Influencing Managers to Change Unpopular Corporate Behavior through Boycotts and Divestitures: A Stock Market Test

Literature › Works

Wallace N. Davidson III; Dan L. Worrell; Abuzar El-Jelly (1995). Influencing Managers to Change Unpopular Corporate Behavior through Boycotts and Divestitures: A Stock Market Test. Business & Society 34(2). Sage Publications / International Association for Business and Society.

Type	journal article
Scope	Wider apartheid literature
Method	econometric
Period studied	1969–1991
Regions	United States, South Africa (as target of US divestment campaigns)
Themes	sanctions divestment, business and state, comparative theory, methods data sources
Citation key	iii1995a

Argument

The authors build an agency-theory model of how stakeholders can change unpopular corporate behavior. In the model, an organized protest works only if it hurts the stock price enough to motivate managers. They run an event study of press-reported announcements from 1969-1991: 59 boycotts and 35 stock divestitures, the latter mostly over South Africa. Boycott announcements are followed by significant negative abnormal returns, while divestiture announcements are not. About 33% of boycotted firms announced behavior changes, against about 11% of divested firms. The authors conclude that, from a pure stock-market perspective, boycotts are a more effective lever than divestment.

Evidence and method

Press announcements of boycotts and divestitures from the Wall Street Journal, Washington Post, LA Times, Chicago Tribune and St. Louis Dispatch, 1969-1991. Returns come from CRSP tapes for NYSE/AMEX firms (8 OTC boycott cases are examined separately). Corporate responses are taken from press stories and annual reports. The divestiture sample (Table 3) is mostly US firms with South African ties.

Place in the debates

• Effectiveness of divestment/disinvestment as pressure on firms doing business in South Africa (see Sanctions, disinvestment and the anti-apartheid movement). Skeptical on the financial channel. Divestment announcements produce no significant stock-price reaction, because other investors buy the shares and expected cash flows are unchanged. Few targeted firms announced changes. The authors allow that media attention may still harm a firm's image.
• Agency theory: do markets or managerial labour markets discipline managers? (see Sanctions, disinvestment and the anti-apartheid movement). Treats both the classic (Alchian/Manne) and Fama views as implying that a negative price reaction can motivate managerial change. The classic view is seen as more immediate and easier to observe.

Findings

• Headline result: boycott announcements depress share prices, while divestiture announcements do not. (p. 2)

  Announcements about boycotts were associated with significant negative market reactions, whereas divestiture announcements resulted in no significant market responses.

• The divestiture sample is dominated by anti-apartheid/South Africa cases, so the results may not generalize to other divestments. (p. 13)

  the vast majority of the 35 stock divestiture announcements were attributable to company ties with South Africa.

• Across the full boycott sample there is a significant negative announcement-day abnormal return. (p. 18)

  On Day 0, the MCPE is -0.32 (z =-2.01, p = .0445), which is significant at the .05 level.

• No significant abnormal returns follow the 35 divestiture announcements (Hypothesis 2 supported). (p. 19)

  All MCPEs are insignificant at the .05 level. This is consistent with Hypothesis 2.

• About a third of boycotted firms announced behavior changes. (p. 21)

  Of the 59 firms with boycotts, there were 20 announced behavior changes. This is about 33%.

• Only a small share of divested firms announced behavior changes (e.g. Chemical Bank, Citibank, Ford and Mobil on South Africa). (p. 21)

  Panel C of Table 5 lists four firms that announced behavior changes after a divestiture. From the original 35 firms hit by divestitures, this is about 11%.

• Some divested firms explicitly refused to change. (p. 21)

  There were three firms (Chevron, Union Carbide, and Warner Lambert) that announced they were not changing their behavior shortly after the divestiture.

• Divestment announcements clustered in 1986-89, coinciding with US sanctions legislation and the end of the Sullivan Principles. (p. 22)

  Almost all the divestitures occurred in the 1980s and involved doing business in South Africa.

• Explanation for the null result on divestiture: buyers absorb the shares, so cash flows are unaffected. (p. 22)

  This is consistent with the expectation that other investors buy the divested stock and with a consequent lack of effect on expected cash flows.

Evidence ledger

Sources used

Source	Kind	Period	How used
Press announcements of boycotts and divestitures in Wall Street Journal, Washington Post, Los Angeles Times, Chicago Tribune, St. Louis Dispatch, 1969-1991	newspapers periodicals	1969-1991	Source of the event sample (59 listed boycott, 8 OTC boycott, 35 divestiture announcements) and of event dates (Day 0).
University of Chicago Center for Research in Security Prices (CRSP) tapes	stock market financial data	1969-1991	Daily stock returns and market index returns used to estimate market models and compute abnormal returns around announcement dates.
Press and annual-report search for boycott endings and corporate behavior changes (Table 5: WSJ, New York Times, Los Angeles Times, American Banker, Black Enterprise, Automotive News, Platt's Oilgram, company annual reports)	newspapers periodicals	c.1969-1991	Used to identify boycott endings and announced behavior changes after boycotts (20 of 59) and divestitures (4 of 35).
Pruitt and Friedman (1986) boycott stock price study, built on Friedman (1985) boycott sample 1970-1980	secondary literature compilation	1970-1980	Prior event-study evidence (21 boycotts, negative returns at days -1 and 0) used as a benchmark for Hypothesis 1 and comparison with the present results.
Jensen and Murphy (1990) CEO pay-performance data	secondary literature compilation	several decades	Cited estimate of pay-performance sensitivity ($3.25 per $1,000) to support the agency-theory argument that stock price drops motivate managers.

New data

Authors built an event sample of 59 boycott announcements for NYSE/AMEX-listed firms (40 beginnings, 10 ends), 8 OTC boycott announcements and 35 stock divestiture announcements (mostly South Africa-related), 1969-1991, from five newspapers, with abnormal returns computed from CRSP; plus a hand-compiled list of announced behavior changes (Table 5).

Inference

Quantification: Statistical analysis.

Standard event-study methodology (Fama, Fisher, Jensen, Roll 1969): market-model regressions on a 150-day estimation period, mean cumulative prediction errors with the Dodd-Warner Z statistic, and nonmatched t-tests comparing boycott and divestiture samples. The agency-theory model is conceptual and is only partially tested. Outcomes are tallied descriptively from press reports.

Key claims

Claim	Basis	Strength
Boycott announcements are associated with significant negative abnormal stock returns. (p. 187)	Event study of 59 listed-firm boycott announcements (CRSP returns, press-dated events); the 40 beginnings show negative MCPEs over days -4 to -2 and day -1.	Systematic data
Divestiture announcements (mostly South Africa-related) produce no significant market reaction. (p. 188)	Event study of 35 divestiture announcements; all MCPEs insignificant at .05.	Systematic data
Boycotts produce more negative abnormal returns than divestitures. (p. 188)	Nonmatched t-test comparing individual firm MCPEs of 40 boycotts versus 35 divestitures; significant on day -1, only marginal over -4 to -2.	Partial data
The market does not react to announced boycott endings. (p. 191)	MCPEs for 10 boycott-ending announcements are insignificant; very small sample.	Partial data
Firms hit by boycotts announce behavior changes more often (about 33%) than firms hit by divestitures (about 11%), implying boycotts are more effective at motivating change. (p. 190)	Hand-compiled press and annual-report search (Table 5); counts of 20 of 59 boycott firms versus 4 of 35 divestiture firms; no controls and press-limited.	Illustrative or anecdotal
A negative stock price reaction is a necessary motivator for managers to change unpopular behavior (the agency-theory model). (p. 176)	Theoretical model drawing on agency literature; only the stock-reaction part is tested empirically and manager response is not tested causally.	Asserted

The author's caveats

Authors note that the divestiture sample is dominated by South Africa-related cases because the press covered them, so results may not generalize. Outcomes are limited to press-reported changes, firms may change behavior without announcing it, announced changes may be window dressing, and the boycott-end sample (10) is small with possible anticipation. Some earlier boycott announcements may not have reached the national press, so the negative reaction may be understated.

Gaps in the evidence

The argument links stock reactions to managerial behavior change without tracking actual changes in policy or sales, or testing whether abnormal returns predict outcomes across firms. Divestiture events are clustered on common dates (e.g., 05/26/87), and the announcement dates and investor sizes are not documented, so event-date clustering and the scale of divested holdings are not addressed. Company records, sales data and fund-level holdings (available from archives and filings) would allow better tests, as would a regression of outcomes on abnormal returns.

Replicability: Data described but not public.