https://cpes.org.uk/om/files/original/dc582e1357a4d0b5aa9738131be00363.pdf 0e1349ae44ec18e998dda6d7789e74bc Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource International Contributions to Labour Studies 4 Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Text Any textual data included in the document International Contributions to Labour Studies (1994) 4, 115-127 'Over here and over there': a comparison of top executive pay in the UK and the USA Brian G. M. Main, Charles A. O'Reilly Ill and Graef S. Crystal Introduction Over recent years there has been an increase in interest, on the part both of the general public and specialist commentators, in the levels of top executive pay being awarded in publicly held firms. This has occurred on both sides of the Atlantic, and the discussion of the topic has often been set in a comparative mode. Critics in the USA claim that Chief Executive Officer (CEO) salaries there are high in comparison with their European and Japanese competitors. And those in the UK often claim that British top executive salaries need to be boosted for the very reason that they are high in the USA, and failure to keep UK salaries in line would result in a harmful 'brain drain' of managerial talent to the USA. The present paper attempts to quantify the size of the wage gap between the salaries of top executives in the UK and those of comparable executives in the USA. Some work in this area has already been carried out by Crystal (1991) who compared the typical pay package of the CEO in a top flight company in each of Japan, Germany, France and the United Kingdom with their equivalent in the USA. He found that the USA and the UK were the closest in terms ofpay generosity. The present study uses data from 1989 for some 915 companies in the USA and from 1990 for some 220 companies in the UK to make a comparison that is applicable to a wider definition of top executive. As will be seen below, an attempt is made to control for the size of the company (measured by sales, i.e. turnover), its performance (measured by stock market return), the executive's length of service with the company, and the executive's length of service in his current post as CEO. From the British perspective, the idea that US levels of pay influence domestic pay levels is more convincing in the presence of some observed labour market flows of executives between the two countries. The claim that executives in one country are paid more than those in another is of no consequence unless there is some evidence that the two are acceptable substitutes. In fact, there are remarkably few cases of British executives going off to assume the CEO role at any US companies. There are, however, some notable examples of US executives making a successful transition into the top jobs at British companies. Sir Richard Giordano at BOC, and *University of Edinburgh, Stanford University and University of California, Berkeley, respectively. The authors are grateful for research support provided under ESRC grant R000232031 and for research assistance from James Johnston. They are also indebted to two anonymous referees for helpful comments. 1052-9187/94/010115 + 13 $08.00/0 © 1994 Academic Press Limited 116 Brian G. M. Main et al. Mr Henry Wendt and Mr Bob Bauman at SmithKline Beecham are well-known examples. It is also true, however, that it requires only the credible possibility of such movement to result in an influence on pay awards. The market for top executives is not a volume market. Each company is faced with a unique situation and a unique list of requirements demanded from its CEO. As Yarrow (1972) makes clear, the hiring/appointing decision being made is not about how many top executives to hire, but about the quality of the executive. There is only one CEO, and most companies are looking for the best person possible to fill that slot. 1 Thus, for the top 100 companies the competition is implicitly for the 100 top quality executives to fill those slots, although, as Yarrow (1972) emphasises, some trade-off between expense and quality is to be expected. In most labour market situations it is necessary to consider the elasticity of supply with respect to the real wage on offer. Failure to offer a sufficiently high wage will result in the individual deciding that he or she is better off not accepting paid employment and either consuming leisure, engaging in home-production or some combination ofthe two. But at this top end of the labour market such considerations are not much in evidence. Save in the case of voluntary retirement, the executives in question are not interested in whether or not paid employment is worth their while--there is little doubt of that. Similarly, in terms of alternative occupations, these individuals have highly specialised portfolios of human capital. They are unlikely to find employment in a different field of work as productive. The major factor conditioning the supply of executives of a given quality into the CEO level at any particular company is the opportunity cost, i.e. what could be earned in alternative companies in similar positions. The market for CEOs is a relatively thin market and notions of 'the going rate' do not make much sense. Nevertheless, widespread use is made of survey data provided by remuneration consultants. For any one company, this approach tends to make of critical importance the CEO compensation packages available at some dozen or so comparator companies. In such circumstances, perceived upward movements in levels of pay can easily become self-fulfilling beliefs as companies move to match 'the market' and, in so doing, actually change the market equilibrium. Cross-Atlantic comparisons must be taken as an important ingredient in this pay process. As emphasised above, the market for CEOs is not a high-volume market. Each situation is different. Each service contract awarded is the result of a unique round of negotiation between the focal CEO and the Board, or its representatives. And in this negotiating process the bargaining strength of the CEO will rest to some extent on the level of pay awards at comparable companies both here and in the USA. As always, in negotiation the 'best alternative to a negotiated agreement' (BATNA) is of paramount importance. To estimate the market value of a UK-based CEO in the USA, it is necessary to standardise for organisational and personal characteristics. To compare what an 1 The recent trend to split the top management job into a Chairman and CEO creates essentially two jobs to fill unless the posts are utilised as part of the corporate succession process as Vancil (1987) describes the practice in the USA. In the USA roles of CEO, Chairman, and President are often structured as distinct stages in the succession process from grooming through induction to retirement. A comparison of top executive pay in the UK and the USA 117 industrial worker could earn in the USA versus the UK1 would require some allowance for the experience of the individual (as proxy for the human capital involved) and for the nature of the work, so for similar reasons we will allow for the particular circumstances of the CEO. In terms of human capital, this is done by controlling for the length of service with the firm and the length of time in office as CEO. The nature of the CEO's job is captured by a size measure of the firm (sales) and a firm-performance measure (stock market return in the previous year). It is worth noting that top executive pay rose dramatically on both sides of the Atlantic during the 1980s, at a time when in both countries' rates of personal taxation were being slashed for these very same highly paid individuals. In the USA top rates of Federal taxation fell from 50% in 1981 to 33% after 1986. In the UK the fall was even more marked, dropping from 83% in 1979 to 40% after 1988. These and other changes occurring at the time may well have altered what was regarded as a fair rate of pay for a CEO. After a period in which there had been severe pay restraints,2 and at a time when notions of the 'enterprise culture' were being popularised and CEOs were being encouraged to see themselves more as entrepreneurs than as 'organisation men', some upward shift in CEO pay levels was to be expected. Within this context, transatlantic comparisons may assume an important part in the feedback mechanism that seems to drive the process of top executive pay determination. This paper utilises two data sets that contain detailed information on the pay of top executives in the UK and in the USA. The nature of this data is discussed in Section 2 of the paper, and a series of comparative analyses is presented in Section 3. The paper concludes with a discussion of the policy implications of the results. 2. Description of the data The data on UK executives come from a sample utilised in Main and Johnston (1993) which was based on two overlapping sample frames that aim at capturing both large and high-paying companies operating in the UK. The companies considered for inclusion are those in the top 500 in 'Britain's 10,000 largest Companies in 1991' and the top 500 companies in the Charterhouse (1990) 'Top Management Remuneration Sample for 1989/90'. Companies were considered for inclusion if they appeared on either list and were also available on Datastream. The sample was also restricted to UK companies with listings on the London Stock Exchange. These considerations and the problems of incomplete data reduced the sample size to 220. Most large publicly held UK firms are included. 1 Even such general comparisons are difficult to make owing to the differences in coverage of the two sides of the Atlantic. Thus, for the USA the Monthly Labor Review reports (May 1991, table 21) that the average hourly earnings in 1989 of non-farm private-sector production or non-supervisory workers was $9.66. For the same year the Department of Employment Gazette reports Ouly 1992, table 5.6) the average hourly earnings over all industries and services of a manual worker as £4.59. The average exchange rate for the year (Economic Report of the President, 1992, table B107) was $1.6382 per £1. This suggests a US advantage of 28%. But the UK figures cover only adults and exclude those whose earnings were affected by sickness in the week of the survey. 2 Almost the entire 1970s in Britain could be viewed as one long period of incomes policy and pay restraint. When the maximum pay increase allowed was £6 per week and inflation was running around 20%, it is easy to see the severe erosion that top executive pay could suffer. 118 Brian G. M. Main et al. Datastream is utilised to provide data for 1990 on company size (measured by sales or turnover) and on CEO remuneration. UK reporting requirements in the area of top executive pay are extremely modest in comparison with the USA, although recent proposals from the Cadbury Committee (1992) will soon result in some improvement in this area. At the moment the principal disclosure requirement in the area of executive remuneration originates in the 1967 Companies Act (currently in the form of Schedule 4 of the 1989 Companies Act). This demands that the company annual report disclose the emoluments (essentially pay plus bonus plus the cash-equivalent of any perquisites but excluding pension contributions) of the Chairman and (if a different person) the highest paid director. In what follows, the highest level of pay reported is used and discussed under the label of 'Chief Executive Officer' (CEO) pay, although it may, in fact, be that of the chairman or of someone who combines both roles of chairman and CEO. Information on the length of service of the CEO was taken from the company's annual report. But this information was surprisingly sparse from this source. Reference to Hambros' Corporate Register and W'ho's W'ho in Industry improved the situation somewhat, but these variables remain as unknowns in a number of instances. Personal descriptors of tenure and length of service are used only in cases where we felt we could identify the person who was earning this level of pay. These considerations reduce the sample size at certain stages of the analysis that follow. The data for the executives in the USA come from the 'United Shareholders Association 1000' data base as developed by Crystal (1990). This provides a wide coverage of companies based in the USA. By careful scrutiny of the proxy statements of each of these companies, Crystal has managed to break down the pay of each top executive into various components. In addition to the usual 'total current compensation' (TCC, base plus bonus), it is also possible to arrive at the amount received by the CEO in the form of stock options, restricted stock, performance units and other forms of compensation. Added to the measure of total current compensation these provide an aggregate that is labelled 'total direct compensation' (TDC). Stock options, performance units and restricted stock are valued at time of issue. Thus the value of the stock options an executive receives is computed by use of the Black-Scholes (1973) option-pricing formula which estimates the worth of such options (with a given strike price and restriction period before exercise) as if purchased on the open market. Owing to the lumpy nature of such awards an attempt is made to smooth these data by using a three-year average. The compensation data extracted from proxy statements is supplemented by use of data on sales and stock market performance taken from Compustat. Thanks to SEC regulations, information on the length oftenure ofthe CEO in his/her post and the length of service with the company are always available on the proxy statement. Table 1 provides an overview of the data with an emphasis on its comparative nature. The pounds-sterling and dollar figures are converted at an approximate 1990 average exchange rate1 of $1.80 per £1. The US data correspond to financial years ending in the spring of 1989 and the UK data pertain to accounting year ends 1 Table B 107 of the 1992 Economic Report of the President records the following average exchange rates in$/£: 1989=$1.6382; 1990=$1.7841; 1991=$1.7674. The $1.80 rate used in this paper, therefore, favours UK salaries over this period. A comparison of top executive pay in the UK and the USA 119 Table 1. Average sample characteristics Characteristic Executives in UK All Subset Executives in US Current paya £ $ Total pay £ $ Value of share holdings £ $ Sales £ $ Market return Current tenure (years) Current service (years) 266,310 479,360 n.a. n.a. 5.5m 9.9m 1,874m 3,373m -0.065 n.a. n.a. N=220 276,000 496,800 n.a. n.a. 2.3m 4.lm 2,112m 3,802m -0.067 5.9 19.0 N=132 437,730 787,920 578,055 1,040,500 16.0m 28.8m 1,801m 3,242m 0.064 9.3 22.9 N=915 a$1£ conversions done at $1.80 per £1. in 1990 (with an even split between spring and autumn year-ends). While the timing is not perfectly coincident, it is, hopefully, close enough to provide a meaningful· insight to the relative magnitudes of the levels of pay involved. In terms of average levels of total current compensation (base plus bonus), pay levels in the USA seem to be some 64% higher than those in the UK (£437,730 vs. £266,31 0). Owing to the lack of comprehensive data, it is not possible to comment meaningfully on the total compensation package in the UK, but total direct compensation raises the average compensation in the USA sample to the equivalent of some £578,055. The impact of the long-term components of compensation (stock options, performance unit grants, and so on) is thus significant, but less dramatic than has been seen in certain well-known cases.l US executives are also seen to hold a greater value of their own company's shares (£5.5 million in the UK versus the equivalent of £16.0 million in the USA). Such personal equity holdings are ignored in the analysis that follows on the grounds that they are generally the result of a personal investment decision rather than a form of executive compensation.2 In terms of size as measured by turnover, the sample averages are very close, both being around £1.8 billion. The stock market return recorded by the two national samples of companies is markedly different, owing in part to the slightly different timing of the observations. In the USA the average return was 6.4%, while in the UK the average return was a more disappointing - 6.5%. In the multivariate 1 See Crystal (1991) for a painstaking analysis of some of these cases, for example, Stephen Ross at Time-Warner. 2 These holdings are distinct from grants of restricted stock which are valued as a component of long-term compensation for the US executives. 120 Brian G. M. Main et al. analysis utilised later, estimates of within-country regressions produce extremely small empirical linkages between CEO pay and the performance of each company in the stock market. This difference in average outcomes is not, therefore, thought to be of vital importance. The second set of entries in the UK column of Table 1 refer to that part of the sample for whom it was possible to establish personal details concerning company and job tenure. The impact on the representativeness of the sample of the atrophy that occurs here seems to be modest in terms of the level of pay and the size of the companies involved (which are little changed). Where the impact is most noticeable is in terms of the average value of share holdings which drops markedly. This suggests that the executives for whom this type of information is difficult to find (under Britain's voluntary reporting system) are possibly those entrepreneurial CEOs who started up their companies and who hold large amounts of equity. For those for whom this personal information is available, the UK level of tenure as a CEO seems much shorter than in the USA, standing at 5.9 years vs. 9.3 years in the USA. On the other hand, average time with the company seems broadly similar, with the UK average of 19.0 years coming close to the US figure of 22.9 years. In part, the short tenure of office in the UK could be a result of eliminating the entrepreneurial types as mentioned above. With this brief descriptive overview of the data complete, we now turn to some multivariate analysis in an attempt to control for the different characteristics and circumstances of the UK vs. the USA sample. 3. Comparative analyses: UK vs. USA The general approach adopted in this section is to estimate, for both the UK and the USA sample, the level ofCEO pay as a function ofthe size ofthe company (sales), its performance (stock market return), the length of time the CEO has held the job (CEO tenure) and the length of time the CEO has been with the company (CEO service). Once these relationships have been estimated, it is possible to ask the counterfactual question of what the typical top executive in a British firm would be paid in a comparable firm in the USA. The equivalent question can also be asked concerning the pay prospects of the typical CEO in the USA with respect to a comparable British company. As use of CEO tenure and CEO service reduces the UK sample rather dramatically, the first column of Table 2 presents the pay equations estimated on sales and stock market return only. It may be noted that there is no great shift in these estimated coefficients when the analysis is extended in the second column of Table 2 to include tenure descriptors available for the subsample. This then gives some confidence in using the subsample of 132 chief executives on which to base our comparisons. The dependent variable is the logarithm of total current compensation, and the size variable (sales) is also in logarithms. It can be seen that, for both the UK and the USA, pay is significantly related to sales with an elasticity of around 0.3. This result is consistent with many other studies of top executive pay and comes close to being regarded as a universal constant (see Baker et al., 1988; Kostiuk, 1990; and Rosen, 1992). A comparison of top executive pay in the UK and the USA 121 Table 2. Pay of chief executives in the UK (standard errors in parenthesesa) Descriptive variable UK Current Pay (base plus bonus, £) USA Current Pay (base plus bonus, $) Constant Sales Return CEO Tenure 10.3090b (0.1509) 0.2981b (0.0224) 0.1033 (0.0847) CEO Service Adjusted R2 F-statistic Sum of squared residuals Standard deviation of residual N 0.45 89.32 39.07 0.4243 220 10.1420b (0.2296) 0.3244b (0.0332) 0.1222 (0.1006) 0.0191b (0.0073) -0.0075b (0.0034) 0.46 29.36 20.72 0.4039 132 11.2840b (0.0806) 0.2921b (0.0111) 0.0974 (0.0815) 0.41 319.77 200.20 0.4685 915 11.2090b (0.0833) 0.3058b (0.0117) 0.0868b (0.0802) 0.0105b (0.0026) -0.0052b (0.0016) 0.43 170.51 194.67 0.4625 915 ·White standard errors bSignificant at a=O.OS. Stock market return plays no statistically significant role. An insignificant or empirically modest effect of company performance on executive pay is, in fact, typical of the findings ofMurphy (1985), Coughlin and Schmidt (1985), Jensen and Murphy (1990), Gibbons and Murphy (1992) and of others who have examined this issue in the USA. Results for the UK by Cosh (1975), Gregg et al. (1993), Conyon and Leech (1993), Conyon and Gregg (1993) also find a dominant size effect and an empiri~ally weak performance effect. In terms of personal characteristics, it is seen that both CEO tenure and CEO service are significant. In both the UK and the USA, the longer the CEO has been in office the higher is his/her level of pay. This may be thought of as a return to human capital as the executive gains increasing expertise in the job. Alternative explanations related to the capture of economic rents also suggest themselves, of course. What is less easily reconciled with standard economic theory is the significant negative relationship between the length of time the executive has been with the company and his/her level of pay. Notions of human capital suggest that this should be a positive relationship. The result may well be an indicator of those executives who have moved to their present company relatively recently, possibly even directly into the CEO job. Such individuals may require higher salaries than long-serving employees owing to the uncertainty about the new employment situation that confronts any potential mover. Loss of pension rights, and so forth, are also manifestations of the costs of such career moves which may be compensated through higher pay in the new job. Alternatively Faith et al. (1984) draw attention to the rent-seeking activity of executives, and explain how executives with long company service are more likely to be benefitting from unmeasured rents in addition to their measured income. 122 Brian G. M. Main et al. The broad agreement between the estimated coefficients in the UK and the USA regressions ofTable 2 should be noted. This similarity has already been reported by Kostiuk (1990) using US data for 1969-71 to compare with Cosh's UK results in terms of size (log of assets) and performance (return on assets). In Table 2 it can be seen that for our samples the estimated coefficients on size (log of sales) and performance (stock market return) are very close. The returns to years of tenure in office and years of service with the company while varying somewhat between countries are nonetheless dimensionally similar for both the UK and the USA. An important difference between the two sets ofestimates rests with the constant terms. Owing to the logarithmic specification of the dependent variable, this implies a proportional difference in the levels of remuneration between the two countries. Part of this difference is due to the measurement in pounds-sterling as opposed to dollars. This accounts for 0.5878 of the difference. But the remainder is due to unmeasured differences between the two countries. These suggest a raw difference in pay of some 62%.1 Before using the results of Table 2 to compute the UK/USA comparisons outlined above, it is interesting to probe the richer US pay data a little further. Table 3 presents an equivalent pay equation to Table 2, but this time, in columns 1 and 2 of Table 3, the dependent variable is the logarithm of total direct compensation. In the subsequent columns 3-6 of Table 3, the shares of this total direct compensation falling respectively as current compensation (base plus bonus), share options, restricted stock and performance units is estimated. For these equations the dependent variable is the fraction (between 0 and 1) of total direct compensation that is received in each form. As heteroscedasticity is likely to be a problem here, it is important to note that the reported standard errors are White (1980) standard errors which correct for bias that might contaminate statistical inference in the presence of heteroscedasticity. Table 3 shows that in terms of total compensation, the relationship is similar to that reported in Table 2. Empirically sales, tenure and service each has a bigger impact (it is a numerically larger dependent variable), but the message is the same: well-paid executives work for large companies and have been in office for a long time but have not necessarily worked for the company much longer than that. In columns 3 to 6 the shares of TDC falling under the various headings is seen to be significantly influenced by sales. Large companies pay a lower proportion in current compensation and a greater proportion in performance-contingent forms such as stock options. This makes sense when one allows for the correlation between size and level of pay. It is clearly far easier to persuade an executive to accept an additional component of his pay in this contingent form when he has already established a high level of'cash' pay. It is simply less risky. For a given time in office, 1 The constants in columns 2 and 4 of Table 2 are 10.1420 and 11.2090 respectively. Log(l.8) =0.5878 of this is due to differences in currency units used. The remainder (0.4792) represents the proportional difference in pay. This suggests a 62% (exp(0.4792)= 1.615) difference between the two pay equations in the up direction or a 38% (exp(- 0.4792)=0.619) difference in the down direction. For 1989 the Statistical Abstract of the United States 1992, table 137 reports a GNP per capita of $14,580 for the UK and $20,910 for the USA, i.e. 43%. Table 3. Components of pay of chief executives in the USA (standard errors in parenthesesa) Descriptive variable Total direct compensation Share of direct compensation in the form of Current Share Restricted Performance pay options stock units Constant Sales Return CEO Tenure CEO Service Adjusted R2 F-statistic Sum of squared residuals Standard deviation of residuals N "White standard errors hSignificant at u=0.05. 11.2510b (0.0973) 0.3239b (0.01357) 0.0340 (0.0950) 0.37 266.25 291.95 0.5658 915 11.1890b (0.0983) 0.3466b (0.0145) 0.0387 (0.0936) 0.0127b (0.0026) -0.0095b (0.0020) 0.39 146.94 280.95 0.5556 915 1.0149b (0.0342) -0.0297b (0.0048) 0.0258 (0.0312) -0.0008 (0.0010) 0.0025b (0.0006) 0.05 12.96 27.21 0.1729 915 0.0478 (0.0288) 0.0111b (0.0039) -0.0223 (0.0259) 0.0002 (0.0008) -0.0016b (0.0005) 0.02 5.64 16.30 0.1338 915 -0.0293 (0.0196) 0.0090b (0.0028) 0.0080 (0.0143) 0.0007 (0.0005) -0.0005 (0.0003) 0.01 3.77 8.37 0.0959 915 -0.0291b (0.0111) 0.0082b (0.0021) -0.0013 (0.0092) -0.0002 (0.0002) -0.0003 (0.0003) 0.02 5.30 4.75 0.0722 915 > n ~ ~. I'll =.00.., .g ~ =na. ~ 'tl ~=... ~ ~ l ;. R ~ > .... N ~ 124 Brian G. M. Main et al. Table 4. 'Over here, over there' CEO pay comparisons (i) Average US executive: actual in US (£) if in UK (£) Current pay 437,730 317,750 (ii) Average UK executive: Current pay Total direct compensation of which: share options restricted stock performance units if in US($) 978,769 1,318,846 0.1116 0.0390 0.0317 actual in UK ($) 496,800 n.a. n.a. n.a. n.a. long company service executives not only receive lower levels of pay (see column 2 ofTable 3) but take a higher proportion of it in cash and a lower proportion in stock options. All of these results are put in context of the UK/USA comparison in Table 4. In part (i) of this table, the average characteristics of the US CEO are used with estimated coefficients from Table 2 (column 2) to compute his 'expected' pay at a UK company with the same size and performance characteristics as his US company. 1 It can be seen that this turns out2 to be £317,750 as compared to his actual pounds-sterling equivalent of £437,730. In other words, he would be some 27% worse off in terms of current pay (TCC) alone. Part (ii) of Table 4 computes the expected experience of the average UK executive if he were paid in the way that the US CEOs are remunerated.3 Using the coefficients from column 4 of Table 3, this procedure predicts4 a level of current compensation of $978,769. Compared to his actual pay plus bonus in the UK of $496,800 (the dollar equivalent of £276,000), this would leave him some 97% better off. 1 The average characteristics of the US executives are taken to be as indicated in Table 1, namely, average sales of $3,242 million (£1,801 million), stock market return of+ 6.4%, current tenure in office of 9.3 years, and current service with company of 22.9 years. 2 Given that sales are measured in logs, the predicted log(£) pay in the UK is: 10.1420+0.3244 X (7.4961) +0.122 X (0.064) +0.0191 X (9.3)- 0.0075 X (22.9)= 12.58745. This implies a pay level of exp(12.58745) =£292,860, but as the regression was in logs and the implicit assumption one of a log-normal distribution for pay, Greene (1990, p. 168) reminds us that the above will be the median rather than the mean. To obtain the mean estimate it is necessary to multiply the above result by exp(s2/2) where s is the standard error of the regression. In this particular case s=0.4039, and the consequent estimate of the mean level of pay is £31 7,750. 3 The average characteristics of the UK executive are taken to be as indicated in Table 1, namely, average sales of £2,112 million (£3,802 million), stock market return of - 6.7%, current tenure in office of 5.9 years, and current service with company of 19.0 years. 4 Given that sales are measured in logs and the average UK sales (£2,112 m) measured in $s is $3801.6 m, the predicted log($) pay in the USA is: 11.2090+0.3058 x (8.24318)+0.0868 x (- 0.067) +0.0105 x (5.9)- 0.0052 x (19.0) = 13.6871, and, with the adjustment from median to mean as described on p. 124 in footnote 2, the predicted level of pay $978,769. A comparison of top executive pay in the UK and the USA 125 In addition, it is possible to predict the typical UK top executive's total direct compensation at the head of a US corporation by using the estimated coefficients from column 2 ofTable 4. This predicts a TDC of$1,318,846 (or some £732,692 at the exchange rate of $1.80 per £1 assumed here). The difference between the TDC and the TCC is shown in Table 4 (using the results from Table 3) to include an expected 11% in stock options, 4% in restricted stock and 3% in performance units. Under this regime TCC accounts for some 81% of the total compensation package. 1 There are no reliable data on UK long-term compensation that permit the reverse comparison with the UK situation. 4. Summary and policy discussion Using two samples oftop executives, one from the UK and one from the USA, this study has attempted to gauge the difference in pay that a top executive in a British company would experience if he could move to a broadly comparable post at the head of a company in the USA. The results suggest that in terms ofcurrent pay (pay plus bonus) the British executive could expect to see his earnings increase by almost 100% (some 97% to be precise). His pay would rise from £176,000 ($496,800) to $978,769. In addition his yearly compensation in terms of stock options, performance units, restricted stock, etc., would bring his total annual compensation up to $1,318,846. Repeating the experiment in the opposite direction, the average CEO in a company in the USA would, if paid in a way typical of British top executives, receive some £317,750 in current pay rather than the $787,920 (£437,730) that he can expect as current compensation in the USA. This represents a fall of some 27%. There is no comparable British data on the components of long-term compensation to allow us to predict what would happen on this dimension. Given the size of the pay differences established above, one would expect that, in the absence ofbarriers, labour mobility would have the effect of closing the gap. But although immigration restrictions present no serious barrier at this level of job, the very specificity of the type of work in terms of the different legal, social and political frameworks within which enterprises operate in the two countries does mitigate against serious international mobility of labour. Indeed, the relatively high rate of pay in the' USA merely serves to underline the extent to which remuneration is demand-determined at this level. As was outlined in the introduction above, the top companies are competing for the very best people to fill these slots. Adding more potential hires to the pool does little to change that situation. One thing that should be emphasised here is that what is being measured above is not a difference in the design of pay (incentive pay schemes, payment by results, etc.), but that the measures above reflect average levels of pay. Differences in pay design will only affect average levels of pay to the extent that they raise productivity levels. 1 There is some difference in the predicted TCC between the method in Table 2 of estimating it directly, and the method of Table 3 of estimating it indirectly as a predicted share of predicted TDC. The first method predicts $978,769 and the second predicts $1,069,754. The difference between the two estimates seems small enough to be put down to statistical imprecision. 126 Brian G. M. Main et al. From a policy perspective it is worth noting that the USA already has much stricter disclosure and governance requirements than apply to the UK. The SEC requires extremely detailed disclosure of all the components of pay on the top five named officers of the corporation. By contrast, even under the new Cadbury (1992) proposals the level of disclosure in Britain concentrates on the chairman and the highest paid director (if different). Although companies will now be required to separate out the base pay from bonus payments for these individuals (and for the board as an entity), there is no move to require fuller evaluation of long-term components of pay such as stock options. In a similar vein, compensation committees (remuneration committees in the UK) are more universally in place in the USA. In the UK they have only become fashionable in the last five years or so, although they will become a board requirement under Cadbury. And non-executive directors are still in a minority on most British boards, in marked contrast to their dominant position on US boards. These facts suggest that the constraint of top executive pay is not achieved through corporate governance requirements. In fact Main and Johnston (1993) actually suggest that sound structure of corporate governance may give a company the confidence and legitimacy that it might otherwise lack to grant high pay awards to its CEO. Furthermore, O'Reilly et al. (1988) demonstrate that complex social forces may well condition the decisions of a remuneration committee in ways that are not consistent with pay restraint. We have seen that the level of CEO pay is significantly higher in the USA than in the UK. In the realm oftop executive pay, corporate practice in terms ofgovernance and disclosure is significantly more comprehensive in the USA than it is in the UK. It would seem wise, therefore, not to place too much faith in stricter corporate governance or disclosure requirements as a way of moderating pay at the top of the corporation. Similarly, however, it would not seem that market forces are particularly strong in this area. Shareholder activism may hold the answer, but the transaction and co-ordination costs involved will very often inhibit any effective action on this dimension. The above discussion offers no clear policy prescription for constraining top executive pay. At a more parochial level, however, the estimates presented above do suggest that British top executives who can claim to be internationally mobile have ample bargaining strength in negotiating the terms oftheir remuneration. From this perspective, pay is indeed less 'over here' than 'over there'. Bibliography Baker, G. P., Jensen, M. and Murphy, K. 1988. Compensation and incentives: practice versus theory, Journal of Finance, vol. 43, July Black, F. and Scholes, M. 1973. The pricing of options and corporate liabilities, Journal of Political Economy, vol. 81, 6737-54 Cadbury, A. 1992. Committee on the Financial Aspects of Corporate Governance, London Conyon, M. and Gregg, P. 1993. Pay at the Top:. A Study of the Sensitivity of Chief Executive Remuneration to Company Specific Shocks, National Institute ofEconomic and Social Research, mimeo Conyon, M. J. and Leech, D. 1993. Top Pay, Company Performance and Corporate Governance, Warwick Economic Research Papers No. 410, Department of Economics A comparison of top executive pay in the UK and the USA 127 Cosh, A. 1975. The remuneration of chief executives in the United Kingdom, Economic Journal, vol. 85, March Coughlin, A. and Schrnidt, R. 1985. Executive compensation, management turnover, and firm performance: an empirical investigation, Journal of Accounting and Economics, vol. 7, no. 2 Crystal, G. S. 1990. The Crystal Report, vol. 2, no. 2 Crystal, G. S. 1991. In Search of Excess. The Overcompensation of American Executives, New York, W. W. Norton Faith, R. L., Higgins, R. S. and Tollison, R., D. 1984. Managerial rent and outside recruitment in the Coasian firm, American Economic Review, vol. 74, 660-72 Gibbons, R. and Murphy, K. J. 1992. Optimal incentive contracts in the presence of career concerns: theory and evidence, Journal of Political Economy, vol. 100, no. 3 Greene, W. H. 1990. Econometric Analysis, New York, Macmillan Gregg, P., Machin, S. and Szyrnanski, S. 1993. The disappearing relationship between directors' pay and corporate performance, British Journal of Industrial Relations, vol. 39, 1-10 Harrison, ]. R. 1987. The strategic use of corporate board committees, California Management Review, vol. 30, 109-25 Jensen, M. C. and Murphy, K. J. 1990. Performance pay and top-management incentives, Journal of Political Economy, vol. 98, 225-64 Kostiuk, P. F. 1990. Firm size and executive compensation, Journal of Human Resources, vol. 25, no. 1 Main, B. G. M., and Johnston, ]. 1993. Remuneration committees and corporate governance. Accounting and Business Research, vol. 23, no. 91A Murphy, K. J. 1985. Corporate performance and managerial remuneration: an empirical analysis, Journal of Accounting and Economics, vol. 7, 11-42 O'Reilly, C. A. Ill, Main, B. G. M. and Crystal, G. S. 1988. CEO compensation as tournament and social comparison: a tale of two theories, Administrative Science Quarterly, vol. 33, June Rosen, S. 1992. Contracts and the market for executives, in Werin, S. and Wijkander,]. (eds) Contract Economics, Oxford, Basil Blackwell Vancil, R. F. 1987. Passing the Baton, Boston, Harvard Business School Press White, H. 1980. A heteroscedastic-consistent covariance estimator and a direct test for heteroscedasticity, Econometrica, vol. 48, 817-38 Yarrow, G. 1972. Executive compensation and the objectives of the firm, in Cowling, K. (ed.) Market Structure and Corporate Behaviour, New York, Gray-Mills Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource 'Over here and over there': a comparison of top executive pay in the UK and the USA Creator An entity primarily responsible for making the resource Brian G. M. Main Charles A. O'Reilly III Graef S. Crystal Source A related resource from which the described resource is derived International Contributions to Labour Studies 4, pages 115-127 Date A point or period of time associated with an event in the lifecycle of the resource 1994