1 ‘A place for capital controls' Economist, 3 May 2003, p 16. This paper is not concerned with the relationship between the effectiveness of legal systems and success in attracting foreign portfolio investment. Gordon Walker has provided a particularly useful reviews of research by Rafael La Porta's team on why and in what ways’ law matters' for foreign portfolio investors: G Walker ‘Upgrading corporate governance in East Asia’ in Keong, L (ed) Corporate Governance: An Asia-Pacific Critique (London: Sweet & Maxwell, 2002 Google Scholar).

2 See eg J Stopford and S Strange Rival States, Rival Firms Competition for World Market Shares (Cambridge: Cambridge University Press, 1995) and the annual World Investment Reports produced by the United Nations Conference of Trade and Development (UNCTAD). For examples of these arguments, see A Perry ‘Effective legal systems and foreign direct investment: in search of the evidence’ (2000) 49 ICLQ 779; and A Perry Legal Systems as a Determinant of FDI: Lessons from Sri Lanka (London: Kluwer Law International, 2001) pp 1346.

4 Dunning, J Multinational Enterprises and the Global Economy (Wokingham: Addison-Wesley, 1993) p 68 Google Scholar.

5 Dunning, n 4 above, pp 79 and 82. One survey of evidence concluded that the market size is the most important location specific advantage, and that the cost of labour, and (historically) the possibility of avoiding tariffs were also important. However, it also concluded that we do not have the information necessary to calculate the relative importance of each of these factors with certainty: UNCTC The Determinants of Foreign Direct Investment: A Survey of the Evidence (New York: United Nations, 1992) p 59.

6 See eg the Global Terrorism Index 2003–2004 produced by the World Market Research Centre, available at http://www.worldmarketsanalysis.com/t-index_2003.html as at 19 August 2003. Index author Guy Dunn reportedly told CNN News that the center ‘created its global terrorism index in response to requests from its business clients. In assessing the risk of 186 countries, it found the isolationist state of North Korea the least likely to suffer a terror attack’: Summary of CNN interview available at http://www.edition.cnn.com/2003/BUSINESS/08/19/global/terror.biz as at 19 August 2003.

7 For an elaboration of the arguments made in this section, see Perry (2000), n 3 above; and Perry (2001), n 3 above, pp 13–46.

8 Although this assertion is frequently made, commentators rarely take the time to spell out its theoretical roots in any detail, let alone to prove it empirically.

9 Bomer, S, Brunetti, A and Weder, B Political Credibility and Economic Development (London: St Martins Press, 1995)Google Scholar.

10 See Perry (2001), n 3 above, ch 2.

11 H Wang ‘Informal institutions and foreign investment in China’ (2000) 13 The Pacific Review 4 at 525. Perceptions of corruption were measured by the Transparency International corruption index: http://www.transparency.org. The time periods examined were 1980–85, 1985–90 and 1990–95.

12 Other factors tested were market size, living standards, economic growth rates, sunken investment and natural resources.

13 Wang, n 11 above, at 527–530.

14 Wang, n 11 above, at 525.

15 Wang, n 11 above, at 525.

16 S Wei ‘Can China and India double their inward foreign direct investment?’ 1999 version of draft paper available at the World Bank's Rapid Response Unit, 30 November, available at http://www.worldbank.org/rru.

17 Coase, R The Firm, the Market and the Law (Chicago: University of Chicago Press, 1988) p 19 Google Scholar.

18 Perry (2001), n 3 above.

19 Source: World Bank figures for 2001, published at http://www.worldbank.org/data/quickreference as at 9 January 2003.

20 Source of FDI data: UNCTAD World Investment Report 2001: Promoting Linkages (Geneva: United Nations, 2001) Annex A, p 301ff. Source of population data for per capita figures: World Bank World Development Indicators Database figures for 1999, published in http://www.devdata.worldbank.org, as at 5 November 2002.

21 Since the surveys are designed to be repeated at regular intervals, researchers should soon be able to track changes in these perceptions over time. Some relevant indicators were not available for some countries in the sample. In these cases, the indicator was excluded from this study. Eg data for questions about the influence firms have over the executive, legislature and ministries, which are intended to measure state capture - a negative thing under the dominant theory - are not available for India and Bangladesh.

22 See http://www.fias.net/services.html.

23 D Kauffman, A Kraay and P Zoibo-Lobaton Governance Matters Policy Research Working Paper No 2196 (Washington DC: World Bank, 1999a) p 1, available at http://www.worldbank.org/wbi/govemance/pubs/govmatters.html. For detailed methodology, see also D Kauffman, A Kraay and P Zoibo-Lobaton Aggregating Governance Indicators World Bank Policy Research Working Paper No 2195 (Washington, DC: World Bank, 1999b), available at http://www.worldbank.org/wbi/govemance/pubs/aggindicators.html.

24 Polls and surveys were conducted by 13 different NGOs, think tanks, international organisations, and political and business risk rating agencies and cover 150 countries. Indicators were produced in 1997–98 and 2000–01.

25 The possibility of effectual relationship was statistically eliminated.

26 Kauffman et al (1999a), n 23 above, pp 16–17.

27 Kauffman et al (1999a), n 23 above, pp 7–8.

28 Source of GRICS data: 1997–98 data set from D Kaufmann, A Kraay, and P Zoido-Lobaton Governance Matters, II: Updated Indicators for 2001–02 World Bank Policy Research Working Paper No 2772 (Washington DC: World Bank, 2002), available at http://www.worldbank.org/wbilgoverrnance/pubs/govmatters2.html. The raw data set was downloaded as a spreadsheet file from the World Bank Institute Governance site at http://www.worldbank.org/wbi/govemance/govdata2001.htm.

29 Usually called the ‘90% confidence interval’. The tentative nature of the data is emphasised by the use of 90%, rather than the stricter, and more common, 95%.

30 Kauffman et al (1999b), n 23 above, p 2.

31 Graphs demonstrating these conclusions have been excluded due to space constraints.

32 Confidence intervals are calculated as +/-1 1.64 x standard error, and range from 0.508 at the narrowest, to 1.296 at the widest - a difference of 155%.

33 Kauffman et al (1999a), n 23 above, p 10. Standard errors measure spread - that is, the average deviation of a particular indicator value (for that indicator in that country) in each individual poll, from the average (mean) indicator value (for that indicator in that country) of all the polls. They are calculated as the sum of the deviations of each observation from the mean of the observations, divided by the sample size. Standard errors in this data range from 0.155 at the lowest to 0.395 at the highest - again, a difference of 155%.

34 D Kauffman and A Kraay ‘Governance indicators, aid allocation, and the Millennium Challenge Account’ draft paper for discussion (Washington DC: World Bank, 2002) p 6, available at http://www.worldbank.org/wbi/governance/mca.htm.

35 Kauffman et al (1999a), n 23 above, p 1.

36 S Djankov, R La Porta, F Lopez de Silanes and A Shleifer ‘The regulation of entry’ (2002) 117 Quarterly Journal of Economics 1 at 5–6.

37 Djankov et al, n 36 above, at 1–2.

38 Source: Entry Regulations’ Topic Reports' generated for all available countries, and for each sample country using interface available at http://rru.worldbank.org/doingbusiness/SnapshotReports/EntryRegulations.aspx, as at 18 December 2002.

39 http://www.rru.worldbank.org/doingbusiness/Methodology/ContractEnforcement.aspx, as at 18 December 2002.

40 S Djankov, R La Porta, F Lopez de Silanes and A Shleifer ‘The practice of justice’ (2003) 118 Quarterly Journal of Economics 3 at 518.

41 Source: Contract Enforcement ‘Topic Reports' generated for all available countries, and for each sample country using interface available at http://www.rru.worldbank.org/doingbusiness/SnapshotReports/ContractEnpx, as at 18 December 2002. For detailed methodology of this study see S Djankov, R La Porta, F Lopez de Silanes and A Shleifer’ Courts' (2003) 118 Quarterly Journal of Economics 2 at 453.

42 Djankov et al, n 36 above, at 7.

43 Djankov et al, n 36 above, at 8.

44 Questionnaires were implemented on behalf of the World Bank in 1999–2000 by the Gallup Organisation (East Asia, Pakistan, Latin America, OECD), AC Nielsen (Eastern Europe and Turkey), the Confederation of Indian Industries (India), the Harvard Centre for International Development (Africa), the Egyptian Centre for Economic Studies (Egypt), Lidee Khmer (Cambodia), the University of Chamber of Commerce (Thailand) and Bangladesh Export Development Project (Bangladesh): G Batra, D Kauffman and A Stone ‘Voices of the firms 2000: investment climate and governance findings of the world business environment survey (WBES)’ discussion draft (Washington DC: World Bank, 2002) p 3 available at http://www.worldbank.org/wbi/governance/wbes. See also G Batra, D Kauffman and A Stone Investment Climate Around the World: Voices of the Firms from the World Business Environment Survey (Washington DC: World Bank, 2003).

45 Batra et al (2002), n 44 above, p 6. The WBES sample was focused on providing a representative selection investment sources, sectors, locations of industry, export orientation and so on.

46 Source: The World Business Environment Survey (WBES) 2000 A data set for each country was generated from the WBES interface at http://www.info.worldbank.org/governance/wbes. All figures are rounded to the nearest whole percent.

47 The Kvetch factor - that is, the propensity of some individuals to complain about anything and everything -was identified and controlled for by asking questions about the quality of services (eg postal), responses to which should be uniform: Batra et al (2002), n 44 above, p 31.

48 Options: always, mostly, frequently, sometimes, rarely, never. ‘Effective’: always, mostly, frequently.

49 Options: always, mostly, frequently, sometimes, rarely, never. ‘Effective’: always, mostly, frequently for ‘effective’ and ‘predictable’; and sometimes, rarely never for ‘required’.

50 Options: fully agree, agree in most cases, tend to agree, tend to disagree, disagree in most cases, strongly disagree. ‘Effective’: fully agree, agree in most cases, tend to agree.

51 Options: very helpful, mildly helpful, neutral, mildly unhelpful, very unhelpful. ‘Effective’: very helpful, mildly helpful.

52 Options: always, mostly, frequently, sometimes, rarely, never. ‘Effective’: always, mostly, frequently.

53 This requires dividing all FDI and legal systems figures by the relevant US figure and multiplying by 100.

54 Ranking for stocks and flows is the same.

55 T Beck, A Demirgüc-Kunt and V MaksimovicFinancial and Legal Constraints to Firm Growth: Does Size Matter Working Paper 2784 (Washington DC: World Bank, 2002) available at http://www.econ.worldbank.org/files/12029_wps2784.pdf. Thanks to my PhD student, Ali Afshar, for pointing this out to me.

56 Response options: no, minor, moderate, major obstacle. ‘Effective’: no obstacle, minor obstacle. See n 45 above for the need to use caution in seeking to rank countries on the basis of a single indicator.

57 This conclusion is based on discussion at an international symposium on the Market for Legal Systems held by the Lord Chancellor's Department in London on 24–25 October 2002. The World Bank Legal and Judicial Reform Practice group is trying to collect such data but it currently limited to very few countries of disparate location: http://www.worldbank.org/legal/database.

58 ‘Law reform: don't sue’ Economist, 26 October 2002, p 34.

59 See eg S Macaulay ‘Non-contractual relations in business: a preliminary study’ in Buckley, P and Mitchies, J (eds) Firms, Organisations and Contracts (Oxford: Oxford University Press, 1996) pp 339–358 Google Scholar.

60 Wang, n 11 above, at 530. See also Perry (2001), n 3 above.

61 Wang, n 11 above, at 525. Wang notes that ‘informal networks in Chinese society stand out for three reasons: ‘(1) the degree to which informal networks contradict and undermine formal institutions is extremely high; (2) the influence of informal networks is unusually strong in every aspect of Chinese life; and (3) the level of institutionalization of Chinese networks is very low‘: at 531. Importantly, she distinguishes between guanxi and bribery on four counts. First, ‘unlike bribery, which is a purely instrumental exchange of favours, guanxi takes place among people who - directly or indirectly - are bound together by sentiments and social etiquette’. Secondly, guanxi ties are long term and are ‘not designed to facilitate any specific exchange of favors’. Thirdly, gifts exchanged in guanxi are often merely symbolic, while ‘in bribery, the practical and monetary value of gifts is the sole criterion for evaluating the relationship’. Fourthly, while bribery is usually an isolated exchange, ‘guanxi is a relationship embedded in a complex social network’: at 537–538.

62 For an interesting comparison of the efficiency of formal and informal systems, see A Stone, B Levy and R Paredes ‘Public institutions and private transactions: a comparative analysis of the legal and regulatory environment for business transactions in Brazil and Chile’ in L Alston, T Eggertsson and D North (eds) Empirical Studies in Institutional Change (Cambridge: Cambridge University Press, 1996) pp 95–128.

63 It should also be noted that foreign investors in India were 21% more likely than their domestic counterparts to report the typical firm in their business would report 100% of their sales. No other statistically significant differences were found relating to tax.

64 See n 45 above for the need to use caution in seeking to rank countries on the basis of a single indicator.

65 Batra et al (2002), n 44 above, p 6, citing an unpublished regression analysis of the WBES data on general constraints to doing business: Nagarajan, N et al Perceptions of the Investment Climate: Foreign vs. Domestic Investors (Washington DC: HAS, 2001 Google Scholar). General constraints examined were financing, taxes and regulations, inflation, policy instability, anti-competitive practices, infrastructure, street crime, exchange rate, organized crime, corruption, judicial system. However, when data was examined at a regional level it seemed that there was no difference between the firms, who reported themselves to be similarly constrained. Batra et al's description of this study is slightly confusing and seems to suggest at one point that characteristics other than foreign ownership account for the differences in firm experience. See Batra et al (2002), n 44 above, p 16. The WBES did seek to ensure that the proportion of foreign to domestic investors surveyed was at least 15% of each country sample: p 4.

66 J Hellman, G Jones and D Kaufmann ‘Far From Home: Do Foreign Investors Import Higher Standards of Governance in Transition Economies?’ draft for discussion (Washington DC: World Bank, 2002) p 3, available at http://www.worldbank.org/wbi/govemance/pubs/farfromhome. html.

67 GRICS figures are calculated by adding 2.5 for each point estimate, in order to remove any minus signs; and then dividing all figures by US figures and multiplying by 100. WBES figures are divided by US figures and multiplied by 100.

68 Hellman et al, n 67 above, p 3.

69 Dunning, n 4 above, p 66. The most important ownership advantages identified in a 1992 survey of evidence were the investor's technological supremacy, product differentiation, and management expertise; and the availability of a concentration of similar investors: UNCTC, n 5 above, pp 55–58.

70 Caves, R Multinational Enterprises and Economic Analysis (Cambridge: Cambridge University Press, 1996) pp 37.46 and 52–53 Google Scholar.

71 Batra et al (2002), n 44 above, pp ix-x.

72 For a critical review of the cookie-cutter approach, see Perry (2001), n 3 above, chs 1–3.

73 Wang, n 11 above, at 544–546.

74 Wang, n 11 above, at 544–546.

75 Wang, n 11 above, at 544–546.

76 Batra et al (2002), n 44 above, p 4.

77 Batra et al (2002), n 44 above, pp 29–32.

78 Batra et al (2002), n 44 above, p ix.