1890s, and from 1910 to 1913. Table 1 reconstructs the numbers of for-
eign insurance companies entering four countries in each decade, com-
pared with the total number of stock companies operating there.11
Two of these countries were net importers with few native offi ces
(Spain and Norway); one was a net importer with a large number of na-
tive offi ces, almost none of which exported (the United States); and one
was a net exporter with some powerful domestic fi rms that harbored
overseas ambitions (Sweden). The numbers entering Sweden, Norway,
and the United States rose during the 1850s, and they grew in Spain
during the 1870s. In the Scandinavian countries, foreign entry peaked
during the 1860s, while in Spain and the United States there was little
or no entry during the periods of civil war in the 1860s. There were fur-
ther peaks of entrants to the United States in 1872, 1876, and 1881, each
higher than the one before. The upswing of the early 1880s was also
experienced by Sweden and Norway. There were thirty foreign fi re in-
9 The Sumitomo Marine & Fire Insurance Co., The First Century, 1893–1993 (privately
printed, 1993), 12.
10 Trebilcock, Phoenix Assurance, vol.1: 315–16.
11 Mutuals are excluded from this table in order to compare like with like. Almost all for-
eign entrants were stock companies. In some countries, the number of tiny mutual offi ces
was very large. There were 1,012 “parish” mutual fi re-insurance associations in Sweden in
1910, for instance.
12 Spectator (May 1871): 417; Insurance Times (hereafter IT) (Oct. 1883): 608; Rund-
schau der Versicherungen (hereafter RdV) 13 (1863): 36–38; Vereinsblatt für Deutsches
Versicherungswesen (hereafter VDV) 2 (1874); Bengt Bergander, Försäkringsväsendet i
Sverige, 1814–1914 (Lund, 1967).
Regulatory Regimes and Multinational Insurers / 63
surers in Sweden in 1862, but seventy-nine by 1884.12 In Spain, by con-
trast, there were still only nine foreign offi ces in 1880, but by 1898, the
number had reached twenty-four.13
Tables 2 and 3 present the foreign share of insurance premiums in
these four countries, plus eight others, between the 1870s and the First
World War.14 Both tables display data for twenty-one different markets
in twelve countries, representing fi re, marine, nonlife, or all branches of
insurance, as the sources allow. Table 2 divides these markets based on
Table 1
Entry by Foreign Insurance Companies into
Four National Marketsa
Spain Sweden Norway U.S.A.
Year ForeignbTotalcForeign Total Foreign Total Foreign Total
Before 1830 — 2 2 — 1 — 3 —
1830–39 2 5 — — 6 — 1
1840–49 1 12 4 4 3 — 94
1850–59 2 30 15 16 11 — 8 253
1860–69 2 24 26 28 17 — 7 450
1870–79 6 32 13 15 13 — 39 759
1880-89 2 35 17 22 18 — 38 817
1890–99 14 59 21 35 11 19 — —
1900–09 20 102 23 34 — — — —
1910–15 24 184 26 37 — — — —
Sources: Rosales and Quiza, “Los Seguros en España, tables 1, 2; Bergander, För-
säkringsväsendet, 125–27; Hägg, Institutional Analysis, table 7; Insurance Inspector-
ate, Private Insurance in Sweden, How it has Developed and How It Is being Super-
vised (Stockholm, 1954), 71–73; Færden, Forsikringsvesenets, 323; U.S. Bureau of the
Census, Insurance Report (Washington, D.C., 1890), table 1, 74–77, supplemented by
data from Cornelius Walford, Insurance Cyclopedia, vols. 1–5 (London, 1871–78);
Donald Armstrong, “A History of the Property Insurance Business in the United States
prior to 1890” (Ph.D. diss., New York University, 1971), tables 7 and 8, pp. 72–74.
a The data for Spain cover all branches; those for Sweden, Norway, and the U.S. cover
property (fi re, marine, and inland) insurance only.
b Foreign = cumulative gross numbers entering in each period. They are not net of
withdrawals.
c Total = total number of foreign and domestic stock companies at end of the given
period.
13 Esperanza Frax Rosales and Jesús Matilla Quiza, “Los Seguros en España: 1830–1934,”
Revista de Historia Económica 14, no. 1 (1996): 183–203.
14 The data on marine insurance must be treated with particular caution. In the age of te-
legraphy, marine insurance appears to have been highly price elastic, and risks could be rela-
tively easily insured in London by shipowners and merchants where local agents could not
offer a competitive premium. Such insurance policies would not appear in national statistics.
Umbach, “Deutscher Transportversicherer,” 429.
Robin Pearson and Mikael Lönnborg / 64
Table 2
Foreign Company Share of National Insurance Marketsa
Low Import Penetration (<30%)
Austria-Hungary (Fire) 1%– 4% 1868–1912
Germany (Fire) 10% 1903–09; (All Branches) 14% 1902–13
Japan (Fire) 14% in 1905, 27% in 1910; (Marine) 16% in 1907, 3% in 1910, 7% in 1913
Norway (Marine) 9% in 1868–69, 11% in 1875–76, 18% in 1914
Sweden (Fire) 3%– 4% in 1887–1914; (Marine) 8%, 1890–05, 11% in 1900, 5% in 1910
Switzerland (Fire) 28%, 1886–87, 14% –15% 1891–97
High Import Penetration (>30%)
Canada (Fire) 66% in 1878, 69% in 1881, 76% in 1882; (Nonlife) 51% in 1904
Finland (Fire) 42% 1895–98; (Transport) 54% in 1895
Italy (Marine) 40%–51% 1887–97, 33% in 1912
Netherlands (Fire) 50% in 1910
Norway (Nonlife) 23% 1868–69, 31% 1875–76, 35% 1881–86, 38% in 1891, 41% in 1896,
45% in 1901–06
Spain (Fire) 37% in 1910; (All Branches) 40% 1909–14, 63% 1914–15
Switzerland (All Branches) 43% in 1886, 37% in 1904, 33% in 1912
Sources: AJ 1–35 (1880–1914); W. Rohrbach, ed., Versicherungsgeschichte ˝
Osterreichs
(Vienna, 1988), 423; IT (1883): 285, 608; U.S. Bureau of Manufactures, Insurance in For-
eign Countries; Gesellschaft fuer Feuer-versicherungsgeschichtliche Forschung e.v., Das
Deutsche Feuerversicherungswesen (Hanover, 1913), vol. 2, table. 13, p. 590; T. C. Kölmel,
Das Auslandsgeschäft deutscher Versicherungsunternehmen in den USA (Frankfurt/
Main, 2000), 132–33; Umbach, “Deutscher Transportversicherer,” 41, 429, 433n127; B. P.
A. Gales and K. Sluyterman, “Outward Bound: The Rise of Dutch Multinationals,” in The
Rise of Multinationals in Continental Europe, eds. Geoffrey Jones and Harm G. Schröter
(Aldershot, 1993), 65–98; Færden, Forsikringsvesenets, 324–25; A. J. Cook, “Insurance in
Spain,” PM, 25 Oct. 1913; Rosales and Quiza, “Los Seguros en España,” 183–203; Hägg, In-
stitutional Analysis, tables 23, 24; K. Englund, Försäkrings och fusioner. Skandia, Skåne,
Svea, Thule, Öresund, 1855–1980 (Stockholm, 1982), 43–44; Statistical Yearbook for Swe-
den (Stockholm, 1914); Borscheid, “Vertrauensgewinn und Vertrauensverlust,” fi gs. 5, 6.
a Annual averages are given where data cover more than one year. Figures are percentages
of premiums earned, gross of reinsurance, except Sweden (1887, 1903, 1913 only) and J apan
(marine), which are percentages of sum insured. Swedish fi gures for 1890, 1895, 1900,
1905, and 1910 are net premiums for Swedish fi rms, gross premiums for foreign fi rms.
their records of low and high import penetration; the dividing line for
foreign market share is set at 30 percent.15
15 Only three of the twenty-one markets moved across this threshold, namely, nonlife in-
surance in Norway, where the premium share of foreign companies increased from 23 per-
cent in 1868 to 31 percent in 1875; and fi re and marine insurance in the United States, as
shown in Table 3, where foreign market share rose above the 30 percent mark between 1872
and 1915, sometimes more than once.
Regulatory Regimes and Multinational Insurers / 65
From the 1870s and 1880s in Austria-Hungary, Switzerland, Swe-
den, in Norwegian marine insurance, and in Germany and Japan dur-
ing the 1900s, foreign insurers achieved only a modest presence. In
some places, they were virtually absent. There were no foreign compa-
nies at all in Russia before 1868, and hardly any in Austria before the
liberal legislation of 1873. An imperial decree of 1871 allowed foreign
direct fi re insurers into Russia and Russian Poland, but most chose to
continue to underwrite reinsurance only.16 By contrast, in the Nether-
lands, Canada, Finland, Spain, and in the markets for Norwegian fi re in-
surance and Swiss transport insurance, levels of foreign penetration were
high. Developing economies lacking local fi rms or state monopolies
necessarily had to fi nd all their insurance with foreign fi rms, as was the
case in Romania before 1872, Japan before 1879, and Turkey and Egypt
before 1897. As late as 1905, Greece, much of Central America and the
West Indies, and smaller Asian countries, such as Siam, had no native
insurance companies at all.17
Table 3
Foreign Company Share of U.S. Insurance Markets
Marine and Inland Fire and Marine
a
Year $m
insured % of $
insured
b $m
premiums % of
premiums
1872 44 25
1875 1,283 8 50 22
1880 1,907 12 46 35
1885 1,833 18 — —
1890 2,637 44 — —
1891 — — 149 27
1895 2,780 58 161 29
1900 5,226 63 182 31
1905 5,690 45 273 24
1910 12,596 59 352 24
1915 21,527 51 434 29
Sources: Calculated from Spectator, 21 Aug. 1919; ICM 11 (1881): 118; The Insurance
Yearbook (New York, 1896–1916); IT (1893): 184.
a The fi rst three rows relate to fi re insurance only.
b Stock and mutual companies, plus Lloyds from 1903.
16 Trebilcock, Phoenix Assurance, vol.2: 142. According to the U.S. consul in Moscow in
1905, foreign insurance companies could operate under a charter, but few did. A few more
operated illegally. U.S. Bureau of Manufactures, Insurance in Foreign Countries, Special
Consular Report 38 (Washington, D.C., 1905).
17 Ibid.
Robin Pearson and Mikael Lönnborg / 66
Increasing foreign entry did not always lead to a decline in market
share for native fi rms. Countries with powerful local companies, such
as Sweden, could accommodate growing numbers of foreign insurers
without any concomitant loss of business, as a comparison of Tables 1
and 2 demonstrates. For the same reason, foreign market share fell in
Switzerland during the three decades before the First World War. For-
eign insurers also made only limited inroads into markets with a strong
tradition of mutual insurance, such as Germany and France, particu-
larly in specialist branches, such as hail or livestock insurance.
Table 3 suggests that the foreign share of the world’s largest prop-
erty insurance market, the United States, rose during the 1870s and the
late 1890s, both decades when total premium volume was growing only
slowly or shrinking. When the U.S. market grew quickly in the 1900s,
the share held by foreign insurers fell. As total premium volume grew
again between 1910 and 1915, however, foreign market share also rose
rapidly. Here too, therefore, import penetration was cyclical, but its re-
lation to the development of the total U.S. market remains unclear, as it
does for other countries.18 Overall, the growth in the amount insured by
foreign companies in the United States was impressive. Between 1899
and 1914, while U.S. GDP grew by 57 percent in real terms and long-
term foreign direct investment in the United States grew by 123 per-
cent, fi re and marine insurance premiums earned by foreign companies
grew even faster, from $51 million to $121 million, or by 137 percent.19
A detailed analysis of the cycles of international insurance growth
during the half-century before 1914 is beyond the scope of this paper,
but it is worth noting that the connection with foreign investment is not
obvious. Consider the fi gures for Britain, by far the world’s biggest ex-
porter of insurance. British companies were fl ooding into the United
States after the Chicago and Boston fi res just as Britain’s net foreign in-
vestment was falling, and the same was true of the insurers’ rush into
the United States during the mid-1890s, when there was an investment
boom back home. In the latter case, some British (and German) compa-
nies sought to take advantage of the weak position of many smaller
American fi rms following the recession of 1893, while they were also
enticed by rival companies’ reports of profi ts on U.S. business in earlier
periods.20 In other words, in each upswing of foreign entry, there may
have been period- or event-specifi c factors helping to explain the trend.
18 This question is explored for the United States in a forthcoming book by Robin Pearson,
Insuring America: Multinational Insurers in the United States, 1850–1920.
19 Angus Maddison, Dynamic Forces in Capitalist Development (Oxford, 1991), 210;
Wilkins, History of Foreign Investment in the United States, 159.
20 Robin Pearson, “British and European Insurance Enterprise in American Markets,
1850–1914,” Business and Economic History 26 (Winter 1997): 438–51.
Regulatory Regimes and Multinational Insurers / 67
Nevertheless, the general patterns of British commerce were refl ected
in the distribution of overseas insurance sold by major exporters, such
as the Phoenix, notably the relative decline of trade and investment in
Europe between 1865 and 1914, and the rise in the importance of the
United States, India, China, and the Dominions.21
One remarkable feature of nineteenth-century insurance was the
range of nations involved in exporting it. By 1870 companies from the
United Kingdom, Germany, France, the Netherlands, Belgium, Austria-
Hungary, China, Canada, and Switzerland sold fi re or marine insurance
in foreign markets. During the following two decades, insurers from
New Zealand, Sweden, Norway, Denmark, Ireland, Singapore, Japan,
Australia, and the United States joined this list. At present, it is diffi cult
to explain such diversity, given how little we know about the insurance
industries of many of the smaller insuring nations. Driving the interna-
tional diffusion of property insurance, however, were those companies
from Britain and Europe whose business was highly dependent upon
foreign income. The largest, the Liverpool, London & Globe, earned 86
percent of its total premiums abroad in 1906.22 However, companies in
other countries depended heavily on exports as well. For example, be-
tween the 1880s and the 1900s, up to 70 percent of the total premiums
earned by the major Swedish insurers came from abroad.23
In sum, although the data are fragmentary, they are suffi cient to
point to three conclusions about the global diffusion of property insur-
ance: fi rst, that it was cyclical in nature; second, that different markets
were riding on different cycles, so that diffusion was not even or simul-
taneous across countries; third, that levels of import penetration in dif-
ferent countries could fall as well as rise during this period. Thus, the
growth of international insurance was not a linear phenomenon. We
can also distinguish between the trajectories of marine and fi re insur-
ance. The former had always exported its services far and wide, even as
the chief locus of marine underwriting moved from the Mediterranean
ports to northwestern Europe between the fi fteenth and nineteenth cen-
turies. By contrast, the business of exporting fi re insurance commenced
in the late eighteenth century from one source—London. By the middle
of the nineteenth century, however, fi re-insurance exports too had be-
gun to fl ow from multiple sources, involving companies from over a
21 Michael Edelstein, “Foreign Investment and Accumulation, 1860–1914,” in The Eco-
nomic History of Britain since 1700, 2nd ed., vol.2, eds. Roderick Floud and Donald N. Mc-
Closkey (Cambridge, U.K., 1994), 173–96; Trebilcock, Phoenix Assurance, vol.1: table 5.6;
vol. 2: table 2.3.
22 Trebilcock, Phoenix Assurance, vols. 1 and 2; and Pearson, “Reinsurance Markets.”
23 Mikael Lönnborg, Internationalisering av svenska försäkringsbolag: Drivkrafter,
organisering och utveckling, 1855–1913, Uppsala Studies in Economic History no. 46
(U ppsala, 1999), 158.
Robin Pearson and Mikael Lönnborg / 68
dozen different countries. In the following section, we examine the in-
stitutional vehicles adopted by insurance companies when exporting
across borders and discuss some of the motives behind their choices.
Institutional Forms of Multinational Enterprise
in Insurance
Like multinationals in other sectors, insurance companies adopted
several different institutional means to enter foreign markets. (See Ta-
ble 4.) Even without considering the impact of regulatory barriers and
different market conditions, the range of institutional choice itself may,
at least in part, explain the nonlinear nature of the diffusion process.
Several forms of multinational activity allowed insurers to hedge their
bets before committing themselves to foreign markets, which made it
easier for them to withdraw when things went wrong.
The most basic form of export was underwriting abroad without es-
tablishing any formal presence there. British companies attracted many
proposals to insure foreign property, both from their fellow country-
men and from other nationals approaching them directly. Some oper-
ated a “home-foreign” department for such business, whose accounts
were kept separate from those of foreign agencies.24 Our impression,
however, from a detailed reading of the insurance press, is that most
exporters were moved at some point to employ agents abroad. In the
United States, these were often general agents, independent fi rms that
sold insurance and held powers of attorney on behalf of a number of in-
surers. American general agents usually confi ned operations to their
home states, but the most successful became huge regional concerns,
underwriting millions of dollars of property and employing hundreds of
staff and agents.25 They charged heavily for their services, but the best
also delivered excellent results.
Insurance companies sometimes preferred to combine the use of
agents with alternative entry options. An unknown quantity of insurance
in the United States, for example, was purchased and sold through
brokers, though companies frequently complained about the costs in-
volved. Another form of foreign entry was the joint venture between of-
fi ces, usually, but not always, of the same nationality. In Austria, for in-
stance, the Imperial of London and the North British of Edinburgh
shared business equally until 1875.26 In various countries membership
24 Trebilcock, Phoenix Assurance, vol. 1: 234–36.
25 One example was the agency of John C. Paige of Boston. By 1880, Paige had two Brit-
ish, two French, and four American fi re offi ces as clients, and had built up “one of the largest
and most thriving insurance agencies in the East.” IT 13 (Dec. 1880): 700.
26 RdV 27 (1877): 37–43.
Regulatory Regimes and Multinational Insurers / 69
in insurance tariff unions—associations of companies that regulated
premium rates and disseminated knowledge of risks and best under-
writing practice among their members—also afforded newcomers some
protection through lower information costs and uniform prices and
policy conditions. Many tariff organizations, such as those formed in
Australia in the 1890s, operated in “weakly collusive” markets and co-
existed with a generally manageable level of nontariff competition from
both foreign and local offi ces.27
In the 1880s, the largest British fi rms began to establish subsidiar-
ies in the United States that were usually registered as American com-
panies in New York. This strategy was also occasionally adopted in Eu-
rope.28 Native offi ces were also purchased, for instance in the United
States by British, and occasionally German, insurers. There were several
factors behind these strategies. Establishing subsidiaries from scratch,
or purchasing an indigenous offi ce and then letting it continue to oper-
ate under its own name, were devices usually motivated, at least in part,
by fear of troublesome state legislation.29 A purchase could be a way of
Table 4
Institutional Forms of Multinational Enterprise in Insurance
1. Unmediated or mediated exporting of direct insurance with no overseas base
Via home–foreign business
Via nonresident general agents
2. Direct insurance via local agencies or branches
3. Direct insurance via brokers or general agents
4. Direct insurance via joint ventures, including coalitions
5. Acquisition of a local offi ce
6. Establishment of local subsidiaries
7. Reinsurance via reinsurance brokers
8. Reinsurance
Via treaties (a regular reinsurance of risks between companies)
Facultative (ad hoc or “one-off” reinsurance arrangements)
27 Monica Keneley, “The Origins of Formal Collusion in Australian Fire Insurance, 1870–
1920,” Australian Economic History Review 42, no. 1 (2002): 54–76. The same is true of
Britain beginning in the 1860s. Oliver M. Westall, “Marketing Strategy and the Competitive
Structure of British General Insurance, 1720–1980,” Business History 36, no. 2 (1994):
20–46.
28 For example, La Foncière of Paris established an offshoot, La Fondiaria, in Florence in
1878. Assecuranz Jahrbuch (hereafter AJ) 1 (1880): 269–70.
29 Dickson, Sun Insurance Offi ce, 226–30; Pearson, “British and European Insurance
Enterprise,” 438–51.
Robin Pearson and Mikael Lönnborg / 70
entering a new market or expanding in an existing one. Sometimes it
was a means of squeezing out local rivals. When the London & Lanca-
shire, for instance, bought the Capital City Fire (Alabama) in 1899, it
removed the last signifi cant native fi re offi ce in the state.30
Reinsurance offered another means of obtaining foreign business.
International fi re reinsurance emerged in Europe during the 1820s in
the form of bilateral quota agreements between British, German, French,
and Belgian companies to take each other’s risks.31 The fi rst, or “direct,”
insurance company, dealing directly with the policyholders, would in-
sure a range of properties in a given category in its homeland, say Lan-
cashire cotton warehouses. It would then turn over, “cede” or “reinsure,”
a fi xed proportion of the sums insured on these risks, say 10 percent,
according to the quota agreed, to the second insurance company, the
“reinsurer,” in return for a premium paid by the reinsurer to the direct
insurer. In this second transaction, the reinsurer had no direct contact
with the original policyholders. Under these early bilateral treaties, the
fl ow of risks also took place in the opposite direction, so that, in fact,
both parties were direct insurers and reinsurers for each other. In this
way, for example, a range of industrial properties in England was rein-
sured by the Imperial Fire Offi ce in the early 1840s with the Compagnie
Royal de Paris, while the French offi ce in return reinsured what proved
to be even more hazardous risks—dockside warehouses at Havre, sugar
refi neries and textile mills at Lille, Rouen, Lyons, and elsewhere—with
Imperial. The fi rst specialist reinsurers appeared fi rst in Germany, in
the 1840s, and then in Belgium, Switzerland, Austria-Hungary, and
Scandinavia during the 1850s and 1860s. Most of these companies were
the offspring of direct insurers faced with shortages of reinsurance ca-
pacity at home and keen to avoid ceding large portions of domestic
business to rivals. Some, such as the Swiss Re of Zurich, quickly con-
structed a large international business for themselves.32 State restric-
tions on direct underwriting by foreign companies, like those imposed
in Prussia and Russia, also encouraged the use of reinsurance. By 1871,
54 percent of the sums insured by native Russian stock companies were
reinsured, mostly with foreign insurers.33 Some of this growing volume
of reinsurance was sold by a new type of specialist international rein-
surance broker that emerged during the 1870s.34 In America, European
30 Post Magazine (hereafter PM), 6 June 1899.
31 Pearson, “Reinsurance Markets.”
32 Robin Pearson, “The Birth Pains of a Global Reinsurer: Swiss Re of Zurich, 1864–79,”
Financial History Review 8, no. 1 (2001): 27–47.
33 RdV 22 (1872): 276–77; RdV 26 (1876): 181–82.
34 There were twenty-fi ve foreign reinsurance bureaus in Paris alone by 1878. RdV 28
(1878): 51–52.
Regulatory Regimes and Multinational Insurers / 71
fi rms dominated the business. In the decades before the First World
War, several of them piggybacked on British direct insurers that had es-
tablished themselves in America much earlier. The Skandia of Sweden,
for instance, entered New York in 1900 as a reinsurer for the Royal of
Liverpool, while the Swiss Re went there in 1910 “under the chaperon-
age” of the Phoenix of London.35
Thus, not only were the institutional forms of multinational enter-
prise in insurance highly diverse, but also several forms were used si-
multaneously by one company as a means of spreading the risk of an
uncertain foreign venture. As noted above, collaboration between in-
surers from different countries when entering a foreign market was not
unusual. This desire to minimize risk was the common motive behind
all institutional choices made by multinational insurers.
For the same reason, some insurers conformed to the Nordic chain
model of multinational enterprise. This model postulates that inter-
nationalization results from the acquisition of market-specifi c knowl-
edge.36 Experience gained from operating in a particular market enables
a company to increase its commitment there. Multinational fi rms go
through a sequential expansion process or “establishment chain.” Re-
strained by “psychic distance” and uncertainty, fi rms at the beginning
of foreign ventures will limit investment and stay relatively close to their
domestic market.37 After gaining international experience, they will then
gradually invest further away, and on an increasing scale, from the do-
mestic market. The presence of cultural affi nities between markets may
help explain why the bulk of the early foreign business of Scandinavian
and Spanish fi rms was done in neighboring countries.38 In Norway, for
instance, two Swedish companies, the Skandia and the Svea, accounted
35 Skandia Corporate Archive, Styrelseprotokoll, 3 Apr. 1900; Swiss Re Corporate Archive,
Swiss Re., Protokolle der Verwaltungsratssitzungen, 14 July 1910.
36 Edith T. Penrose, The Theory of the Growth of the Firm (Oxford, 1959); Yair Aharoni,
The Foreign Investment Decision Process (Boston, 1966); Richard M. Cyert and James G.
March, A Behavioral Theory of the Firm (Englewood Cliffs, N.J., 1963); Stephen H. Hymer,
The International Operations of National Firms: A Study of Direct Foreign Investments
(Cambridge, Mass., 1976).
37 Jan Johanson and Jan-Erik Vahlne, “The Internationalization Process of the Firm: A
Model of Knowledge Development and Increasing Foreign Market Commitments,” Journal
of International Business Studies 8, no. 1 (1977): 23–32; Mats Forsgren, “Some Critical
Notes on Learning in the Uppsala Internationalization Process Model,” working paper, no.2
(Uppsala, 2002); Ingmar Björkman and Mats Forsgren, “Nordic International Business Re-
search: A Review of its Development,” International Studies of Management and Organiza-
tion 30, no. 1 (2000): 6–25.
38 Spain’s fi rst multinational insurer, La Union y El Fénix Español, founded in 1864, com-
menced direct underwriting abroad in Portugal in 1868. José Luis Garcia-Ruiz and Leonardo
Caruana, “The Internationalisation of the Business of Insurance in Spain, 1939–2005,” in In-
ternationalisation and Globalisation of the Insurance Industry, eds. Borscheid and Pearson,
66–83.
Robin Pearson and Mikael Lönnborg / 72
for over half of the premiums earned by foreign companies in 1868. The
political union between Sweden and Norway (1814–1905) obviously
played a role, but probably more important was the fact that Norwegian
investors held up to one-quarter of the share capital of these Swedish
companies.39 Even where insurers operated at great distances, national
connections mattered. In the 1890s, German insurers were emboldened
to venture across the Atlantic by the availability of reinsurance from
reputable German fi rms, such as the Munich Re. This made them “in-
dependent from foreigners and enabl[ed] them to master every risk,” as
one contemporary enthusiastically put it.40 Psychic distance also did
not equate to geographic distance where ethnic ties connected overseas
settlements of merchants with insurers in their home countries. The
fi rst fi re and marine insurance companies established by Chinese inves-
tors, for instance, initially targeted Chinese communities in Singapore,
the Philippines, and California for their overseas operations.41
Imitation could also induce entry. Brief spells of high profi ts for
those who had already crossed the Atlantic, especially in the mid-1870s
and mid-1880s, encouraged others. The British insurance press criti-
cized a “follow-my-leader” mentality among British offi ces in this re-
spect.42 Many companies, however, carefully scrutinized the trans-
atlantic performance of others before coming to a decision.43 Beginning
in the 1880s, German insurers were attracted to America by the experi-
ence of their own pioneers, such as the Magdeburg, as well as by the
search for higher premium yields in the face of falling profi ts on Euro-
pean business.44 Exogenous shocks also had an impact on levels of en-
try. After the great fi res in Chicago and Boston in 1871–72, British com-
panies’ reputation for streamlining claims adjustments secured them
an increased fl ow of business, and also encouraged newcomers to sign
up, although such effects were not always long lasting.45
Competitive conditions, including high expense ratios and moni-
toring problems, helped determine changes in the institutional vehicles
chosen by foreign insurers. Although an insurance company might ini-
39 Karl Færden, Forsikringsvesenets historie i Norge, 1814–1914 (Oslo, 1967), 324–25.
40 AJ 20 (1899): 158.
41 These were the Ren He and the Ji He companies, founded in 1876 and 1878, respec-
tively. Zheng Kang, “Assurance modernes en Chine: Une continuité interrompue (1801–
1949),” Risques 31 (July–Sept. 1997): 103–20.
42 PM, 20 Aug. 1881.
43 PM, 27 Mar. 1886, 3 Aug. 1889.
44 AJ 2 (1881): 367–68; AJ 18 (1897): 184; AJ 17 (1898): 145; AJ 20 (1899): 158; PM, 8
Apr. 1882. The average premium yield per $100 insured in the United States for seventeen
British and German fi re offi ces operating there in 1878 was over twice as high as the yield
from their business in the rest of the world. Calculated from IT (Aug. 1879): 573.
45 Mikael Lönnborg and Robin Pearson, “Multinational Insurers and Catastrophic Loss:
Responses to the San Francisco Earthquake, 1906,” unpublished paper.
Regulatory Regimes and Multinational Insurers / 73
tially employ several different institutional forms when entering a new
market, the subsequent rise or fall of a company’s confi dence in that mar-
ket could lead to rationalization. In the early 1880s, some British fi rms,
such as Commercial Union, dispensed with general agents in the United
States and began to do all their business directly through their own local
agents.46 Intractable monitoring problems, such as deliberate conceal-
ment of information about a risk by local agents in order to increase busi-
ness and their commissions, and anxiety about profi ts, however, could
lead insurers either to abandon a market entirely, as the Japanese marine
companies did in London in the late 1890s, or to replace direct under-
writing by reinsurance, as occurred widely in Austria during the 1870s.47
In sum, one can identify four principal phases in the organization
of insurance exports: fi rst, direct “home-foreign” underwriting of risks
abroad, commencing in Britain in the 1780s and undergoing a relative
decline from the 1850s on; second, direct underwriting through net-
works of commissioned agents, also dating from the 1780s; third, the
use of reinsurance and cooperative agreements to underwrite in foreign
markets indirectly, commencing in Europe in the 1820s; and fourth, di-
rect investment in establishing subsidiaries and purchasing native of-
fi ces, beginning in Europe and North America in the late 1870s.48 For
the most part, however, no linear sequence was involved. From the
1870s on, most of these organizational methods for exporting insurance
coexisted and were often used simultaneously by companies in order to
spread the risk of venturing into foreign markets.
Regulatory Regimes in Insurance
We have identifi ed nine categories of insurance regulation in over
twenty states in Europe, Asia, and the Americas during the nineteenth
century. (See Table 5.) The character of the regulatory regimes in differ-
ent countries was determined by the extent to which they regulated
some or all of these categories, the stringency and transparency with
which such regulations were applied, and the extent to which there were
multiple layers of subnational regulatory and fi scal authority. In the
following survey, our focus is on property insurance, particularly the
46 Insurance & Commercial Magazine (hereafter ICM) 14 (Jan.–June 1883): 334.
47 Nippon Life Insurance Company, The 100-Year History of Nippon Life: Its Growth
and Socio-Economic Setting, 1889–1989 (Osaka, 1991), 46, discussing the Nippon Sea and
Land Insurance Company; The Tokio Marine & Fire Insurance Company, The Tokio Marine
& Fire Insurance: The First Century, 1879–1979 (Tokyo, 1980), 36–45; Trebilcock, Phoenix
Assurance, 1: 317; 2: 141.
48 “Home-foreign” premiums declined from 28 percent of foreign premiums earned by
the Phoenix of London in 1846–50 to 8 percent in 1866–70. Trebilcock, Phoenix Assurance,
vol.1: 190, table 5.6.
Robin Pearson and Mikael Lönnborg / 74
regulation of foreign fi re-insurance and reinsurance companies oper-
ating in Europe and North America.49 Unless otherwise stated, how-
ever, the regulations and fi scal burdens examined below applied to all
insurance companies, both native and foreign.
Most countries required some form of license for insurance compa-
nies to do business in their territory. In the United States, domestic as
well as out-of-state and foreign companies were required to obtain state
licenses with various conditions attached, which we discuss below. In
Western Europe and Scandinavia, licensing requirements were minimal.
Here, governments viewed liberal legislation as a means of securing com-
petition and facilitating the lowest premiums for households and busi-
nesses. By contrast, Central and Eastern European governments com-
monly used licenses, often tied to onerous conditions, to restrict entry.
In many German states before unifi cation, licenses were diffi cult to ob-
tain. In the two decades following the Prussian insurance law of 1837,
only six non-Prussian companies were able to obtain full licenses for
the whole of Prussia. When company licensing was relaxed in Prussia
and other German states during the early 1860s, there was a surge of ap-
plications: one hundred sixty-four licenses were issued in Germany to
out-of-state German and foreign insurers between 1862 and 1865.50 Al-
though regulation became more liberal, German unifi cation did not help
Table 5
Typology of Insurance Regulation
1. Company licensing
2. Regulation of insurance contracts, including standard policies, property valuations,
valued policy laws, coinsurance clauses
3. Regulation of loss adjustment
4. Licensing/monitoring of agents
5. Taxation
6. Deposit requirements
7. Supervision and reporting requirements, including legislative investigations
8. Company business, including separate lines, reinsurance, restrictions on advertising
capital, proof of demand, capital requirements, insurance limits
9. Antitrust, including rate regulation, rating bureaus
49 In this paper, “foreign” only applies to insurance companies whose head offi ce is lo-
cated in another country. In federal polities, such as the United States and Germany, the
term was also often applied by state authorities to companies from other states operating in
their territory. Such companies are here referred to as “out of state.”
50 Calculated from Jahrbuch für das gesamte Versicherungswesen in Deutschland,
1864–66.
Regulatory Regimes and Multinational Insurers / 75
to rationalize the great diversity of state regulatory regimes. The licens-
ing laws of the regions annexed by Prussia after the war of 1866 retained
their validity, and within the German empire after 1871 state licenses
were still required for foreign insurance companies. This continuing
state particularism in Germany lends support to the behavioral model
of regulation, in which the amount of regulation is determined by the bal-
ance of power among different groups—politicians, civil servants, native
companies—pursuing their own vested interests.51 On occasion, licens-
ing wars could break out between states. For ex ample, when Bavaria re-
fused licenses to Baden-based insurance companies without heavy im-
posts, Baden retaliated by refusing permits to Bavarian insurers.52
Accompanying state licensing was a range of other regulations to
which underwriters and policyholders from all companies had to sub-
mit. In parts of Europe, and to a lesser extent in the United States, the
processes of underwriting and loss adjustment, and the insurance con-
tract itself, were the subject of regulators’ attention as they became con-
cerned by the increasing complexity of policy language and terms and
by the rise of litigation between claimants and companies. Contracts
came under the scrutiny of European licensing authorities at an early
date, although the extent to which standard policy terms were enforced
before 1914 is not clear.53 American regulators prescribed a standard
language for all policies written on certain lines of insurance, but not
before much preparatory work had been done by the industry itself. The
fi rst serious attempt to develop a standard policy in the United States
was made by the National Board of Fire Underwriters in 1868. Massa-
chusetts adopted a standard form in 1873, and several states followed
suit in the 1880s. In line with economist George Stigler’s theory of how
regulatory processes are “captured” by the subjects of regulation, it has
been argued that the standard policy was designed to favor the industry
itself.54 Wisconsin’s standard policy of 1895, for example, contained many
conditions, and the breach of any of them invalidated the policy. As late
as 1917, 28 percent of all fi re-insurance policies in the state were voided
for some breach of the conditions specifi ed in the standard form.55
51 Richard. A. Posner, “Theories of Economic Regulation,” Bell Journal of Economics and
Management Science 5, no. 2 (1974): 335–58; Sam Peltzman, “Toward a More General The-
ory of Regulation,” Journal of Law and Economics 19, no. 2 (1976): 211–40; Kenneth J.
Meier, The Political Economy of Regulation: The Case of Insurance (Albany, 1988).
52 Ludwig Arps, Auf sicheren Pfeilern: Deutsche Versicherungswirtschaft vor 1914 (Göt-
tingen, 1965), 37–48.
53 Police validation of fi re-insurance policies became a legal requirement in Hanover in
1828, Hesse in 1833, Prussia in 1837 (insurance on goods) and 1841 (buildings insurance),
and Wurttemberg in 1852. The nature and scope of this validation is not known.
54 George J. Stigler, “The Theory of Regulation,” Bell Journal of Economics and Manage-
ment Science 2, no. 1 (1971): 3–21.
55 Meier, Political Economy, 54.
Robin Pearson and Mikael Lönnborg / 76
Another type of contract regulation in the United States was the
valued policy law. This fi xed the value of the insured property, and in
the event of a total loss, that amount had to be paid out to the policy-
holder, regardless of the actual value of the property destroyed. The
fi rst such law was passed in Wisconsin in 1874, and by 1900 they were
in force in nineteen states and territories, mostly in the South and the
West.56 Insurance companies fought these laws through the state courts,
arguing that valued policies drove up loss ratios, increased moral haz-
ard, and encouraged overinsurance and fraudulent arson.57 On occa-
sion, they withdrew entirely from states imposing such laws.
In Central Europe, there was considerable intervention in the pro-
cess by which policies were issued. Several German states required policy-
holders to obtain from the police or local authorities valuation certifi -
cates for the property proposed for insurance. This conformed to a long
tradition in which the insurance of property by private citizens was re-
garded as a public act of direct concern to neighbors. Because the pri-
vate morality of “suspect” groups undertaking such public acts, such as
aliens, Jews, and socialists, was not deemed trustworthy, a degree of of-
fi cial monitoring was required. By contrast, in liberal regimes, such as
Britain and Scandinavia, the authorities left the responsibility for this
type of monitoring to the tariff organizations.
Agents, as well as their companies, required a license in some Amer-
ican states and in parts of Europe. In Missouri, under a law of 1866,
only the agents of foreign companies had to be licensed, but most states
did not discriminate in this way.58 Where agents’ licenses were not re-
quired, the offi cial publication of agents’ names was sometimes adopted
as an alternative to direct screening. In several German states, begin-
ning in the 1820s and 1830s, all agency appointments were subject to
police permission. In Prussia, this involved an investigation of the “reli-
ability” and “irreproachability” of candidates proposed for agencies, their
fi nancial circumstances, any previous criminal record, and any former
political affi liations (Antecedentien). According to one critic, it was this
system of police monitoring of agency appointments, and not any in-
trinsically hostile attitude of the Prussian Ministry of Trade to free trade
in insurance, that explained the diffi culty of obtaining licenses in Prus-
sia and the relatively slow growth of the industry there before agency li-
censing was abolished in 1859.59 The behavioral model of regulation
implies that having the ability to monitor agents and scrutinize claims
56 See the list in Hayden’s Annual Cyclopedia of Insurance in the United States (Hart-
ford, 1906–7), 638–46.
57 ICM 9 (June–Nov. 1880): 248.
58 Western Insurance Review (hereafter WIR) 1 (1867–68): 310–11.
59 RdV 9 (1859): 45–47.
Regulatory Regimes and Multinational Insurers / 77
granted Prussian police a power over their local communities that they
were likely to exercise and defend vigorously, even to the detriment of
the public interest.60
Most public authorities viewed insurance as a source of tax revenue
from an early date. England imposed its fi rst duty on fi re-insurance pol-
icies in the 1690s. In many other countries, fi scal regimes were pre-
dominantly subnational, and in the United States, they were almost ex-
clusively so.61 Massachusetts imposed a stamp tax on policies in 1785,
and New York introduced a punitive 10 percent tax on agency premi-
ums for out-of-state and foreign insurers in 1824.62 Starting in the
1850s, other American and German states and Canadian provinces in-
troduced premium taxes, stamp duties on policies, and, in the case of
Pennsylvania in 1864, a tax on dividends. In many places, there were ad-
ditional local and municipal taxes. Several Belgian cities, for example,
began to tax resident fi re-insurance companies during the late 1870s.63
Municipal charges, such as the New York levy of 1856, went toward the
upkeep of city fi re brigades, while state taxes went toward the mainte-
nance of state insurance departments, or, as in the German Duchy of
Hesse in 1871, they were used for other “public and communal pur-
poses.”64 In the United States, taxation became increasingly discrimi-
natory against foreign companies. In 1862, for example, Massachusetts
imposed a 4 percent premium tax on foreign insurers, but just 1 percent
on local companies. In Kansas in 1899, foreign insurers faced a 6 per-
cent tax on their gross premiums, while local offi ces paid 2 percent.
This took place during a period when all companies were facing an in-
creasing fi scal burden.65
Many governments required insurers to deposit a sum with the state
treasury as a guarantee of solvency and a means of maintaining a suffi -
cient reserve of funds to reinsure the unexpired terms of all outstanding
policies and pay all outstanding losses. In the United States, compulsory
deposits ranged from $10,000 in some smaller states to $200,000 in
New York and Illinois. Authorities usually specifi ed the securities in
which deposits were to be made. In Spain, for example, under the budget
law of 1893, these were government securities, bank-mortgage bonds,
60 Alf Lüdkte, Police and State in Prussia, 1815–1850 (Cambridge, U.K., 1989); Reinhart
Koselleck, Preußen zwischen Reform und Revolution (Munich, 1989).
61 The U.S. exception was a federal stamp duty on all fi re-insurance policies introduced in
1862.
62 The latter was reduced to 2 percent in the wake of the great New York City fi re of 1835.
State of New York, New York Insurance Reports—Condensed Edition, vol. 1: 1830–53 (Al-
bany, 1873), xxiv.
63 Deutsche Versicherungs-Zeitung 21 (1880): 603.
64 RdV 22 (1872): 411.
65 WIR 3 (1869–70): 382; Rough Notes 60 (1917): 104.
Robin Pearson and Mikael Lönnborg / 78
or railway bonds.66 Over time, more deposit requirements were gener-
ally added. This made it diffi cult for smaller companies to invest abroad,
particularly to establish branch networks, without weakening their fi -
nancial position back home. Setting up overseas branches thus increas-
ingly became an option only for the largest fi rms.
Some countries also imposed capital requirements on foreign com-
panies. In a blatant attempt to disadvantage European insurers, a New
York law of 1879 required that all companies operating there have a
paid-up capital of 100 percent. Other U.S. states prohibited companies
from advertising the value of their subscribed capital, which comprised
the shares issued by a company and subscribed to by their owners, be-
fore the latter had paid for them. The regulators’ argument was that the
large differential between subscribed and paid-up capital of most Brit-
ish and European companies—the proportion of paid-up capital was
commonly between 10 percent and 25 percent—amounted to a decep-
tion of American policyholders, whose native companies typically had
all their subscribed-share capital fully paid up. Advertisements showing
impressive amounts of subscribed capital implied that foreign compa-
nies had a solidity, which, it was claimed, they did not possess. Share-
capital stipulations were also a feature of insurance regulation in Eu-
rope, but they did not become the prohibitive device that some U.S.
regulators sought during the 1870s and 1880s.67
Many states also required regular company reports. Massachusetts
introduced a general reporting requirement as early as 1818. In other
places, including the Netherlands after liberalization in 1880, Spain be-
fore 1893, and some American states such as Missouri before 1869, no
offi cial monitoring of insurance-company accounts was in force. In gen-
eral, however, in Europe, North America, and Japan, there was a trend
after 1850 toward more formal supervision of the industry. By the 1860s,
several German states had standing government commissions on fi re
insurance. Switzerland established a federal insurance bureau in Berne
in 1885. During the early 1900s, statutory insurance bureaus or offi ces
of inspection were established in Austria, Denmark, Germany, Italy,
Japan, Montenegro, Norway, Spain, Sweden, and Turkey.
In the United States, fi ve states—New Hampshire, Vermont, New
York, Massachusetts, and Rhode Island—had established permanent
insurance departments by 1860. Others followed after the Civil War.
66 Rosales and Quiza, “Los Seguros en España,” 198. For similar requirements in Russia,
see RdV 22 (1872): 21; Paul J. Best, “Insurance in Imperial Russia,” Journal of European
Economic History 18, no. 1 (1989): 139–69.
67 Cf., for example, the Spanish insurance law of 1908, which required all companies seek-
ing a license to have just 25 percent paid up on their shares: Rosales and Quiza, “Los Seguros
en España,” 188–89.
Regulatory Regimes and Multinational Insurers / 79
Such departments were run by commissioners or superintendents who
were political appointees and were frequently criticized for their lack of
expertise and their interference in company affairs. In the South and
the West, they could become a vehicle of populist hostility against east-
ern and foreign corporations. In some states, for instance New York in
the 1900s, insurance departments also fell victim to political squabbling
between the superintendent, who developed a power base of his own in
the legislature, and the governor, who could only recommend his dis-
missal to the legislature. Here, too, the behavioral model of regulation
appears applicable. American and foreign companies alike complained
that some state insurance departments were ineffi cient or corrupt. Sev-
eral New York offi cials were brought before the courts in this period
and charged for pursuing their own interests by dubious means.68
In the United States, however, probably more than in most other
countries, bureaucratic control was highly contested during this period.
In particular, the constitutional independence of American courts pro-
vided a medium through which business regulation was interpreted and
amended.69 One legal manual, published in 1893, cited over seven thou-
sand judgments, overwhelmingly made in U.S. courts, which defi ned
the law on fi re insurance. Courts deliberated over policy contracts, in-
surable interest, disclosure and utmost good faith, liability, the pro-
cesses of underwriting and loss adjustment, licenses, and deposits. Sev-
eral rulings also clarifi ed the legal jurisdiction and the rights of parties
in interstate and foreign insurance contracts.70 There is scope for fur-
ther research on this subject, but the power of judicial mediation may
have helped to offset the impact of unpredictable and discriminatory
regulation against foreign companies and to sustain their confi dence in
the American market as it moved into the Progressive Era.71
A growing volume of miscellaneous regulations in different coun-
tries also covered general insurance-company operations. The most oner-
ous of these was the “proof of demand” (Bedürfnis) requirement, which
bedeviled foreign insurers trying to expand in Central and Eastern Eu-
rope before the 1870s. Under the Prussian law of 1837, for example,
new companies wishing to obtain licenses from the Ministry of Interior,
68 ICM 14 (Jan.–June 1883): 146–47, 154.
69 Tony A. Freyer, “Business Law and American Economic History,” in The Cambridge
Economic History of the United States, vol. 2: The Long Nineteenth Century, eds. Stanley L.
Engerman and Robert E. Gallman (Cambridge, U.K., 2000), 435–82; Edward A. Purcell Jr.,
Litigation and Inequality: Federal Diversity Jurisdiction in Industrial America, 1870–1958
(New York, 1992).
70 George A. Clements, Digest of Fire Insurance Decisions in the Courts of the United
States, Great Britain and Canada (New York, 1893).
71 Foreign insurers, of course, could face hostile juries as well as sympathetic judges in the
United States, most notably when involved in claims litigation after the San Francisco earth-
quake of 1906. Lönnborg and Pearson, “Multinational Insurers and Catastrophic Loss.”
Robin Pearson and Mikael Lönnborg / 80
or licensed companies wishing to appoint additional agents in cities,
had to provide proof that there was a “demand” for their services. This
proof took the form of a testimony from the local city or county admin-
istration. It does not seem to have been based on any scientifi c investi-
gation of data. Indeed, critics of the regulation pointed out that many
existing insurance companies in small towns employed as their agents
local mayors and councilors, who therefore had a vested interest in pre-
venting the admission of rival offi ces.72 The requirement was abolished
in Prussia in 1859, but it survived in other places, such as Austria and
Baden.
Some states prohibited reinsurance in unlicensed out-of-state and
foreign companies. New York, Massachusetts, and Kansas passed such
regulations in the 1870s, and there was a further wave of such regula-
tion in other American states in the later 1890s and early 1900s. As new
lines of insurance evolved, regulators also became increasingly deter-
mined to keep them separate. First, their evaluation of rates and com-
pany reserves depended upon the examination of distinct lines of busi-
ness. Second, regulators were concerned that losses in one line should
not be covered by transfers at the expense of policyholders in another.
The New York law of 1879, for example, required any foreign company
that sought a license to underwrite in New York, and that sold some
combination of fi re, transport, and life insurance in other markets, to
restrict itself to just one line of business in the United States.73 Few Eu-
ropean countries had such regulations. Prussia did restrict life-assurance
companies to one line in 1868, but most other countries did not.
Our account suggests that a wide spectrum of regulatory regimes
emerged in insurance before 1914, ranging from liberal to monitorial
and prohibitive.74 This spectrum is illustrated in Table 6, which displays
the regulatory environment for out-of-state and foreign insurers. Liberal
countries were those without state supervision, or with minimal regula-
tion and taxation. Monitorial nations had standing insurance commis-
sions or inspection departments; they had general licensing and deposit
requirements; and they imposed other types of regulation, such as policy
conditions and loss-adjustment procedures. Countries whose require-
72 RdV 13 (1863): 368–72.
73 WIR 3 (1869–70): 689–70; RdV 30 (1880): 358–59; ICM 7 (July 1879): 101. Some
states were quick to follow New York’s lead. In Ohio, for instance, a law came into effect in
1880 banning companies from writing both fi re and life insurance. ICM 8 (Jan. 1880): 107.
74 Morton Keller has described four types of business regulation, namely, American “plu-
ralist,” German “corporatist,” British “contract liberal,” and French “dirigiste.” However,
these have questionable relevance to insurance where the multiple layers of regulation in
countries such as Germany and Belgium defy any such neat typologies based on political
economy. Morton Keller, “The Pluralist State: American Economic Regulation in Compara-
tive Perspective, 1900/1930,” in Regulation in Perspective: Historical Essays, eds. Morton
Keller and Thomas McCraw (Boston, 1981).
Regulatory Regimes and Multinational Insurers / 81
ments were prohibitive excluded particular categories of foreign insurer
for certain periods, or exercised strict bureaucratic or police controls
over companies, agents, and policyholders, effectively limiting entry.
Many regimes moved back and forth along this spectrum at differ-
ent times. In some midwestern American states, regulation shifted from
liberal to monitorial to prohibitive as populist feeling against out-of-
state insurers ran high toward the turn of the century. Most German
states moved from prohibitive to liberal by the 1860s and 1870s, and
then switched to monitorial regimes during the 1900s. The Netherlands
changed from prohibitive to liberal after 1880, while Spain and Japan
moved from liberal to monitorial after 1893 and 1900, respectively.
Table 6 suggests a general convergence of regulatory regimes of the
monitorial kind around the turn of the century, indicating a trend to-
ward international standardization before 1914.75 One can also, how-
ever, observe persistent national differences among regulatory regimes
throughout the period. Much German insurance regulation, for example,
concerned relations between companies, agents, and customers at the
Table 6
Regulatory Regimes in Insurance before 1914
Liberal Monitorial
aProhibitive
Belgium Austria (from 1873, 1905) Austria (to 1873)
France Brazil (1901) Bavaria (to 1872)
Germany Canada (from 1854) Hanover
Hamburg Denmark (1904) Hesse
Bremen Finland (1892) Hesse-Darmstadt
Lubeck France (1905)bNetherlands (to 1880)
Frankfurt Germany (1901) Prussia (1837–59)
Oldenburg Italy (1904) Russia
Brunswick Japan (1900) Saxony (1828–63)
Electorate of Hesse Montenegro (1904) Switzerland (to 1885)
Saxon duchies Norway (1904)
Mecklenburg duchies Prussia (from 1859)
Small principalities Spain (from 1893, 1908)
Japan (1861–1900) Sweden (1904)
Netherlands (1880–1920) Switzerland (from 1885)
Spain (1829–93) Turkey (1904)
Sweden (to 1904) U.S.A. (Massachusetts) (1855)
United Kingdom
a Dates in italic = the fi rst state supervision legislation.
b Life insurance only.
75 So far, we have found no evidence of formal international collaboration among insur-
ance regulatory bodies during this period.
Robin Pearson and Mikael Lönnborg / 82
point of sale. There was less emphasis on regulating annual reports, de-
posit and capital requirements, and on taxation. This contrasts with the
United States, where much state and city regulation was concerned with
investor and shareholder interests and, to an extent, with extracting tax
revenue.
How can these differences be explained? First, cultural and politi-
cal traditions were important. Laissez-faire attitudes largely overrode
public-welfare concerns in determining the scope of British insurance
regulation. In Germany, the concept of the patriarchal state infused the
highly bureaucratic policing of insurance, even after licensing controls
were relaxed in the 1860s. In the United States, the strength of early sa-
lus populi traditions in state and community government helps explain
the support for the early adoption of formal supervision. As elsewhere
in the U.S. economy, the transition in insurance during the second half
of the century to what historian William Novak has called a “new con-
stitutional regulatory regime,” based on the concept of due process of
law and limitations on police power, was refl ected in an increasing vol-
ume of litigation between companies and regulators.76
Second, market structures also infl uenced national differences in
regulatory systems. The ease of incorporation in the United States, the
large numbers of insurance companies, and the level of competition
may explain the diffi culty regulators faced in dealing with the persistent
problem of illegal “undergrounders.” By contrast, in Central and East-
ern Europe, where the number of companies was smaller, the licensing
system was less porous. There, state monopolies in building insurance
and compulsory insurance were factors in shaping regulatory practice.
A recurrent argument in some countries concerned the need to prevent
insurance premiums from fl owing abroad in order to support the do-
mestic institutions.77 This view was not generally aired in the United
States, where occasional calls for nationalization or compulsion were
quickly denounced as state despotism and criticized as a species of “ori-
entalism.”78 In parts of Europe, but not in the United States, state in-
surance monopolies created vested interests within the very bureaucra-
cies that provided the regulators.
Thus, the operation of regulatory regimes in insurance lends sup-
port to the behavioral model of regulation advanced by economists Sam
Peltzman, Richard Posner, and others. The problems for private insur-
76 William J. Novak, The People’s Welfare: Law and Regulation in Nineteenth-Century
America (Chapel Hill, 1996).
77 Argued by regulators, for example, in the Swiss cantons. Martin Körner, Banken und
Versicherungen im Kanton Luzern vom ausgehenden Ancien Régime bis zum ersten Welt-
krieg (Lucerne, 1987), 157.
78 Spectator 12 (1874): 146–47.
Regulatory Regimes and Multinational Insurers / 83
ance companies in Central and Eastern Europe, especially those from
outside the state, caused by the interlocking interests of civil servants
and state insurance organizations, were exacerbated by the frequent ar-
bitrariness of regulatory practice. This capriciousness contrasted
sharply with the seemingly more transparent, certainly more public, li-
censing system in the United States. Prussian regulators were career
civil servants or local magistrates from the gentry or from aristocratic
backgrounds. Early U.S. insurance commissioners came from mercan-
tile or professional backgrounds and were generally sympathetic to the
needs of private industry. Their business leanings caused them to be
more willing than their colleagues in Germany, Austria-Hungary, or
Russia to compromise with companies in disputes over licensing. An-
other factor that contributed to national differences may have been the
more intense press scrutiny of regulators in the United States, com-
pared to Europe. The monopoly position of public societies in building
insurance, the presence as offi cers of public societies of the same state
offi cials who regulated their private competitors, and the resistance
among civil servants to any attempts to reduce their regulatory powers
combined to make many European regulatory regimes substantively
different from their U.S. counterparts, even though many of the objects
and methods of regulation were the same.
Conclusion
What of the impact of regulation on international insurance? It is
true that, generic regulations notwithstanding, not all foreign insurers
were treated alike within a jurisdiction, and not all responded to regula-
tions in the same manner.79 There is some indication that more oner-
ous deposit requirements made it harder for smaller companies to ex-
pand abroad toward the end of the nineteenth century. It does not seem,
however, that there was ever any “race to the bottom,” in which foreign
insurers searched for markets with no prudential regulations. Some of
the evidence we have presented suggests that insurance companies re-
garded regulation as a credible commitment and that legislation made
the conduct of foreign business less uncertain. First, prudent regulations
to an extent protected companies from new entrants. Second, placing
itself under supervision helped legitimize a company and enhance pub-
lic confi dence in its operations. From this point of view, it was more
79 Wilkins has demonstrated this in the case of U.S. life insurers in Germany before 1914,
where New York Life, thanks to the diplomatic talents of its vice president, coped with hostile
regulators more effectively than the other American offi ces. Wilkins, Emergence of Multi-
national Enterprise, 103–7.
Robin Pearson and Mikael Lönnborg / 84
important to expand in markets with sustainable growth and a stable
institutional environment than merely to engage in a hunt for unregu-
lated markets.
The international convergence of regulation may have facilitated
cross-border trade, especially for larger companies underwriting around
the globe.80 The managers of Sun Insurance of London, for example,
constantly drew on their experience of operating in many different mar-
kets when advising their overseas agents and forming responses to
changing competitive and legal regimes.81 On the other hand, our data
indicate that similar regulatory regimes had a differential effect on the
internationalization of insurance before 1914, suggesting that there was
no straightforward relation between regulation and growth. Comparing
Tables 2 and 3 with Table 6, we fi nd that liberal and monitorial regimes
coexisted with both high and low degrees of import penetration. For-
eign companies enjoyed high levels of market share in Spain under both
liberal and monitorial regulation before and after 1893, while their share
of the highly competitive market in liberal Sweden remained low. Fire-
insurance imports were low in monitorial Japan, Germany, and Swit-
zerland, but high in monitorial Canada, Finland, and Turkey. Under the
extensive state supervisory system in the United States, levels of import
penetration fl uctuated above and below the 30 percent mark between
the 1870s and the First World War. In Germany, a market split between
stock and mutual insurers, powerful state monopoly institutions, and
highly bureaucratic regulatory practices made life diffi cult for foreign
companies. In the United States, despite increasing levels of discrimi-
nation, regulation, or its interpretation by an independent judiciary, was
mostly benefi cial for foreign fi re insurers. State supervision seems to
have improved consumer confi dence in the industry, as did educational
efforts to raise awareness of hazards and promote fi re prevention. Only
under prohibitory regimes was there a clear association with low levels
of direct underwriting by foreign fi rms. In Russia, for example, foreign
companies focused mainly on reinsurance, capturing over half of that
market after 1870.
We cannot yet answer one key question: whether or how regulatory
regimes were factored into risk assessment and premium rates.82 Any
convergence of regimes may have been manifested in rate convergence
80 Peter Borscheid, “Systemwettbewerb, Institutionenexport und Homogenisierung: Der
Internationalisierungsprozess der Versicherungswirtschaft im 19. Jahrhundert,” Zeitschrift
für Unternehmensgeschichte 51, no. 1 (2006): 26–53.
81 Robin Pearson, “Las compañias extranjeras en España, 1800–1939,” in Empresas y
Mercado del Seguro en la España Contemporánea, eds. Jerònia Pons Pons and Maria Ange-
les Pons Brias (forthcoming).
82 We wish to thank Janette Rutterford for raising this point with us.
Regulatory Regimes and Multinational Insurers / 85
across markets, but we have almost no evidence on this at present.83 Pre-
miums, of course, were determined by a wide range of variables, includ-
ing technological change, changes in risk composition, a fi rm’s capacity
for portfolio diversifi cation across different markets, and administra-
tive and marketing costs, as well as the costs associated with regulatory
regimes. The effect on premium rates of any convergence in regulation
may have been either reinforced or offset by other variables.
Our analysis also indicates that nonregulatory factors helped to de-
termine levels of cross-border trade in insurance. These include the di-
verse choice of institutional forms for multinational enterprise that may
have offset regulatory disincentives to entry; the follow-the-leader ef-
fect; exogenous shocks, such as great fi res, which increased the demand
for risk spreading; bilateral affi nities, including close cultural and some-
times political ties between markets; competitive conditions; and the
opportunity costs of entering new and distant markets. In general, in-
ternational insurance was driven by a search for stable markets and the
need to sustain an incremental growth of business. By contrast, there is
little evidence that insurance companies explicitly followed their clients
abroad before 1914. This seems to have been largely a phenomenon of
the period after the Second World War.84
There is evidence from Scandinavia, Spain, Switzerland, and China
that insurers chose a chain process of internationalization. For some
companies, it was clearly important to acquire prior knowledge of more
distant markets, through reinsurance, by contacting general agents, or
through ethnic connections with local settlements of merchants, before
venturing further from their home markets. The choice of different
forms for international insurance was infl uenced not only by the need
to comply with regulatory requirements, but also by considerations of
costs, effi ciency, monitoring, and psychic distance.
Since the early nineteenth century, insurance has been character-
ized by international arrangements to diversify risks and benefi t from
different business cycles. Insurance has attracted a lot of attention from
83 There is only limited evidence of premium rate regulation through most of the period
we cover. In Austria, the power to examine policy conditions and premium rates was ex-
tended in 1860 to provincial commissioners appointed by the Ministry of the Interior, but
how far they amended rates is unknown. In Russia, the premium rates of the state-chartered
fi re-insurance companies were fi xed by their charters. In the United States, despite the ex-
tension of antitrust regulation to fi re insurance—sixteen states had enacted so-called anti-
compact laws by 1907—no state directly regulated premium rates before Kansas did so in
1909. The Kansas law was contested through the courts, until it was fi nally upheld by the Su-
preme Court in 1914. Anticompact laws counted from Hayden’s Annual Cyclopedia of Insur-
ance in the United States, 1906–7 (Hartford, 1907), 28–45.
84 Mikael Lönnborg, “Svenska försäkringsbolag tidiga internationalisering,” Nordisk
Försäkringstidskrift no. 4 (2000): 312–36; and “Skandiakoncernens internationella
verksamhet 1887–1995,” Nordisk Försäkringstidskrift no. 3 (2002): 1–20.
Robin Pearson and Mikael Lönnborg / 86
domestic legislators, whose goals have included upholding the solvency
of fi rms and securing public confi dence in a product that has a major
impact on the economy. Other factors shaping regulation include con-
sumer lobbying, the self-interest of politicians and regulators, the un-
evenness of enforcement in some places, protection for native compa-
nies, and maximizing state revenues. This paper represents an initial
exploration of the effect of this regulation on the diffusion of insurance
around the world before 1914. This effect was highly differentiated by
market, and by method chosen for international underwriting, such as,
for example, whether this was done directly by agency or indirectly by
reinsurance. We can, however, draw two general conclusions that speak
to current debates about the impact of deregulation on the globalization
of insurance.85 First, each category of regulatory system before 1914,
under different circumstances and in different markets, could enhance
the internationalization of this fi nancial service—outright deregulation
was not necessary to achieve this result. Even prohibitory regimes could
stimulate internationalization by creating an increased demand for re-
insurance. Second, the rate of internationalization could decline in
some places, under different regulatory regimes, even while the insur-
ance industry itself was growing rapidly.
85 Swiss Re, “Emerging Markets: The Insurance Industry in the Face of Globalisation,”
Sigma no. 4 (2000).
