We turn now from the labor market to the relationship between the corporation and that other factor of production, capital. In the previous chapters, we have traced changes in the American system of corporate governance and Wnance: in the late nineteenth and early twentieth centuries, large American corporations were formed, and initially con-trolled, by the guiding hand of investment banks; they emerged gradually from bank control until, during the Fordist era, they were essentially controlled by their managers;
since the late 1970s, they have existed under the constant scrutiny of Wnancial markets.
11.4.1. CONTROL BY OUTSIDERS, OR QUASI-INSIDERS?
America’s system today is the archetypal LME Wnancial system. In most large com-panies, shareholding is widely dispersed so that shareholders cannot exercise direct control. Shareholders do, however, assert the right to control. This assertion is evidenced in the rhetoric of shareholder value; in institutions such as the market for corporate control, disclosure requirements, and other legal protections for the rights of minority shareholders; and in very high levels of executive pay.
Outside of the LMEs (i.e., both in CMEs and also in a host of countries that don’t really fall in either category), most companies are eVectively controlled by some combination of insiders (managers) and what I will call quasi-insiders: founding families, shareholders who have bought large blocs for strategic reasons, banks with which the companies do business. The Japanese keiretsu system described in Chapter 10 is one example of this, but should not be regarded as typical – quasi-insider control looks very diVerent in diVerent countries. Germany lacks keiretsu-like groups, but does have widespread cross-shareholding between companies. In Germany, banks may also own shares in their customers; this, together with their role as trustees managing (and voting) the shares of others, can give them considerable inXuence as shareholders; on their inXuence as lenders, more in a moment. In some countries – Sweden, South Korea, and South Africa, for instance – a large share of industry is grouped into a handful of corporate pyramids, which are controlled (often by one family) from the top (LaPorta, Lo´pez-de-Silanaes, and Shleifer 1999); another example of such pyramids, of course, is the old zaibatsu system in Japan, abolished by the US occupation authorities after World
164 THE GLOBAL ENVIRONMENT OF BUSINESS
War II. While these mechanisms diVer, what they have in common is that the inXuential shareholders are strategic shareholders, who tend to hold their shares for very long periods, and who typically own them not as a passive portfolio investment but as part of a more complicated business relationship with the Wrm.
Banks in LMEs play a relatively passive role in the aVairs of their customers. In many other countries they are inclined to take a far more active role, and should be regarded as a second category of quasi-insider, along with strategic shareholders (in Germany, as just noted, the two can be the same). Indeed, before Hall and Soskice’s the LME/CME distinction became current, a common way of classifying capitalist economies was in terms of whether their Wnancial systems were ‘‘bank centered’’ or ‘‘stock-market cen-tered’’ (Franks and Mayer 1990). It turns out that all the CMEs fall into the former category, and the LMEs into the latter. Today, more attention is focused on the role of block-holders than of banks, but the banks in CMEs play a role which parallels, and complements, that of strategic shareholders.
11.4.2. PROTECTION OF MINORITY SHAREHOLDERS AND MINORITY CREDITORS
Why do quasi-insiders exercise such inXuence in most countries and not in LMEs? One answer is that in most countries it is more common for a single shareholder to hold a controlling interest in a large company (this is not so in Japan, where in statistical terms shareholding is fully as dispersed as in the US, but in Japan eVective control resides with banks and cross-shareholding networks, so we have control by strategic investors even if they don’t own large blocs). That answer just pushes the question back: why, outside the LMEs, are controlling interests likely to be held by a few strategic investors? We also need to understand how quasi-insiders are able to exercise control.
LMEs tend to have strong provisions for minority shareholder protection (MSP).
MSP is any application of the principle that all holders of common shares in a company should be treated equally. This principle can be reXected in a number of diVerent legal or regulatory requirements, and in corporate practice. Gourevitch and Shinn (2005, p. 48) prepared an index of MSP for diVerent countries; this included the provision for getting information about the company (both through company disclosure and through independent audits), oversight (which they treat as a function of the propor-tion of independent members on the board of directors), control (the extent to which insiders and strategic investors are prevented from interfering with the market for corporate control), and incentives (the extent to which executive compensation is structured to align executives’ interests with those of shareholders). With the same caveat about adding apples-and-oranges that I gave for Estevez-Abe et al. (2001) indices of
employment and unemployment protection, an extract from Gourevitch and Shinn’s index is reported in Table 11.2. Notice that countries with the strongest MSPs are former British colonies: legal traditions have an eVect. The two former British colonies that come in relatively low on this list – India and South Africa – had vigorous post-independence regimes of import substitution industrialization (ISI), and their corpor-ate governance was reshaped in that period.
Countries with strong MSP also tend to be countries where it is uncommon for a single shareholder to have a controlling interest (Gourevitch and Shinn 2005, Figure 3.3). One way of explaining this empirical relationship is that shares are more valuable to those with controlling blocks when MSPs are weak, because with weak MSPs large blockholders can use the company’s resources to help themselves at the expense of minority shareholders.
Another interpretation is that in return for the privileges they enjoy when MSPs are weak, large block-holders provide a service: monitoring the performance of the Wrm’s managers, and step in to sort things out if the company is doing badly. An investor does not want to hold a large, undiversiWed block of a company’s shares unless she can monitor and intervene in this way, but since it is costly to play this role she would not be willing to do so if she had to share the beneWts equally to a lot of free riding minority shareholders.
The diVerence in the role played by banks has a similar explanation: small creditor protection (SCP). We don’t have information about such a wide range of countries as in the case of MSP, but Frankel and Montgomery (1991) provide a comparison of the US, Britain, Germany, and Japan – two LMEs and two CMEs. We saw in Chapter 9 that American and
Table 11.2. Minority shareholder protection index for selected countries
Source: Gourevitch and Shinn (2005), Table 3.1.
166 THE GLOBAL ENVIRONMENT OF BUSINESS
British banks take care not to be seen as managing a company, because if they did and if the company subsequently went bankrupt, a court would treat them as an owner, not a lender:
all other creditors – suppliers, for instance – would be repaid before the bank. Therefore, whatever information the bank might have about the way the company is managed, and whatever worries it may have about being repaid, it maintains a strictly arm’s length relationship with the company. The bank’s only real sanctions are to refuse further credit or to call in loans outstanding. It will have larded the loan agreement with covenants detailing circumstances – most commonly, balance sheet benchmarks – which would place the borrower in default, giving bank the option of calling in the loan. The German or Japanese bank has a wider range of intervention options because minority creditor protections are weaker. Because they can intervene, they also have an incentive to monitor the company more closely – the information is more useful to them.
Our story so far, then, is that quasi-insiders take strong roles in companies, except in the LMEs. They do so because, except in LMEs, MSP and SCP are weak. But (at the risk of sounding like a broken record here), why are MSP and (as far as we know from a few important cases) SCP weak in so many countries?
Mark Roe (2003) oVers a functional explanation: the countries that have weak MSP are those that need weak MSP, and they need it because employment protection is strong. They have strong employment protection not for functional reasons, but from the vagaries of politics. He assumes that quasi-insider monitoring is more costly, overall, than monitoring conducted through Wnancial markets. He also assumes that it is more eVective. Strong employment protection accentuates the principal-agent problem be-tween shareholders and managers, since managers not looking out for the shareholders’
interests can easily allow the Wrm to get locked into a high-cost workforce. Therefore, countries with strong employment protection need strong quasi-insiders. Weak MSP is an institutional solution to this problem. Comparing Table 11.2 and Figure 11.1, we do indeed see that, of the countries included in both, those with the strongest employment protection (Austria, Belgium, Germany, Italy, and Japan) are also those with the weakest MSP. Those with the weakest employment protection and the strongest MSP are, of course, the LMEs. Gourevitch and Shinn (2005) noted this relationship and a similar strong correlation between employment protection and the prevalence of controlling blockholders (see Gourevitch and Shinn 2005, Table A.13).
Roe assumes that market governance is less costly than quasi-insider governance and will be more eYcient when agency costs are lower; that employment security raises agency costs; and that governments will act to adopt eYcient systems of corporate governance. It is possible that all of these are true, but Roe does not demonstrate them: they are assumptions he uses to explain some empirical correlation. Gourevitch and Shinn have a diVerent explanation for the same correlations, based on the constitutional structures of the countries in question. I will return to that below. First, let us consider some of the practical implications that quasi-insider control has for corporate governance.
11.4.3. WHAT DO QUASI-INSIDERS DO?
The eVects of quasi-insider governance are in the eye of the beholder: it has its fans, and its detractors. Let us consider two aspects – the governance function (monitoring and intervention), and the eVect on the character of investment.
Quasi-insiders can keep a close eye on a company, can intervene early if things are going badly, and have a broad range of intervention tools rather than a few crude clubs.
If we assume that the intervention is an eYcacious response to problems that are causing avoidable losses, then earlier, gentler intervention seems better than waiting for the company to lose much of its value before it is reorganized via the market for corporate control. Of course, some part of our assumption might be incorrect.
There are at least three reasons why quasi-insider control is likely to produce diVerent investment choices than market control: information, time horizon, and attitude toward risk. Quasi-insiders get information from close monitoring; markets don’t have a visible monitoring mechanism, but market prices are based on some information – including, plausibly, things the inside monitors might have missed. Takeo Hoshi, Anil Kashyap, and David Scharfstein (1991) test the eVects by comparing the investment behavior of Japanese companies that are part of a keiretsu, and those which are independent. They Wnd that the investment behavior of Wrms which are independent is inXuenced by their cash position, while that of the keiretsu members is inXuenced by the Wrm’s proWtability. This is evidence that a keiretsu – a network of companies controlled by quasi-insiders – does a better job of channelling its members’ investable funds to proWtable uses, than Japanese Wnancial markets do for independent Wrms. Depending on your perspective, that says something good about quasi-insider control, or something bad about Japanese Wnancial markets.
Quasi-insider systems are often said to have, patient capital, on the grounds that strategic investors take a long view, while stock markets look at the most recent Wnancial report. This is as much about information as patience: minority shareholders often have no information about a company other than what is in its Wnancial reports, while quasi-insiders can know quite a bit more about the company’s plans and prospects. Patience can also reXect a lack of liquidity in large-block investments.
Quasi-insiders are likely to be more risk averse than outsiders. In the case of large block shareholders, this is because their portfolios are undiversiWed. Banks will be risk averse because part, if not all, of their Wnancial interest in the company is in the form of loans earning Wxed rates of interest, giving the bank little to gain from its client’s spectacular proWts and much to lose from the Wrm’s insolvency.
Much of what is written on this topic takes the form of advocacy of one system of the other. The VOC analysis has striven to tone down the discussion, by making the argument that quasi-insider control and market control are good for diVerent things.
Patient capital Wts well with highly skilled teams of workers with good job security, who work together to make incremental improvements in complex products. Automobiles
168 THE GLOBAL ENVIRONMENT OF BUSINESS
are one example, manufacturing equipment is another. (In the Hall–Soskice world, patient capital and secure workers are complementary inputs – a world apart from Roe’s story that quasi-insider control is a costly but necessary response to politically driven job security). Similarly, liberal Wnancial markets are seen as Wtting well with Xexible labor markets – capital and labor are complements, and it does no good to be able to scale capital quickly up or down if labor can’t be treated in the same way. Again, the system of governance is suited to particular products and particular approaches to production. A Hollywood movie, a new software product, an oil Weld – investments and workforces put together for the occasion.