3.1 Political identity measure

The FEC publicly discloses campaign contributions over $200.5 Disclosures in this database contain the name of the donor, the size of the donation, the committee receiving the donation, and the committee's party affiliation. We identify a manager's political orientation by collecting information about their individual contributions to either Republican or Democratic candidates or both. The candidates could be running for the Senate, the House of Representatives, or the Presidency and contributions can include party committees. We do not collect information on firms’ political action committees (PACs) since they tend to contribute to different parties simultaneously (Cooper et al., 2010).

We then use PO to assess the degree of conformity of the two merging firms’ executives’ ideologies. Our key independent variable, PMA, is a dummy variable, which allows us to flag a merger as having a common political identity if both the acquirer and target share the same political values. Hence, PMA would be equal to 1 if the POs of merging firms were greater than 0.25 (i.e., Republican managers in both firms), less than −0.25 in both firms (i.e., Democrat managers in both firms), and 0 otherwise.7

Christensen et al. (2015), Hutton et al. (2014), and Chin et al. (2013) perform various validation tests such as self-reported party affiliation, and they find that such measure truly captures the political orientations of the manager. Moreover, the extant literature in political science confirms that individual political donations are a result of their personal ideologies, which is the basis of our manager-level time-invariant approach in measuring a manager's political identity (Ansolabehere et al., 2003; Ensley, 2009; Francia, 2003; Francia et al., 2005).

3.2 Descriptive statistics

Table 1 shows the breakdown of acquisitions in our sample by year, and Table 2 documents the breakdown by acquirer industry. Both tables report the numbers for the full sample as well as by PMA indicator. We notice that the distribution in the PMA subsample across the years in Table 1 and across industries in Table 2 is representative of the full sample. Of all mergers in our sample, approximately 24% are flagged as PMAs.

Table 3 presents sample summary statistics8. We notice that acquirers in PMAs are smaller firms, have a lower Tobin's Q, and have worse pre-merger stock performance.9 We also notice that PMAs execute larger deals in absolute terms and also relative to their size.

Other trends in the table suggest that PMA deals happen more within-industry,10 are less likely to be tender offers, less likely to pay all cash, and more likely to pay all stock, compared to non-PMA counterparts.

3.3 Entropy balancing

h() is a maximum distance deviation constraint assigned by the researcher and is equal to h(ω_i) = w_i log(ω_i) and m_r represents the n^th order of balance in moment distribution assigned.

Using entropy balancing on the first and second moments, we balance the distribution of all regression covariates using pre-merger levels. This is achieved by assigning weights to observations such that PMA and non-PMA mergers have similar first and second moments along with observable covariate distributions.

4.1 Univariate results

In this section, we start by analyzing how the market responds to acquisitions in which the target and the acquirer share a similar political ideology. We do this by examining the abnormal returns for the two merging firms around the announcement period. We follow the standard event study methodology of Brown and Warner (1985) to calculate the CARs around the announcement date. We use the market-adjusted returns model of Brown and Warner (1985) in which the abnormal return for firm i on day t is equal to firm i ’s return on day t minus the return of the market on day t.11 For the market return, we use the CRSP value-weighted index return. To generate the CARs, we would then sum the daily abnormal returns for the firm over the event window. Various papers have shown that the use of simpler models in event studies yields results that are no different from those of more sophisticated models (Brown & Warner, 1980, 1985; MacKinlay, 1997).

Table 4 reports the CARs for the acquirer, the target, and the combined firm around the acquisition announcement for the full sample and also broken down between firms engaged in PMAs and non-PMAs. The table reports the 7-day window [−3,+3] where the event on day 0 is the acquisition announcement date. In creating the combined entity CARs, we do a weighted-average of the CARs of both the acquirer and the target where the weights are the market value of the firm 4 days before the announcement (Bradley et al., 1988; Kaplan & Weisbach, 1992) and then adjust those weights to reflect the bidder's toehold. By examining the combined entity, we can observe the total economic value of the acquisition for the overall shareholders (Kaplan & Minton, 2006).

We can see from Table 4 that acquirers tend to have their stock prices drop around acquisition announcements. This finding is in line with that of the extant literature on acquirer announcement CARs involving public targets (Fuller et al., 2002) Moreover, the drop observed in an acquisition where firms have common political ideologies is higher than those who do not. Looking at a 7-day window, we can see that acquirers who share similar political beliefs with the target had negative abnormal returns of −3.1%, a figure that is double in non-politically identical acquisitions. This difference in the magnitude of the abnormal returns is statistically significant at the 5% level as can be seen from the reported t-stat. The evidence from the Wilcoxon rank-sum test is even stronger for the 7- day window with a z-value of almost 2.5 providing further support that acquirer CARs are different across PMAs and non-PMAs. Nevertheless, Table 4 provides preliminary evidence that PMAs tend to be more detrimental to the acquirer's shareholders when announced.

We now move to examine the target firms’ return response to the acquisition announcement. Table 4 shows that target CARs were almost as high as 20%, in line with the extant literature. Given our earlier finding that politically identical acquirers have lower CARs than non-politically similar acquirers, it is possible that such acquirers suffer from a more significant wealth transfer to the targets that arises from overpayment. To support such a hypothesis, politically similar targets would have higher CARs than other targets. When looking at the last two columns of Table 4, we can see that targets in PMAs and non-PMAs both observe CARs that range between 18% and about 20%. Furthermore, there is no statistically significant difference between targets’ CARs across both PMAs and non-PMAs. This lack of statistical difference is evidence against the notion that acquirers in PMAs overpay relative to acquirers in non-PMAs. We can, therefore, interpret the findings for the acquirer and target CARs as PMAs being value-destroying as opposed to observing wealth transfer from the acquirer to the target.

Our next step in assessing the market's assessment of the merger announcement, we examine the combined CARs. Table 4 reports combined CARs ranging from 1.1% to 1.2% for the full sample providing evidence that the combined entity has a higher value for all the shareholders involved. This positive combined CAR finding is in line with that of Andrade et al. (2001), Moeller et al. (2004), and Wang and Xie (2009). Looking across politically similar and non-similar acquisitions, we find that the combined CARs are quite close and lack any statistically significant differences. This lack of statistical significance could indicate that the overall shareholders’ gain is similar in both types of acquisitions. In other words, even though politically similar acquirers observed lower CARs, looking at the overall shareholder base of the combined entity, this lower CAR loses its substance.

4.2 Multivariate results

In this section, we further analyze our initial findings that acquirers in PMAs have lower CARs than acquirers in non-PMAs using multivariate analysis that controls for factors driving M&A announcement returns. We use the 7-day acquirer CAR as our dependent variable while the primary independent variable is our PMA measure. Recall that this measure takes a value of 1 if the top executives of the acquirer and target share the same political ideology and 0 otherwise.

Our multivariate setting controls for the various deal and firm characteristics that have been shown to affect merger announcement returns. For deal characteristics, we control for acquirer size and deal size relative to the acquirer (Moeller et al., 2004), acquisition source of financing (Amihud et al., 1990; Travlos, 1987), whether there was a tender offer (Jensen & Ruback, 1983), whether the target is related to the acquirer or not (i.e., a focus vs. a diversifying deal; Morck et al., 1990). For firm characteristics, we control for Tobin's Q, leverage, operating cash flow, and past stock performance (Moeller et al., 2004).12 Corporate governance also has an important impact on mergers and their outcomes (e.g., Carline et al., 2009; Cosh et al., 2006; Masulis et al., 2007). We control for the board's structure, using staggered/classified board indicator as a proxy to control for corporate governance. Staggered boards have been found to be a highly correlated measure with the firm's governance and a very potent anti-takeover measure. (Bebchuk & Cohen, 2005; Bebchuk et al., 2009; Guo et al., 2008).13

It is an observable fact that ideologies, cultures, and even races are to a certain extent geographically clustered for various reasons. These reasons include income inequality, education, environmental, and other factors. Political ideologies are no different Motyl et al. (2014). Furthermore, Wee et al. (2016) find that proximity between the merging firms significantly increases the likelihood of completion as well as an increased likelihood of the bid to receive a positive recommendation from the respective boards. The authors argue that this result is driven by an information channel given the proximity. As such, we control for this channel as one could argue that our measure is picking up geographical proximity instead of political ideology. We follow Uysal et al. (2008) and construct our geographical proximity as a dummy variable, which is equal to 1 if the distance between the headquarters of the target and acquirer is more than 100 km and 0 otherwise. We collect the zip codes of the headquarters as reported in CRSP, and use Google maps Application Programming Interface (API) to extract the latitudes and longitudes at the zip code level to calculate the geodetic distances.14

The standard errors in all of our regressions are adjusted for heteroskedasticity and clustered at the year level.15

Table 5 reports the multivariate regressions for acquirer CARs, target CARs, and combined CARs, in panels A, B, and C, respectively. Column 1 reports the specification with a single PMA variable that does not differentiate between political parties. We can see from panel A, column 1, that the coefficient on PMAs is statistically significant at the 1% level. The coefficient indicates that politically matching acquisitions have almost 2% lower CARs than non-matching acquisitions.

Column 2 repeats the specifications of column 1 while having distinct dummy variables for Republican and Democrat PMAs. This exclusion allows us to understand whether this political similarity effect is pronounced for a certain political orientation or observed across both conservative and liberal orientations. The rationale behind this is that individuals’ decision, as stated earlier, are shaped by their political values and therefore might take different actions based on their openness to new ideas and new experiences and how much they value job security (Atieh et al., 1987; Conover & Feldman, 1981; Gosling et al., 2003; Joe et al., 1977). The shaping of decisions by political values would have a bearing on our setting when considering that Republicans tend to have a higher preference for the familiar and conventional as well as not forthcoming with their trust (Adorno et al., 1950; Carney et al., 2008; Glasgow et al., 1985; Jost et al., 2003; Tomkins, 1963). Hence, we would expect that our findings would be more driven by Republicans than Democrats given their greater stress on familiarity and shared identity. When we examine Democratic and Republican matches each distinctively, we find results that are similar concerning the adverse effect for PMAs, albeit at a slightly lower significance for Democratic PMA. Both types of matches observe statistically and economically lower CARs.

As for the controls’ coefficients, we find, in contrast to Moeller et al. (2004) and Masulis et al. (2007), that Tobin's Q and operating cash flow do not affect the acquirer's CARs. Furthermore, whether the deal offer is a tender offer (i.e., hostile or friendly) does not affect the acquirer CARs as documented by Schwert (2000). We fail, however, in contrast to the extant literature, to find any effect for the acquirer's size, leverage, or its pre-announcement stock run-up. Moreover, no effects were found for deal characteristics such as deal size, method of payment, or type of deal being a focus acquisition or a diversifying one. The other covariate results, however, are in line with the findings of Ishii and Xuan (2014), which use a similar sample of acquisitions done by S&P 1500 firms.

In summary, the results from panel A in Table 5 support our earlier result from Table 4 on the negative relationship between acquirer CARs and the extent of political match between the two management teams even after controlling for various drivers of acquirer announcement returns.

Table 5, panel B, reports the multivariate analysis for target CARs over a 7-day window and PMA as the main regressor. As in Table 4, we fail to find any difference between PMA and non-PMA as indicated by the insignificant coefficient on PMA in column 1. This result takes into account controlling for size, leverage, operating cash flow, staggered board identifier, and pre-announcement stock run-up of the target as well as deal characteristics. The deal characteristics are deal size, method of financing, whether the offer was a tender offer, deal locality, and the industry of the target relative to the industry of the acquirer. The finding is robust to separating the matches along party lines in column 2. This finding means that there is no relationship between target announcement returns and whether the target's management shares the same political values as the acquirer's.

In panel C of Table 5, we repeat our earlier analysis but with the dependent variable being the combined entity of acquirer and target 7-day CARs. Unlike Table 4, we do find a relationship between the combined CARs and PMA. We find that PMAs have a statistically and economically lower significant combined CAR by 1.7%, ceteris paribus, while controlling for the firm and deal characteristics. For firm characteristics, we use the combined acquirer and target assets, Tobin's Q, leverage, staggered board, and operating cash flows. In column 2, which separates PMA along the party lines, we observe evidence that suggests both Republican and Democratic PMAs have lower combined CARs at the 5% and 10% significance levels, respectively. Hence, the multivariate analysis shows evidence consistent with PMAs hurting the combined entity wealth.

The results from the various panels show that mergers in which the TMTs share a similar political orientation seem to damage acquirer shareholders’ wealth. The damage is significant both statistically and economically. Such an adverse reaction to the merger announcement is not felt by target shareholders when the merger is politically matched. Hence, we do not find evidence that this wealth loss by the acquirer is a result of wealth transfer to the target.

In the next section, we analyze the various characteristics of PMAs and explore possible channels that would explain this result.

5.1 Political similarity and takeover premiums

The acquirer's expected return from a merger depends on whether information advantage can be realized in the bidding process. As such, an informed acquirer will price such information in the bidding process and pay an optimal premium to close the deal, compared to an uninformed bidder. Given the trust hypothesis and the better information flow in PMAs, we would expect that such mergers would be better priced. As such, one would expect the takeover premiums in PMAs to be lower. It is also possible that a bargain between the acquirer and the target, where the target is willing to trade a lower premium for executive seats in the newly merged entity, leads to a more moderate takeover premium. In fact, Bradley et al. (1988) suggest that favorable post-acquisition contracts, and other favors extended to target executive post-merger, are different kinds of premium that are unobserved in the takeover offer price. These explanations are not at odds with our findings that a shared political ideology may be at work in facilitating a takeover. Therefore, we test whether PMAs indeed pay lower takeover premiums by running the regressions in Table 6 where the outcome variable is takeover premium.

We follow Schwert (1996) and estimate the takeover premium as the merger offer price (per share) over the target's share price 42 trading days before the merger. We use the offer price information for the target from SDC platinum, and we gather the target's pre- announcement share price from CRSP, with the merger announcement date as the reference when we look up the 42 days prior target share price. Ideally, we would like to measure the right base target price before the offer to accurately measure the takeover premium. This base price, however, is unobservable, and most studies in this literature resort to using 2 or 3 months before the first bid as the target base price (Eckbo, 2009).16 Table 3 shows that the overall sample has a mean takeover premium of 38%. PMAs have a mean takeover premium of 32%, whereas the non-PMA counterparts paid an average of 40%. These numbers are well in line with the numbers reported in the literature survey for takeover premiums (Eckbo, 2009).17

The results we obtain from the regressions are consistent with the hypothesis that PMAs pay smaller takeover premiums, compared to their non-PMA counterparts. On average, mergers where the acquirer and target share the same political orientation pay about 5% less in takeover premiums, and this effect is found only in Republican mergers. In these regressions, we control for known determinants of deal premiums such as the number of bidders, percentage of the transaction paid in cash, whether the deal is a tender offer, locality of the deal, as well as acquirer and target characteristics like size, leverage, Tobin's Q, cash flow to assets, staggered board, and past performance.

5.2 Political similarity and target executive retention

One possible channel that could explain the adverse market reaction and the lower takeover premium is the target managers’ retention (Bargeron et al., 2009; Harford, 2003; Hartzell et al., 2004). The extant literature is split on the effect of managerial retention on mergers and announcement returns with findings ranging from a positive relationship (Matsusaka, 1993), to no relationship (Bargeron et al., 2009; Martin & McConnell, 1991), to a negative relationship (Wulf, 2004).

The preference to work with familiar individuals and those that share similar values could result in a higher retention of the target's executives. The potential for higher managerial retention would go along with the targets’ desire to merge with politically similar firms as they might expect to have a higher chance to maintain their positions with an acquirer who shares their identity and feels more comfortable around them. The target's management concern is probably not unwarranted as they tend to face significant career losses upon the acquisition of their firms (Hartzell et al., 2004). However, any more significant retention of executives based on their sharing of a similar political orientation could be a driver behind the considerable value destruction of PMAs (Ishii & Xuan, 2014). To investigate this hypothesis, we look at the proportion of the target's executives that remain as members of the TMT of the acquirer post-merger.

To check for target's managerial retention for each merger in our sample, we examine the members of the target's TMT 1 year before the merger as reported by Execucomp. We then track these managers and observe the ones still serving as top executives in the new combined entity 1 year post-merger. Our measure of retention is the number of managers from the target's TMT divided by the total number of executives in the combined entity post-merger. Column 1 in Table 7 reports a tobit regression of the target managerial team retention rate as a dependent variable against PMA along with the controls for the usual set of acquirer and deal characteristics. We can see that PMAs have a statistically higher retention rate of the target's executives. To assess the economic magnitude, we compare the coefficient to the mean target retention rate of 0.06. Column 1 reports a PMA coefficient of 0.1677. Therefore, a PMA would have a target management retention rate that is more than 4x greater than the mean retention, ceteris paribus. Column 2 reports the same specification while separating the PMA variable along party lines. We arrive at similar findings with similar economic and statistical significance as in column 1. All regression specifications in this table are tobit regressions and left censored to 0.

To summarize, we find that in PMAs, there is a higher retention rate of the target's management. This result holds across both ends of the political spectrum, conservative and liberal. This finding goes to show that managerial retention is a critical element in mergers where the top managers share a CPI. However, retention of poor managers on the basis of familiarity, political being one of them, could be one channel resulting in the lower value of PMAs.

5.3 Political similarity and operating performance

We have found in Section 4 that markets respond negatively to mergers in which the TMTs share a CPI. One channel for this negative response is that market expects those mergers to perform worse than non-PMAs. To test whether this is true, we examine the operating performance of the merged firms over the 3 years following the merger completion date. We try to assess whether a systematic difference between the performance of merged firms arises due to the CPI of the management of the merging firms. Examining the post-merger performance allows us to separate which of the competing hypotheses regarding PMAs is supported. The positive side of similarity and its associated trust tells us that with a greater information flow and better affinity to work together, PMAs should do better post-merger. However, as discussed earlier, such trust could have its perils in making managers less accurately assess the potential merger and miss out on significant pitfalls. This preference for being in a familiar setting and conforming to the group without adequately assessing the consequences could lead to a deterioration in the post-merger performance.

In assessing post-merger performance, we use the methodology proposed by Ghosh (2001). We first start by finding a match from each of the acquirer's and target's industry where the industry is defined using the two-digit SIC code. The match is based on firm size and operating performance 1 year before the merger (Loughran & Ritter, 1997). Ghosh (2001) defines operating performance as cash flow margin calculated as cash flow in year t divided by sales in year t-1. Cash flow is sales minus the cost of goods sold and selling and administrative expenses plus depreciation and amortization expenses (Healy et al., 1992). After determining the acquirer and target matches, we create the proforma data of the matched firms by aggregating the data of the two matched firms. We then track the performance of the acquirer 3 years after the merger completion date and adjust it annually for the aggregated matched firms’ performance (Franks et al., 1991; Rau & Vermaelen, 1998). Finally, we define the average of the match-adjusted cash flow margin of the acquirer over the 3 years following the merger as the post-merger performance.

Table 8 reports the results of the analysis of post-merger operating performance. Column 1 reports the regression of post-merger performance on PMA along with controls for firm and deal characteristics. We find that PMAs have on average 3.3% lesser cash flow margin in the 3 years following the merger relative to non-PMAs. On the other hand, non-PMAs have an average of 5.78%, and hence PMAs have cash flow margins that are 43% lower than their non-PMAs counterparts. The result is robust across party lines when we split the PMA variable in column 2. Both Republican and Democratic PMAs seem to carry adverse effects on post-acquisition performance with Democratic PMAs having a much lower post-merger performance than Republicans. Therefore, PMAs perform worse in the 3 years following the merger with this underperformance found across both party lines.

This result does not necessarily contradict Elnahas and Kim (2017). While they only examine the political identity of the acquirer, we focus on the identity of both the acquirer and the target in assessing the long-term performance of the acquisition. It is important to stress that we do not find evidence that politically oriented managers are necessarily bad, compared to other managers. Rather, it is when politically oriented managers acquire firms of a similar ideology that we find detrimental long-term performance effects on the combined entity.

The findings in this table go to support the hypothesis that having a common political ideology hurts the merger as they fared worse operationally.

5.4 Political similarity and probability of acquisition

The extant literature suggests that political differences can act as barriers to what could be a logical merger (Hoffman et al., 2007). To further understand the propensity of managers to be with similar managers in terms of their values and ideologies, we investigate the effect of a CPI on the probability of the merger occurring. We try to assess whether mergers are more likely to happen if there was a CPI across the two TMTs. We first start by reporting the proportion of PMAs in our full sample, which is 0.242 from Table 3. That is, 24.2% of the mergers in our full sample were mergers where the top executives had a CPI. We also report the proportion of PMAs along the party lines of Republican and Democrat, which are 21.3% and 4.6%, respectively. When reporting the proportion of PMAs along a particular party line, we remove from the sample mergers matched by the opposing party values. We can see, hence, that most of the PMAs tend to be along conservative (i.e., Republican) lines. This is not surprising given that most of the managers in our sample tend to be conservative. The low incident rate for Democratic mergers is consistent with the numbers seen in the literature and reflects the observed underlying distribution of manager-level donations (Chin et al., 2013; Christensen et al., 2015; Hutton et al., 2014). Hence, it is not surprising to see in our data a lower incident rate for democratic mergers, compared to Republican ones. To gauge whether the observed frequency is systematically higher, we compare it to a random sample of acquisitions.

As a first analysis to assess the likelihood of PMAs occurring, we start by creating two random samples of mergers. In the first sample, we match each acquirer with a random target drawn from the target's industry and merger year. We repeat this step a thousand times and report the proportion of PMAs from those random draws under the column header Random Target–Mean in Table 9. We report the overall PMAs, Republican-matched acquisitions, and Democrat-matched acquisitions. The second step entails matching the targets to a random acquirer from the acquirer's industry and merger year along with repeating the step a thousand times. The proportions of PMAs are reported under the column header Random Acquirer–Mean.

To evaluate whether our observed likelihood of PMAs is higher than what would happen if firms were merged randomly, we test for the statistical significance between our sample's proportion of PMAs and that of the two randomly generated samples. Compared to the first random sample of mergers in which the targets were random, we fail to find any difference in PMAs’ proportions. However, in the sample of only randomized acquirers, we do find statistically significant differences. We find that PMAs occurred 4.7% more than relative to a sample of random acquirers. Similar differences are found when mergers occur along a Republican ideologies line. Interestingly, we find that for Democrat-matched firms, the actual proportion was almost twice that of a sample of random acquirers and targets.

We further investigate this result by conducting a conditional logit model for target and acquirer selection similar to Bena and Li (2014), Kuhnen (2009), and Dyck et al. (2010) using matched case-control observations. Specifically, we analyze the impact of PMAs on the selection of acquirer (target) to the target (acquirer). To do this, we create all possible hypothetical acquirer-target pairs within each industry and year and flag the actual deals that happened using a dummy variable that takes the value of 1 for actual deals and 0 for hypothetical deals. We run a conditional logit regression using this dummy variable as the dependent variable, PMA as our main independent variable of interest, and also control for all observable covariates for either the acquirer (when we randomize the targets) or for the target (when we randomize the acquirer) and report the findings in Table 10.18

Columns 1 and 2 of Table 10 show the coefficient and odd ratio estimates from conditional logit models of the fixed acquirer and a control group of all possible target pairs within that industry and year. Columns 3 and 4 show the opposite, whereby we fix the target and run the model using a control group of all possible acquirer pairs within that industry and year. The results in columns 3 and 4 clearly show more consistency, higher coefficient estimates, and larger magnitudes of odds ratios for the independent covariate PMA. This evidence is consistent with the notion that targets are more likely to choose to be acquired by an acquirer of a similar political ideology than the other way around. All regression specifications include year, industry, and deal-level fixed effects (FE).

The results in general point toward the influence of PMAs on the initiation and completion of mergers. This evidence supports the notion that firms in PMAs tend to miss out on other investment opportunities and push forward, possibly at lower standards, for mergers with politically similar firms. The different findings, nonetheless, across the random acquirers and random targets samples indicates that it is targets that seek politically similar acquirers. This would fit with the career concern notion that target executives face during mergers and would go along with our earlier finding on the higher retention of managers in PMAs. In other words, targets would reach out to firms with similar political background and orientation on the premise that they are more likely to stay on in the new entity after the merger. Moreover, it seems that the effect is observed across both the Republicans and Democrats. It might be surprising to find that Democrats have such a tendency to merge with similarly minded people when considering it is the Republicans who have a higher preference for familiar settings (Jost et al., 2007; Jost et al., 2009). However, liberals’ greater openness and tolerance (Adorno et al., 1950; Tomkins, 1963; Jost et al., 2003) seem to result in a corporate culture within the firm that might not be acceptable to others (Briscoe et al., 2014). This corporate culture, in turn, pushes liberals to prefer to merge with entities that would maintain their culture and not suppress it (Block & Block, 2006; Wilson, 1973). Indeed, 38% of M&A expert survey respondents conducted by Accenture Research identified corporate culture, as measured by behavioral norms and beliefs, as the most critical factor for merger integration.19 Our results in Table 9 confirm these observations.

5.5 Political similarity and acquirer's management post-merger bonus

One source of value destruction in M&As is the improper compensation of CEOs (Grinstein & Hribar, 2004). In this subsection, we examine whether acquirers’ TMTs receive a more significant bonus upon the completion of a PMA. We focus on cash bonuses, in line with Grinstein and Hribar (2004) reporting that bonuses are almost always the form of reward for merger completion. We examine the bonus received by the top five managers of the firm in the year of the merger's completion as reported in Execucomp. We focus on the top five managers to be aligned with our political orientation measure of the firm as the influence of lower managers would be very minimal on the merger's decision and outcomes. Table 11 shows the effect of having a PMA on the management's bonus using a tobit regression. Our dependent variable is the natural logarithm of the sum of the top management's bonuses in the merger completion year. Our primary independent variable is the indicator variable PMA. Column 1 reports the estimation that shows that managers receive bonuses that are approximately 35% higher in PMAs. We get this result after controlling for various acquirer and deal characteristics as well as year and acquirer industry fixed effects. Acquirer characteristics are computed for the year of the bonus award. This effect is significant at the 1% level. Along with Grinstein and Hribar's (2004) findings for CEO bonuses, we find that managers’ bonuses are positively associated with acquirer size and relative deal size. Also, we find that bonuses are higher with higher leverage, lower pre-merger stock performance, cash-financed deal, and diversifying, friendly, and local deals. Column 3 repeats the specification for columns 1 while splitting the PMA variable along party lines. We find that both Republican and Democratic PMAs receive higher bonuses with Democratic PMAs having bonuses that are multiples in size when compared to Republican.

To further understand the findings on managerial bonuses across party lines, we look at the explanatory effect of potential merger performance, measured as the market's response to the merger announcement (Grinstein & Hribar, 2004). This potential explanatory effect is captured by adding an interaction term between the dummy variable PMA and the acquirer's 7-day CARs as in columns 2 and 4. We find in column 2 that executive teams in PMAs still receive bonuses that are significantly higher, and higher market responses to the announcement would lead to a further gain in the bonus, relative to executive teams in non-PMAs as can be seen from the statistically significant interaction effect. The economic magnitude of the effect of PMAs conditional on the potential performance of the merger is notable. A one standard deviation increase in the 7-day acquirer CAR for a PMA would result in an additional increase in the TMT bonus of almost 35%. Similar results are also found in column 4 when we separate PMAs by party. Republican executives in the acquiring firm would see higher bonuses as the announcement CARs went up relative to non-Republican PMAs in the interaction variable, Republican PMA x Acquirer CAR. For Democratic PMAs, the average TMT in a PMA would receive a bonus that is about 136% higher. However, this effect is independent of the merger performance as measured by CARs. Overall, our results indicate that managers in PMAs receive greater bonuses upon the completion of the merger, notwithstanding the adverse market reaction to the merger's announcement. Further, it seems that the better the market's response, the higher the bonus for Republicans but not Democrats who, on average, receive higher bonuses. The potential to be handsomely rewarded, regardless of the shareholder wealth effect, acts as an additional incentive, along with the familiarity preference, for managers to pursue mergers in which the target shares a similar political identity.