Activist Hedge Funds and Takeovers: Their Effects on Employment and Performance

This paper analyses the impact of activist hedge funds (AHFs) on post-merger workforce downsizing and operating performance. AHFs have been widely criticized for achieving short-term gains at the expense of other stakeholders, such as employees. The results show that AHF ownership and presence in acquiring firms is a significant determinant of post-merger employment reductions. There is little evidence that these mergers and acquisitions have better operating performance relative to other takeovers. However, there is a negative effect of AHF ownership on labour productivity. Overall, the results are consistent with the view that AHF involvement in takeovers does not lead to sustained gains in performance.


Introduction
Shareholder activism has become an important corporate governance mechanism.Specifically, activist hedge funds (AHFs) now play a major role in mergers and acquisitions (M&As) and subsequent corporate restructuring (Boyson, Gantchev and Shivdasani, 2017;Gantchev, Sevilir and Shivdasani, 2020;Greenwood and Schor, 2009).As in takeovers initiated by other types of investors (see Amess, Girma and Wright, 2014;Conyon et al., 2002), restructuring initiated by AHFs can involve employment reductions (Chen, Meyer-Doyle and Shi, 2021;DesJardine and Durand, 2020), wage cuts (Brav, Jiang and Kim, 2015) and divestitures (Gantchev, Sevilir and Shivdasani, 2020), all aimed at securing immediate gains in company value and operating performance.Some argue that these initiatives breach long-term implicit contracts with employees (Agrawal and Lim, 2021) and facilitate wealth transfer from employees to shareholders (Coffee and Palia, 2016).However, whether restructuring after takeovers with AHF involvement leads to performance improvements is uncertain, with mixed evidence on post-takeover performance (Gantchev, Sevilir and Shivdasani, 2020).
AHFs differ from most other investment funds in their investment time horizons and strategic intentions (Cumming and Wood, 2019;Gospel and Pendleton, 2014).As AHFs tend to make shortterm investments (on average between 1 and 2 years) (Brav et al., 2008), they are likely to initiate or seek changes that can be implemented more or less immediately after the transaction (Brav, Jiang and Kim, 2015).Often AHFs are the instigators of takeovers aimed at industry restructuring, as in the case of Trian and Pentair plc. 1 Such restructuring may increase the share price in the short term, but have little impact (or even adverse effects) on longer-term performance.This has been the primary criticism of hedge funds: they extract short-term value and then exit, leaving others to 'pick up the pieces' in the longer term.
Whilst there is now extensive, though mixed, evidence on the performance outcomes of AHF interventions to force governance changes, and there is some evidence on AHF effects on takeover targets (Boyson, Gantchev and Shivdasani, 2017), there is very little evidence on the employment and performance effects of AHF involvement in takeover acquirers.Further, there is almost no evidence at all on the effects of AHF activity in the UK, despite the UK being the second most important location of hedge fund activity worldwide after the USA.This paper adds to the evidence base by analysing the relationships between AHF ownership, posttakeover workforce downsizing and long-term operating performance.To this end, we address two main questions.Are there greater employment reductions where AHFs are blockholders in the acquiring firm?Do takeovers with AHF involvement have superior post-merger profitability and productivity?
To answer these questions, we analyse the impact of AHF ownership on the probability and magnitude of post-merger workforce downsizing as well as on post-merger operating performance over a 3-year period.Our analysis is based on 421 M&A transactions involving UK publicly listed companies during 1990-2017. Of these, 26 (28) . Of these, 26 (28) acquirers have at least one AHF blockholder at the beginning (end) of the financial year in which the M&A deal is completed.Although AHFs are involved in relatively few takeovers, our results show significant relationships between AHFs and post-merger employment changes.AHF ownership in the acquirer is positively associated with the likelihood of employment reductions, along with the magnitude of employee change.The profitability of acquirers with AHF ownership does not differ from that of other acquirers.However, there is strong and consistent evidence that acquiring firms with AHF ownership experience lower labour productivity post-takeover.Furthermore, takeovers with AHF presence experience changes in profitability and productivity that are significantly worse than in other takeovers.
The main contribution of this paper is that it shows that AHF ownership has a set of adverse impacts on post-takeover employment, profitability and productivity, and these findings contrast with those from the USA that emphasize the positive effects of AHF activity (e.g.Brav, Jiang and Kim, 2015;Gantchev, Sevilir and Shivdasani, 2020).Our findings support those views which suggest that initial gains are dissipated longer term (Coffee and Palia, 2016).We find all pain, little gain.An important contribution of the research is that it shows that AHFs do not have to mount public campaigns to make an impact, if the corporate governance system is more 'shareholderfriendly' than that found in the USA.To date, there have been few public campaigns by AHFs in the UK.Our focus has been guided by the nature of the UK corporate governance system, which generally favours 'behind-the-scenes' relationships with managers rather than overt conflict and public campaigns.Our results bear out the validity of our approach.An important implication of our results, therefore, is that the nature of corporate governance regimes should be factored in when evaluating the impact of particular actors on company behaviour and performance (Buchanan, Chai and Deakin, 2020).Institutional context affects how actors achieve their objectives.
Although there is an extensive literature in management, economics and finance on the effect of takeovers on employees (Amess, Girma and Wright, 2014;Dessaint, Golubov and Volpin, 2017;Geurts and Biesebroeck, 2017) and companies (King et al., 2004;Renneboog and Vansteenkiste, 2019), there is very little on AHF involvement in takeovers, and its effects, even in the USA (as opposed to other governance interventions).Those studies focusing on AHF involvement have tended to examine the effects on targets rather than acquirers (e.g.Boyson, Gantchev and Shivdasani, 2017).AHFs might be expected to have substantial effects on takeover outcomes because of their concern to influence management decision-making.We find this, but the effects are either negative or weak, unlike recent US studies (e.g.Gantchev, Sevilir and Shivdasani, 2020).The adverse effect on productivity is especially notable and, taken in conjunction with results showing employment reductions, is consistent with those perspectives on takeovers that emphasize the damaging effects of loss of human capital (Chen, Meyer-Doyle and Shi, 2021).The implication of our findings for managers is that they should seek to avoid AHF attention, especially if planning M&As.If AHFs are present, managers need to be judicious in responding to any AHF pressure for restructuring post-takeover.
The paper proceeds by reviewing the literature and developing the hypotheses.The data and key variables are then discussed, results are presented and conclusions are drawn.

AHF ownership, mergers and restructuring
There is growing interest in the effects of activist investors on investee companies (Cumming and Wood, 2019;Gillan and Starks, 2000;Goergen et al., 2018;Gospel, Pendleton and Vitols, 2014;Wright et al., 2007).Activism can be defined as 'taking an equity stake in a company with the intention to influence the control or the management of the company' (DesJardine andDurand, 2020, p. 1055).Shareholder activism has become a broadly used investment strategy associated with the increasing willingness of investors to trigger corporate change and intervene in management's decisions (Bratton and McCahery, 2015;Filatotchev and Dotsenko, 2015).An important issue in the literature on activism is the nature and extent of the impact on company operations and performance (Briggs, 2007;Coffee and Palia, 2016).
Activist investors come in various forms, including some pension funds (e.g.CalPERS) (Jacoby, 2007), 'social activists' such as unions and religious organizations (Ahn and Wiersema, 2021) and some sovereign wealth funds (such as the Norwegian Government Pension Fund-Global) (Goergen et al., 2018).Hedge fund activism is distinctive as it focuses on securing changes in management strategy or organization to achieve short-term abnormal stock returns (Kahan and Rock, 2007).AHFs have primarily financial motives and tend to target underperforming companies (Boyson, Gantchev and Shivdasani, 2017) to improve performance via governance, financial and strategic changes.Hedge fund activism therefore differs from relational investing aimed at achieving longer-term improvements in performance, and from activism aimed at improving social or environmental performance.Indeed, much of the criticism of AHFs is that short-term stock price gains are achieved at the expense of longer-term performance, including R&D investment (Coffee and Palia, 2016), social and environmental concerns (DesJardine, Marti and Durand, 2021) and the interests of other stakeholders, such as employees (Agrawal and Lim, 2021).
Hedge funds typically pursue more aggressive investment strategies than other institutional investors such as mutual funds or pension funds, and thus can have a greater impact on company decision-making.Originally, the term was used to define funds that hedged their investments, but it now covers a wider range of investment strategies (Gospel and Pendleton, 2014, p. 9).The distinctive features of hedge funds are that they are relatively undiversified, lightly regulated and open only to wealthy and sophisticated investors, presumably with relatively low risk aversion.Most are registered offshore.The structure of management fees and returns to the general partners of the fund is such that they are incentivized to pursue risky and short-termist strategies (Coffee and Palia, 2016).
Hedge funds have a variety of investment strategies, the most common being 'directional' (whereby funds exploit and even cause market movements).Activism is less common and can vary from being the primary investment focus through to occasional activism (Buchanan, Chai and Deakin, 2020).Currently, 897 hedge funds worldwide are recorded as activist shareholders in the Activist Insight database.Among these, 112 hedge funds are classified as 'primary focus' activists and 206 as 'partial focus' activists.2Our attention centres on such funds.In 2018 (2019), 55 (55) UK firms were targeted by such activist investors, a steady increase from 2014.Of these, between 15% and 33% have been M&A related (Booth, 2020).
Although AHFs are becoming increasingly important sources of activist pressure on company managements in the UK (Becht et al., 2009;Becht et al., 2017), most research on hedge fund activism has been conducted in the USA (where such activity is most widespread; Gospel and Pendleton, 2014;Wiersema, Ahn and Zhang, 2020).This research focuses on observable interventions, such as proxy fights, to put pressure on management (Aguilera, Federo and Ponomareva, 2019).The objectives of activism include changes to governance, financial structure or assets, and the sale of the company.Acquisition of seats on the board of directors is the most common campaign objective (Ahn and Wiersema, 2021, p. 105).AHF campaigns tend to be highly public, with AHFs competing for shareholder votes against incumbent managements.The public nature of activism in the USA (which has tended to define how hedge fund activism is viewed) may well owe much to the nature of the US corporate governance system, whereby managers have had a series of legal and other protections against shareholder 'sovereignty', such as 'poison pills' and staggered boards.Opposition to management has almost inevitably taken place in the public arena.Investor activism has been enhanced by recent legal reforms, giving AHFs more freedom to mount campaigns (Ahn and Wiersema, 2021).
As institutional theory predicts, investor behaviour is likely shaped by institutional and systemic contexts (DiMaggio and Powell, 1991;Scott, 2007).The differences between the corporate governance systems of the UK and the USA mean that AHF activism may take different forms.The more 'shareholder-friendly' regime in the UK, and the long-standing preference for governance relationships to take place privately and discretely (Black and Coffee, 1994;Holland, 1998;Gospel and Pendleton, 2005), suggests that AHF activity is more discrete, and less confrontational.It often comprises 'behind-the-scenes' relationships with managers (Levit, 2019;McCahery, Sautner and Starks, 2016).
Whilst highly public and oppositional US-style activism does occur in the UK, often instigated by US-based hedge funds, it tends to be rare.The most successful cases of US funds prompting change in UK companies have used collaborative activism (with managers) rather than 'an aggressive, table-thumping, American-style campaign' (Booth, 2020, p. 10).The important methodological consequence is that evaluation of the full effects of AHF activity in the UK cannot be solely linked to observable activist intervention campaigns or events.As the mere presence of AHF ownership can put pressure on management, studies of the effects of AHFs should focus on presence as well as specific interventions (DesJardine and Durand, 2020).
The focus of our research is AHF involvement in M&As.This is novel, as previous work on AHFs has mainly focused on interventions to force gover-nance, board composition or strategy change.The rationale for our approach is the steep increase in M&A activity by hedge funds in recent years (Reetun, 2020). 3We examine an area of AHF activity that has not been widely examined previously.We do this by studying acquiring firms with AHF blockholders in the immediate run-up and aftermath of takeovers.
AHF interventions in takeovers can take a variety of forms.In the USA, Greenwood and Schor (2009) and Boyson, Gantchev and Shivdasani (2017) report that AHFs target firms with more opportunities for restructuring (such as highly unionized companies; see Brav, Jiang and Kim, 2015) and urge third parties to acquire the targeted firm. 4Alternatively, AHFs may campaign against those M&A deals offering lower premiums to the acquired firm's shareholders. 5Gantchev,  Sevilir and Shivdasani (2020) argue that AHFs discipline 'empire-building' managers by constraining them to make fewer acquisitions, preventing acquisitions aimed at diversification.
Equally, there are cases in our dataset where AHFs pressured managers to mount several takeovers with a view to both diversification and disposal of surplus activities.In 'behind-thescenes' negotiations with management, AHFs often push for divestments, along with employment reductions and lay-offs (Chen, Meyer-Doyle and Shi, 2021).A good example in our dataset is the takeover of former Monopoly-maker Waddington by the John Mansfield Group in 1999.Mansfield's management were encouraged by an AHF investor to acquire struggling companies, such as Waddington, with a view to developing parts of them and disposing of the remainder (The Independent, 1999).
According to 'wealth transfer' theory (Bacon, Hoque and Wright, 2019;Shleifer and Summers, 1988;Wright et al., 2007), activists initiate or encourage M&A deals and subsequent restructuring to achieve short-term abnormal stock returns at the expense of other stakeholders such as employees.This is because such interventions are typically aimed at companies believed to be underperforming, where there is scope for quick and substantial improvements.The expectation is that the stock market will respond positively to such initiatives to enhance labour productivity and operating performance.These takeovers can therefore lead to substantial workforce downsizing and wage cuts (Amess, Girma and Wright, 2014;Goergen, Brewster and Wood, 2013).Some studies show that reductions in labour costs are the main source of M&A gains (Dessaint, Golubov and Volpin, 2017), and that most of the M&A premium is recovered by such reductions (Krishnan, Hitt and Park, 2007).
US evidence shows a decline in employment after AHF interventions (Brav, Jiang and Kim, 2015).Allaire and Dauphin (2016) report a reduction of around 15% in the workforce of firms targeted by AHFs relative to workforce changes in other firms.DesJardine and Durand (2020) find that AHF interventions are associated with a 4.6% decline in workforce size in the year after the intervention, and 7.7% within the 5-year period. 6Chen, Meyer-Doyle and Shi (2021) conclude that activist-induced workforce downsizing leads to an undesirable loss of human capital as valuable employees leave firms targeted by AHFs, with potentially negative effects on company performance.
However, there is limited empirical evidence on the impact of AHFs on employment in the UK.Extant evidence tends to be case study based (Gospel et al., 2011), as in discussions of the effect of hedge fund interventions on employment in Cadbury's plants after the Kraft takeover7 (Pendleton and Gospel, 2014).A good illustration of the processes at work is provided by the former CEO of one of the companies in our dataset: I was approached by an activist combining investing with consulting… They then approached our lead-ing shareholders… and proposed using United Industries as an acquisition vehicle to vacuum up small UK-quoted engineering companies and then they would cut their costs… After he resigned in protest at the adoption of this strategy: … a successor was appointed with the simple brief to acquire almost anything and then use these consultants to cut its costs… The consultants duly went to work… the result of this acquisition and cost-cutting strategy?Just two years after I left United Industries as a profitable cash generative group…, it collapsed and was broken up.(Brown, 2017, pp. 84-85) Whilst these accounts suggest that AHFs affect company decision-making, more systematic and generalizable evidence is needed to determine whether AHFs have adverse impacts on labour after M&As, given that public debate often highlights this outcome (Ahn and Wiersema, 2021;Filatotchev and Wright, 2017).Based on the discussion above, the supposition is that the short-term focus of AHFs is especially associated with cost savings post-takeover, and that labour bears the brunt of these.Based on this discussion, we arrive at the following hypothesis: H1: Acquiring firms with AHF ownership will have larger employment reductions post-merger.

AHF ownership and post-merger profitability
Although AHF interventions are motivated by a concern to improve company performance, evidence is mixed.Some find evidence of performance improvement in the short term (e.g.Brav et al., 2008), others report little change (e.g.deHaan, Larcker and McClure, 2019), whilst others find significant decline in the longer-term performance (e.g.DesJardine and Durand, 2020).Nevertheless, most studies show improvements in operating performance after AHF interventions (Denes, Karpoff and McWilliams, 2016), possibly as a result of divestment of underperforming assets (Chen and Feldman, 2018;Clifford, 2008).Much of the extant evidence relates to performance improvement close to the intervention.Brav et al. (2008) and Clifford (2008) report ROA improvement for the 2-year period after the AHF intervention, whilst other researchers' similar findings relate to the year following the event (e.g.González and Calluzzo, 2019;Boyson and Mooradian, 2011).Longer term, Bebchuk, Brav and Jiang (2015) and Bebchuk et al. (2020) find only weak evidence of profitability improvements in the 5 years after activist interventions.Conversely, Greenwood and Schor (2009) and deHaan, Larcker and McClure (2019) find no improvement in operating performance measures after AHF interventions.Boyson and Pichler (2019) report that the intervention strategies of AHFs are not associated with ROA, but are associated with higher cash flow returns.DesJardine and Durand (2020) find that firms targeted by AHFs improve their market value and profitability in the short term (2 years), but they incur substantial losses in their operating cash flow performance during the subsequent 5-year period.
However, these results relate to US settings: alternative settings may mean that AHF strategies are less successful in enhancing performance.This supposition is borne out by Buchanan, Chai and Deakin's (2020) study of AHFs in Japan, which finds mostly insignificant changes in shortterm performance, but significant negative changes in longer-term performance, including both firm profitability and market capitalization.
Concerning AHF involvement in M&As, there is limited and mixed empirical evidence on operating performance changes post-acquisition.Wu and Chung (2021) find that acquirers with AHFs improve their performance post-takeover.Hege and Zhang (2020) conclude that, when acquirers are targeted by AHFs, large acquirers do not secure improvements in operating performance, but small ones do.In contrast, Boyson, Gantchev and Shivdasani (2017) show that firms targeted by AHFs in M&A deals, but not acquired, significantly improve their performance in comparison to non-targeted firms.Further research reveals that AHFs accelerate forced CEO turnover after poor M&As and only allow managers to make selective acquisitions that generate high shareholder returns (Gantchev, Sevilir and Shivdasani, 2020).
Overall, AHFs may focus on improving the short-term profitability of the targeted firms by actively participating in their governance.The above discussion leads us to the following hypothesis: H2: Acquiring firms with AHF ownership will have improvements in their profitability during the 3-year post-takeover period.

AHF ownership and post-merger labour productivity
Post-merger restructuring initiated by AHFs may improve labour productivity by shedding less efficient operations and labour.Brav, Jiang and Kim (2015) show that US plants, targeted by AHFs, improve their labour productivity by reducing their workforce and work hours.Analysing the spillover effects of AHF interventions among the competing firms, Aslan and Kumar (2016) conclude that both firms targeted by AHFs and their rival firms improve their productivity.Also, Brav et al. (2018) report that improvement in the innovation activities of the firms targeted by AHFs improves productivity.However, Chen, Meyer-Doyle and Shi (2021) find that firms targeted by AHFs achieve lower performance improvements when these firms experience higher departure of valuable employees (i.e.employees with stock options).Targets of AHFs may also substantially reduce their strategic investments, operating expenses (including wages) and R&D expenditure (DesJardine and Durand, 2020).Similarly, Japanese companies targeted by AHFs do not experience significant changes in labour productivity (Buchanan, Chai and Deakin, 2020).
A key issue is whether AHFs put longer-term performance of the targeted firms at risk by focusing on short-term returns, possibly by transferring wealth from other stakeholders (Agrawal and Lim, 2021;Coffee and Palia, 2016).The argument here is that short-term abnormal returns are generated by measures which destroy valuable human capital in the company, thereby damaging productivity and profitability in the longer term.Supporting this 'wealth transfer' argument, Brav, Jiang and Kim (2015) report a reduction in the productivityadjusted wages of employees, despite productivity improvements after AHF interventions.Expecting such interventions, valuable employees may accelerate their departure from the targeted firms (Chen, Meyer-Doyle and Shi, 2021) and consequently these firms may incur 'unexpected' corporate outcomes in the longer term (Buchanan, Chai and Deakin, 2020).
Previous studies find that post-merger restructuring involves large-scale workforce reductions (Amess, Girma and Wright, 2014;Conyon et al., 2002) and lay-offs (Kuvandikov, Pendleton andHiggins, 2014, 2020), which may impact on labour productivity.Some studies conclude that M&As enhance labour productivity, although such transactions result in post-merger workforce downsizing (McGuckin and Nguyen, 2001;Siegel and Simons, 2010).During the post-takeover integration process, AHFs may also initiate additional labour cost savings-oriented restructuring, which may substantially impact on the performance and productivity of the merging firms.As there are limited and contradictory views on the association between AHF ownership and post-merger labour productivity, we test the following hypothesis: H3: Acquiring firms with AHF ownership experience improvements in their labour productivity during the 3-year post-takeover period.

Approach and sample
We analyse the impact of ownership by AHFs on employment changes and operating performance using data on 421 M&As occurring between UK listed firms during 1990-2017.During this period, AHF activity has developed from small beginnings in the early 1990s to more extensive activity in the 2000s, especially since 2014.The sampling period ends in 2017, since we require 3 years of posttakeover performance data.The overall approach is to examine the effects of both the presence and level of ownership of AHFs in the acquirer in the lead-up to the takeover and shortly afterwards on various measures of employment change and operating performance.The rationale for this approach is that AHFs will influence the objectives (e.g.efficiency improvements) and character of the takeover, along with the measures taken in the immediate aftermath to secure the fruits of the transaction.A key assumption is that AHFs take an ownership position in a company specifically to secure changes to governance, management or strategy (Briggs, 2007).
During the period there were 1,147 full takeovers (where acquirers secured more than 50% of the target's equity) of UK listed companies by other companies listed on the London Stock Exchange.Takeovers involving property, financial and utility companies are excluded because these firms have either different assets characteristics, reporting requirements or regulatory regimes.This leaves us with 746 transactions.Of these, we retained 421 (56% of the total), after removing 325 cases where acquirers conducted multiple acquisitions within 3 years of the observed takeover (making it difficult to isolate the effects of particular takeovers).Our final sample is similar to other studies of UK takeovers (e.g.Mira, Goergen and O'Sullivan, 2018).
Transaction-related data, including M&A dates, premium, takeover mode and payment type were collected from Thomson Reuters.These data were merged with financial and employment data from Datastream.Ownership by large shareholders was hand-collected from company annual reports, downloaded from the Filings Expert database.The identity of AHFs was checked against the Activist Insight database.
As outlined earlier, AHFs' influence in the British corporate governance system can be secured privately rather than via US-style public campaigns.We therefore analyse the relationships between AHF presence (measured by ownership and simple presence) and post-merger workforce changes.Next, we analyse the impact of AHF presence on post-takeover operating performance.We run a series of regressions based on the following:

Dependent variables
Post-merger Employment change%, Employment decline%, Employment growth% and Employment reduction variables are used to measure post-merger workforce downsizing.Employment change% by the end of the first financial year after the takeover is calculated by subtracting the combined pre-takeover employment of the acquirer and target from the employment of the post-merger firm at the end of the first full posttakeover year, and then dividing the difference by the average of post-and pre-takeover employment (Davis et al., 2014).To analyse the effects of AHFs on employment, we create two truncated variables to be used in tobit regressions: Employment decline% uses the absolute values of employment reductions, with zeros for cases with employment growth; whereas Employment growth% records the percentage growth where growth has occurred, with zeros where there is employment decline.Employment reduction -an alternative binomial measure -is derived from the above information on employment change and is equal to one if there is employment decline by the end of the first full financial year after takeover, and zero otherwise.
Performance is measured by ROA, ROS and Productivity variables in each of the three posttakeover years.ROA is defined as earnings before interest, taxes, depreciation and amortization divided by the book value of total assets at the beginning of the year.ROS is arguably a more suitable measure for M&As, because there may be large changes in asset valuation posttakeover (Krishnan, Hitt and Park, 2007).We control for industry-level profitability changes by using industry-adjusted measures for each firm in the regressions.Productivity is sales divided by the annual average number of employees (Amess, Girma and Wright, 2014;Datta et al., 2010;Goergen, O'Sullivan and Wood, 2014).As the Productivity variable is not normally distributed, we log-transform this variable (Brav, Jiang and Kim, 2015).

Independent variables
Our main independent variables are the equity ownership of the AHFs in the acquirers (i.e. percentage ownership) and a dummy variable, which indicates whether an AHF is present.We collect data on AHFs from the pre-takeover year and takeover completion year annual reports of the acquiring firms. 8We use four measures of AHF ownership and presence: AHF, AHF presence, post-merger AHF and post-merger AHF presence.Ownership is an often-used measure of investor power in corporate governance research (Goergen, 2018).As AHFs often take small stakes as a 'trojan horse' for intervention in the management of firms, their mere presence may impact on employee and company outcomes.
The hedge funds in the sample are compared against the activists in the Activist Insight database.This database distinguishes between four groups of activist shareholders according to their focus on activism: 'primary focus', 'partial focus', 'occasional focus' or 'engagement focus'.If a hedge fund has a 'primary focus' or 'partial focus' on activist strategies, and media sources confirm this, we classify this hedge fund as an AHF.Activists with an 'occasional' or 'engagement' focus are not classified as AHFs in our analysis, because we cannot be sure that they actively intervened in the observed takeovers.They are classed as institutional investors instead.Finally, any activist investors other than hedge funds are classified as institutional investors.
As shown in Table 1, a total of 24 AHFs have large equity stakes (exceeding 3% of equity) in 26 (28) acquirers at the beginning (end) of the takeover year.Seventeen acquiring firms have AHF ownership throughout the takeover period.AHFs are present at some stage in 37 (8.8%) of our sample cases.Further, 7.6% of all worldwide operating AHFs (as classified in the Activist Insight database) are present in our sample.9On average, these AHFs own 9.4% (11.9%) of the acquiring firm shares at the beginning (end) of the takeover year.
The pattern of AHF activity in our sample is similar to that described in the literature (e.g.Shi et al., 2020). 10Except for three cases, AHFs have exited by the end of the first full financial year after the takeover.Eighty-one percent of takeovers with AHFs have taken place since 2000, with a small majority of these in the period before the financial crisis of 2007-9.

Controls
It is important to control for performance in both target and acquirer prior to the takeover, as there could be path-dependency effects on post-takeover outcomes (Goergen and Renneboog, 2004).Controls are included for profitability (ROA), Leverage (the ratio of total debt to total assets), Cash holdings (the ratio of cash and cash equivalents  1A. to total assets) and Acquirer (Target) size, measured by market value.Since the post-merger company may change its employment by divesting business units, a dummy (Divestments) records whether divestments occurred by the end of the first post-takeover financial year.One-hundredand-one acquirers (24%) make such divestments (similar to Maksimovic, Phillips and Prabhala, 2011).
As other large shareholders may influence management decisions, we control for blockholdings (3% or more) by investment institutions (Institutions (post-merger)) and individuals (Individuals (post-merger)) in the acquiring firm at the beginning (end) of the takeover completion year.
We also control for transaction characteristics as they can affect employment and performance post-takeover (Lehto and Böckerman, 2008;Martynova and Renneboog, 2011).Hostile takeover indicates whether the initial bid was rejected by the target (Franks and Mayer, 1996;Weir and Wright, 2006).Cash records whether the transaction was paid wholly in cash.Related acquisitions are those where both the target and acquirer are in the same sector (as per Datastream Industrial Classification Level Four).Since high takeover premiums can lead to employment reductions (Krishnan, Hitt and Park, 2007), we include Premium (i.e. the percentage difference between the purchase price and the market price of the acquired firm's shares 30 days before the first announcement of the takeover, divided by the latter; Hayward and Hambrick, 1997).Relative employment -the ratio of employment in the target to that in the acquirer prior to the takeover -is included as it may affect the capacity of the acquirer to digest the target.
Table 1 provides descriptive statistics of all the variables: Table 1A in the online Supporting Information provides their definitions, while Table 2A shows the correlations among these variables.

Univariate analysis
Table 2 contains a univariate analysis of workforce downsizing (measured with employment changes in percentage and the employment reduction dummy variable) and operating performance (measured by ROA, ROS and Productivity across the 3 years post-takeover).In Panel A, one set of figures compares acquirers with AHF investors pre-takeover with other firms, whilst the other set compares post-merger firms with AHF investors with the rest.Employment change is significantly more negative amongst those acquirers with AHFs before as well as after the takeover.The difference in Employment change% between these two subsamples is 25.30 percentage points and significant at the 0.1% level.However, Panels B and C show that the differences in operating performance measures between firms with AHFs and those without are insignificant.These initial results suggest that the presence of AHFs just before and after takeovers impacts employment changes but not performance.

AHFs and post-merger restructuring
Table 3 presents the results of probit regressions predicting the probability of workforce downsizing (i.e.Employment reduction as the dependent binary variable) by AHF ownership or AHF presence.All columns report the marginal effects of the independent variables.Columns 1 and 2 indicate that AHF and AHF presence have significant effects on the probability of post-merger employment reductions, whilst columns 3 and 4 present that Post-merger AHF has similar effects on em-ployment.If an acquirer has an AHF investor, then the probability of a workforce reduction postacquisition increases by 39.8% (significant at the 0.1% level).The marginal effects for the pre-merger AHF measures are somewhat larger than those for the comparable post-merger measures, suggesting that AHF involvement prior to the takeover has a larger impact.Overall, there is some support for H1, but with the proviso that AHF presence matters while the magnitude of ownership does not.
The results also suggest that Institutions and Individual investors do not affect post-merger workforce downsizing, but the probability of such downsizing depends on the performance of the acquiring firm and relative size of the merging firms.All columns show that Relative employment size is a significant determinant of downsizing, confirming that employment reductions are made in larger targets.Further, Employment reduction depends on Acquirer ROA, but not on Target ROA, as shown in all columns.Among the deal characteristics, Premium is weakly associated with the probability of downsizing, while other deal characteristics, as well as the Leverage and Cash holdings of both firms, are insignificant.

AHFs and the extent of workforce downsizing
Table 4 reports the results of regressions analysing whether AHF ownership explains the magnitude of employment changes.Panel A uses AHF (presence) and Panel B uses Post-merger AHF (presence) as the main independent variable.Columns 1 and 2 use ordinary least squares (OLS) regressions to explain employment changes (measured by Employment change%).The remaining columns use tobit regressions, as the dependent in these columns is censored.Columns 3 and 4 use Employment decline%, while columns 5 and 6 use Employment growth% as the dependent variable.
Panel A shows that AHF is significantly associated with Employment change% and Employment decline% (columns 1 and 3), but not with Employment growth% (column 5).Column 3 indicates that a one standard deviation increase in AHF ownership leads to a 0.77 percentage point higher workforce downsizing.Similarly, columns 2, 4 and 6 use AHF presence to analyse whether the size of employment changes differs between acquirers with and without AHFs.These results are consistent with the above, as acquirers with AHFs make significantly greater employment reductions.Institutions and Individual owners are not associated with the size of the employment reductions in any of the columns.
Panel B repeats the above regressions using Postmerger AHF (presence) as the key independent variable.Columns 1, 3 and 5 show that AHF ownership at the end of the takeover completion year is not associated with the size of post-merger employment change.However, columns 2 and 4 show that Post-merger AHF presence is significantly associated with workforce reductions (at the 1% and 0.1% level, respectively), while column 6 shows that this variable is negatively associated with employment growth.
All columns show that Institutions and Individual owners do not impact on the size of workforce downsizing.Further, Acquirer size is negatively associated with post-merger employment growth, while Acquirer ROA has a positive effect.Relative employment also plays a significant role in determining the size of post-merger workforce reductions.The other control variables are not significant.
To sum up, the results suggest that acquirers with AHFs immediately before and after the takeover make greater employment reductions.Overall, there is strong support for H1.

The impact of AHFs on long-term operating performance
Panel A of Table 5 reports the regressions that analyse the effect of Post-merger AHF ownership on firm profitability and labour productivity.Controls for the deal characteristics are not included as the initial analysis showed that they had insignificant effects, with a negligible impact on model fit.The pre-takeover performance controls were replaced with similar controls for the post-takeover period.Note that the sample size declines somewhat by year due to delisting of some firms. 11here ROA or ROS is the dependent variable, the coefficient on the Post-merger AHF variable is insignificant (except column 4).However, in columns 3, 6 and 9, Post-merger AHF is negatively and significantly associated with Productivity.For example, in column 3 a one standard deviation increase in post-merger AHF ownership leads to 0.49 percentage point lower Productivity.Overall, AHFs with higher equity ownership in the acquiring firms do not impact on profitability, but they negatively impact on productivity.
Panel B of Table 5 reports the regressions that analyse the role of Post-merger AHF presence, as opposed to level of ownership, in explaining postmerger performance.These regressions do not show a consistent difference in profitability (ROA or ROS) between acquiring firms with and without AHFs.However, the presence of AHFs has a consistently negative relationship with productivity, as per columns 3, 6 and 9. Overall, Productivity declines during the longer-term post-merger period.
In sum, there is weak evidence that acquirers with AHFs earn a higher level of profit in the first two post-takeover years (DesJardine and Durand, 2020), but thereafter AHF ownership or presence has no effect.Thus, AHFs have at best a short-term   effect on performance.Conversely, there is a consistently negative effect of post-merger AHF presence on productivity.On balance, therefore, H2 is not supported, whereas H3 is upheld.

Robustness analysis
To check the robustness of our findings we conduct additional tests.First, AHF and Post-merger AHF data are merged to create a larger group of acquirers who have had AHF owners either at or immediately after the M&A.Thereby, we differentiate all acquirers with AHFs from other acquirers without AHFs.The employment, profitability and productivity regressions are re-run with the new variables.The results (reported in Tables 3A and  4A in the online Supporting Information) show that acquirers with AHFs are significantly associated with employment reductions (at the 5% level or better), in all but one estimation.Second, the regressions reported in Table 5A (in the online Supporting Information) also confirm the above findings (reported in Table 5) by showing that there is no relation between AHF ownership (presence) and profitability during the post-takeover period, but there is a significant negative association in three of four estimations between AHF ownership (presence) and productivity in the first 2 years after the takeover.Third, we analyse post-takeover changes in 'raw' (i.e.not adjusted with industry-median performance) profitability and productivity measures.These Change in ROA (ROS or Productivity) variables are computed as the percentage change in ROA (ROS or Productivity) for each post-merger year, relative to the weighted average ROA (ROS or Productivity) of the acquiring and acquired firms in the year prior to takeover.The results of the univariate analyses (reported in Table 6A in the online Supporting Information) indicate that acquirers with AHFs experience significantly negative Change in ROA and Change in ROS during the post-takeover years (as in Greenwood and Schor, 2009;Klein and Zur, 2009).Table 6 reports the results of the regressions which use these new dependent variables.Panel A uses Post-merger AHF, whilst Panel B uses the interaction between AHF and Post-merger AHF (see above) as the main independent variable.In addition to the controls used in Table 5, these regressions also include a dummy variable for Employment reduction along with pretakeover weighted average ROA (ROS or Produc-tivity) of the acquiring and acquired firms in the year prior to takeover, as post-takeover operating performance changes also depend on the initial profitability (productivity) of the merging firms.
Panel A indicates that Post-merger AHF leads to a negative change in the profitability of the acquiring firms during the two post-takeover years, measured with ROA (columns 1 and 4) and ROS (columns 2 and 5).The panel also indicates that Post-merger AHF leads to a negative Change in Productivity during the three post-takeover years (columns 3, 6 and 9).Panel B shows that the interaction between AHF and Post-merger AHF is negatively associated with the post-merger changes in profitability and productivity of the acquiring firms, but these associations become insignificant in the third year.
As we use AHF ownership for the beginning and end of the takeover completion year, and the employment variables from the first full financial year and operating performance variables from the three financial years after the takeover completion year, it is less likely that endogeneity bias impacts on our main results.Nevertheless, we address endogeneity concerns via Durbin-Wu-Hausman tests (Wooldridge, 2020).First, we regress AHF (Post-merger AHF) on all control variables and five instrumental variables, such as acquirer labour costs, target labour costs, acquirer cash flow, target cash flow and acquirer capital expenditure.Then, we test the endogeneity of AHF (Post-merger AHF) by using a residual from the first regression as an additional regressor in our main regression model.Our results do not change qualitatively (see Table 7A in the online Supporting Information).

Discussion and conclusions
Our results show that M&A deals where AHFs are present involve employment reductions.We find that AHF ownership in the acquirers determines the probability and size of these post-merger workforce reductions.In contrast, individual ownership and ownership by other institutional investors is not associated with employment reductions.Overall, these results suggest that AHFs focus on reducing labour costs after M&A deals and put pressure on managers to bring these about.
We also find that post-merger workforce downsizing is associated with lower labour productiv-   ity than in other merged firms.There is no clear effect of AHF presence on levels of profitability.
When changes in performance are examined, there is clear evidence that firms with AHF involvement have worse changes in profitability and, to a lesser extent, productivity, in the first 2 years after the takeover.As the observed employment reductions do not correspond to performance improvements, these results suggest that post-merger workforce reductions initiated by AHFs have a negative impact on long-term human capital investments (Strine, 2016).Overall, our findings indicate that AHFs have an adverse impact on takeovers.This contrasts with the majority of findings from US studies, which show at least positive outcomes in the first year or two after the transaction.One explanation might be that UK AHFs target potential acquirers who are worse performers than other firms mounting takeovers, and that this path dependency is carried over into post-takeover performance.However, there is no evidence to support this: pre-takeover ROA and ROS differ little between acquirers with and without AHF.Another might relate to the nature of the AHF involvement.In the US studies, attention has focused on AHF interventions to enhance governance and management, thereby forcing managers to improve performance.Involvement in takeovers -the topic of our research -is more uncertain.Many takeovers fail to improve performance due to lack of synergy (Renneboog and Vansteenkiste, 2019), and AHFs are ill-suited to guiding managers to achieve successful outcomes.
A novelty of our study compared with previous studies, nearly all of which have been carried out in the governance environment of the USA, is that we focus on AHF ownership, both its magnitude and presence, rather than on observable activism.This represents something of a gamble in that we are not able to observe how AHF owners influence company managers.Nevertheless, our results show that our approach is reasonable as, even though there are only a small number of cases of AHF ownership in our sample, AHF presence is clearly associated with employment changes post-takeover.There is much less evidence of an AHF effect on performance, and it may well be that more public interventions are necessary to bring about longer-term performance improvements.However, the fact that US studies of interventions have also often found that performance improvements are short-term lends credence to our approach.
A broader contribution of our study is that it implies that actors adapt to the governance system they find themselves in, thereby reinforcing the importance of institutional norms and contexts.Thus, although AHF types cut across national boundaries, with particular investment strategies being common in the USA, Europe and beyond, the means of executing these strategies can vary across national contexts.Thus, system norms modify actor behaviour (Jacoby, 2007).If AHFs (and other investors) ignore important norms in their country of operation, they may well be ineffective in their strategies, as found by Buchanan, Chai and Deakin (2020).In our case, AHFs operate 'quietly' rather than 'loudly' (as is the norm in the USA).This reflects the importance of discretion and privacy in the British corporate governance system.
The study has some clear implications for managers of companies mounting takeovers: it is better not to have AHFs on the ownership register, either in the lead-up to the transaction or in the immediate aftermath!Of course, managers of listed companies cannot usually choose their investors and AHF ownership stakes are often unwelcome.However, managers can influence to some degree the types of investors that are attracted to their companies: in some cases in our sample, company managers had clearly invited AHF involvement and in at least one case the company had put itself at the disposal of the AHF as a takeover vehicle and to facilitate industry restructuring.The implication is that managers wanting to avoid AHF involvement should frame their takeover intentions in a way that emphasizes synergies and market growth rather than restructuring and rationalization.Where AHFs are involved in takeovers, managers may need to resist pressures from AHFs for restructuring, especially if managerial judgement suggests that AHFs' objectives could have adverse impacts.They might draw on the support of other longer-term shareholders to assist them in this.
A possible limitation of our study is that we do not report findings on the role of AHFs in target firms.The literature has found that AHFs sometimes pressure targets to be taken over, so that the AHFs benefit from the takeover premium.Equally, AHFs may obstruct takeovers to secure the higher premia typically arising from contested takeovers.In the early stages of our research, we investigated the role of AHFs in the targets, but the results were non-significant.Hence, we have not reported these results.Ideally, we would have wanted more cases of AHF involvement, but importantly the small number of cases is nevertheless associated with significant outcomes.A further limitation is that we do not observe the character of the pressure that the AHFs put on management.It may well be that company managers are influenced by the mere presence of AHFs rather than active pressure from AHFs.Unfortunately, we are unable to determine precisely how far this is the case.There is evidence in our sample that firms are willing conduits for AHF acquisition strategies.This is an inevitable limitation when evaluating the influence of actors in the UK corporate governance system, since governance processes are conducted behind closed doors much of the time.Further research of a qualitative nature is needed to deepen our understanding of how AHFs affect post-takeover employment changes and company performance.

Acquirers with AHF ownership sub-sample
This table reports the mean, median and standard deviation of all variables used in the analyses.The sample includes 421 M&A deals involving UK listed companies during 1990-2017.M&A data are collected from Thomson Reuters database, firm-level data are collected from the Datastream, Filings Expert and Activist Insight databases.Panel A displays data only on AHF ownership in the sub-sample of acquirers with AHFs, whereas Panel B displays data on all variables for the full sample.All variables are defined in Table

Table 2 .
Post-takeover employment change, profitability and productivity Notes: This table presents the results of the univariate analyses on post-merger employment changes (Panel A) and post-merger operating performance (Panel B).Differences between mean scores are tested using t-tests and chi-square tests.Significance levels: *** p < 0.001.

Table 3 .
Determinants of post-takeover workforce downsizing: probit regressions (marginal effects) Notes: This table reports probit regressions estimating the probability of post-merger workforce downsizing.Employment reduction indicates whether there is an employment decline by the end of the first full financial year after the takeover.All regressions include year dummies.Significance levels: * p < 0.05; ** p < 0.01; *** p < 0.001.

Table 4 .
AHF (post-merger AHF)and the size of workforce downsizing: OLS and tobit regressions Notes: This table analyses the effects of AHF ownership (or AHF presence) on post-merger employment change.In Panel A, the variables of interest are AHF and AHF presence.In Panel B, the variables of interest are Post-merger AHF and Post-merger AHF presence.Year and industry dummies are included.The significance levels of the coefficients are based on heteroscedasticity-robust standard errors.Standardized beta coefficients are reported.Significance levels: * p < 0.05; ** p < 0.01; *** p < 0.001.
Year and industry dummies are included.The significance levels of the coefficients are based on heteroscedasticity-robust standard errors.Standardized beta coefficients are reported.

Table 6 .
Post-merger AHF and changes in annual operating performance: OLS regressions Year and industry dummies are included.The significance levels of the coefficients are based on heteroscedasticity-robust standard errors.Standardized beta coefficients are reported.
© 2021 The Authors.British Journal of Management published by John Wiley & Sons Ltd on behalf of British Academy of Management.