The Global Economic Impact of Private Equity Report 2008 - World ...

sectors. In Retail Trade, the cumulative impact of the privateequity transaction after five years yields a 9.6% loweremployment level than would have occurred if the targetshad the same growth rates as the controls. The cumulativefive-year impact is 9.7% lower employment in Services forthe targets compared with the controls. 18Figure 12 shows the variation in the differences betweendifferent types of private equity transactions. There are fewconcerted differences across the categories: each displaysa similar pattern. One exception is the fact that the periodof reduced employment growth is considerably larger andconcentrated immediately after the transaction for thesecondary buyouts, 19 which presumably have alreadyundergone a restructuring under their previous owner. 20By contrast, public-to-private buyouts experiencereduced employment growth in the first two years after thetransaction relative to divisional and independent buyouts.C. Firm-level changes including greenfield entry,acquisitions and divestituresIn our establishment-based analysis in the last two sections,we focused our attention on the ultimate outcomes forestablishments and workers in target firms at the time of thedeal. Alternatively, one could focus on similar outcomes forthe entire target firm. While the analysis contained in the priortwo sections is an appropriate way to trace the employmentimpacts of private equity transactions for establishments andworkers at target firms at the time of the buyout, it ignoresthe opening of new establishments and other actions thatprivate equity firms or other subsequent owners may takethat impact employment at target firms post-buyout.In this final section, we address this shortcoming of theestablishment-based analysis. However, to do this we, bynecessity, restrict our attention to the subset of target firmsthat we can observe for some period post-buyout. With theLBD, we are able to readily follow establishments over time,even if they undergo ownership or other changes. Trackingfirms over time is more problematic due to mergers andother events that lead to changes in the firm-level identifiersin the LBD. This was not an issue in our establishmentanalysis above, since we only needed to find the target firmin the year of the private equity transaction and then follow itsestablishments over time, regardless of any changes in theirassociated firm identifiers.The disappearance of a firm ID in the LBD can be associatedwith a firm’s death, where all the firm’s establishments areshut down, or some form of organizational change such asa merger. It is possible to utilize the LBD’s robust longitudinalestablishment linkages to provide a rich description of theseorganizational changes. Such analyses are, at this point, verytime-consuming and resource-intensive. Thus, we restrict ourfirm-level analysis to a subset of target firms and similarcontrol firms that we can observe two years after the buyout.In order to be able to identify target firms that we canobserve two years after the buyout, we must restrict theset of transactions in this firm-level analysis relative to theestablishment-level analysis given the differences in thematching of the private equity transaction data at theestablishment and firm level. For the establishment-levelanalysis, as discussed in Section 3.B, the matching of theprivate equity transactions to the LBD is based on a 3-yearwindow centred on the transaction year. This is feasible andreasonable because with the establishment-level analysiswe use the information from the Capital IQ data to date thetransaction and, given high-quality establishment longitudinalidentifiers, only use the set of establishments that exist inthat deal year. In contrast, for the firm-level analysis we areforced to restrict attention to matched cases where thematch to the LBD occurs in the transaction year. Thismatching restriction implies that even before restricting ontwo-year survivors, our firm-level analysis starts with about65% of the matched transactions used in the establishmentlevelanalysis. Using this subset, we impose the furtherrestriction that we observe the target firm in the transactionyear and two years later. This latter restriction impliesthat in the firm-level analysis we have about 55% of thematched transactions in the establishment-level analysis(approximately 1,300 transactions). Note, however, thatconditional on matching in the transaction year, this survivorrestriction yields 97% of the employment from all the firmsthat match in the transaction year. Thus, it is the matchingrestriction and not the survivor restriction that matters mostfor our firm-level subset.Given that we track the target firms for two years and finda comparable group of controls, we can compare growthin employment and the number of establishments at targetfirms with similar control firms. In addition, the LBD permits usto overcome many of the challenges that plagued earlier workin that we can identify the components of the firm changesdue to greenfield entry, the closing of the establishmentsowned by the firm, acquisitions of new establishments anddivestitures of establishments owned by the firm. However, itis important to note that, since we restrict attention to firmsthat can be followed for at least two years post-private equitytransaction, the firm-level analysis focuses on a subset of18While not reported in Figure 11, we have also examined the patterns for the FIRE (Fire, Insurance, Real Estate) broad sector. We find even morevolatile patterns for FIRE than for Retail Trade and Services and a very large net difference between targets and controls.19It is important to note the differences in scale for the figure depicting secondary buyouts – we chose to use a different scale given the very largenet negative difference between targets and controls for the first year after the transaction.20In unreported analyses, we examine relative establishment growth rates across age and size classes. There are some differences, but the posttransactionpatterns are quite similar across age and size classes. One notable difference between private equity transactions and controls isthe pre-transaction net growth differences for very young (between 0 and 4 years old) establishments: very young control establishments havesubstantially higher net growth compared with targets prior to the private equity transaction. This pattern likely reflects differences between targetsand controls in the age distribution of the parent firms of very young establishments. That is, among very young establishments, targets are likelypart of older firms that are ripe for restructuring.The Global Economic Impact of Private Equity Report 2008 Large-sample studies: Employment 51