Corporate Sovereignty Without Territory: Perceived Illegitimacy, Pay and Brokered Extraction from the Zaibatsu, 2037–2047

1. Introduction

The largest corporations of the Sprawl are called zaibatsu, and in the years around the Straylight run they were commonly described as something closer to organisms than to firms: vast, self-perpetuating structures whose employees entered young and were expected to stay for life. Several, Maas Biolabs among them, housed senior research staff in company arcologies, and the case files we analyse show that all of the major firms supplied medical care, adjudicated internal disputes and guaranteed the security of their personnel. None of them held territory in the sense a state does. Their jurisdiction ran with their personnel, and a map would not show it.

This arrangement fits neither standard category: corporate theory treats employment as a contract either party may end, and state theory ties sovereignty to territory. The Sprawl itself gave the zaibatsu room to occupy the space between. Work on the Atlanta end of the corridor has found that the old municipalities of the Boston–Atlanta Metropolitan Axis still issue permits, yet the power, transit and dome systems that cross their boundaries are run under agreements negotiated among staff of zaibatsu-affiliated utilities and cleared with their employers. Kessack (2043) described the welfare side of the same arrangement: housing, medical care and pensions delivered by firms to a workforce that no municipality claims.

A second corporate form of the period offers a useful contrast. Studies of Tessier-Ashpool succession describe a closed family line that owned an orbital habitat outright and governed it through whichever members and clones were awake from cryogenic suspension. That authority was dynastic and territorial, bounded by Freeside and Villa Straylight. Zaibatsu authority is institutional and personal in a different sense: it attaches to the employee for the length of a working life, and it is exercised by an impersonal hierarchy that no family owns.

If that authority is governmental, its subjects should behave accordingly. Our test uses extraction, the Sprawl's term for a brokered defection in which a specialist removes an employee from one zaibatsu and delivers him or her to another, usually against the resistance of the origin firm. Extraction is costly and dangerous, which makes the decision to accept an offer a strong revealed preference. We ask whether that decision is driven chiefly by the pay offered or by the employee's judgement that the origin firm has broken its side of a governing bargain. Earlier work on exit from Sprawl institutions (Okonkwo, 2041) and firm-level studies of staff loss at Maas Biolabs (Ishikawa-Brandt, 2042) and Sense/Net (Vireo-Tamm, 2039) described departures but lacked a comparison group of employees who were approached and stayed.

We write in 2048, a little over a decade after the Straylight run. The best-known extraction of the post-Straylight period, the attempted transfer of a senior Maas Biolabs biochip researcher to Hosaka, informed our interview protocol but lies outside our corpus.

2. Methods

Our primary source is a set of case files from three extraction practices, deposited with the Sprawl Institute between 2045 and 2047 under an agreement that removes the names of employees and clients. Each file records an approach: a broker's initial contact with a named employee on behalf of a destination firm. Approaches were made between 2037 and 2047. Four employees had been approached more than once; for each we kept only the first approach, so every employee appears once. We retained every first approach with a completed intake record and a documented decision, giving 103. In 45 the employee accepted the offer and in 58 the employee declined. Of the 45 acceptances, 39 ended in completed extraction and 6 were aborted before transfer.

The primary outcome is the employee's decision: acceptance or refusal. Aborted cases are acceptances, since the decision is recorded even though the transfer failed, and they remain in the primary analysis. As a sensitivity analysis we refitted the model with completed extraction as the outcome, omitting the six aborted cases (n = 97). Drawn from the same files, brokers and years, the decliners form a comparison group from the population actually at risk. Where a completed extraction involved Hosaka as origin or destination, we checked dates and destination against the incident files of Hosaka's internal security division; 21 of 22 such cases matched.

Perceived governance illegitimacy was scored from the broker's intake notes, which record the employee's account of the origin firm before any decision was made. Five items were each scored 0 (absent), 1 (mentioned) or 2 (described as decisive): arbitrary discipline, breach of housing or medical provision, withdrawal of a promised lifetime guarantee, reprisal against family members, and opaque internal adjudication. The sum runs from 0 to 10. Before coding, an archivist at the Institute separated each intake note from its decision log and later correspondence, so that the first author and a research assistant coded all files without access to the outcome. On 30 files coded independently, agreement was good (weighted κ = 0.78), and disagreements were resolved by discussion. The pay differential is the destination firm's offer expressed as a percentage above the employee's current package, including housing and medical value where the file states them.

We fitted a logistic regression of acceptance on illegitimacy score (per point), pay differential (per 10 percentage points), years of service (per 5 years) and origin firm. With 45 events and seven parameters, we used Firth's penalised likelihood to reduce small-sample bias. To compare the two main predictors on a common scale we also report odds ratios per standard deviation, and we tested equality of the standardised coefficients with a Wald test.

Between 2046 and 2047 we held structured interviews with 11 extraction brokers, including principals of the three depositing practices. Interviews covered fee setting, risk assessment, medical screening and routing. Together these brokers had handled 34 of the 39 completed extractions; for each of those 34 we recorded the primary motive the handling broker assigned to it.

3. Results

Of the 103 approaches, 45 (44%) were accepted. Maas Biolabs was the most frequent origin firm, followed by Hosaka, Ono-Sendai and Sense/Net (Table 1). Hosaka was the destination in 15 of the 39 completed extractions. Mean illegitimacy scores were 6.3 (SD 1.9) among those who accepted and 4.1 (SD 1.8) among decliners. The mean pay differential offered was 30% (SD 18) to those who accepted and 26% (SD 17) to decliners.

In the adjusted model, each additional point of perceived illegitimacy raised the odds of acceptance by a factor of 1.55 (95% CI 1.25–1.92, p < .001). Each ten points of pay differential gave an odds ratio of 1.08 (95% CI 0.93–1.26, p = .32), an interval that includes 1. Per standard deviation (2.2 points of illegitimacy; 18 percentage points of pay), the odds ratios were 2.57 (1.62–4.06) for illegitimacy and 1.15 (0.88–1.52) for pay, and the Wald test rejected equality of the two standardised coefficients (p = .003). Years of service showed no clear association (OR per 5 years 0.91, 0.70–1.18), and the four origin-firm terms were jointly non-significant (χ²(4) = 3.1, p = .54). A reduced model containing only the two main predictors gave an illegitimacy odds ratio of 1.52 (1.25–1.85).

With completed extraction as the outcome (39 completed, 58 declined), the illegitimacy odds ratio was 1.62 (1.29–2.04) and the pay odds ratio 1.11 (0.94–1.31), so the conclusions do not depend on how aborted cases are classified.

Brokers' own attributions point the same way. Among the 34 completed extractions handled by an interviewed broker, the broker named a governance grievance as the primary motive in 21 (62%), pay or position in 8 (24%) and other reasons, mostly research freedom or family relocation, in 5 (15%). Of the individual items, withdrawal of a lifetime guarantee and reprisal against family members were the most often scored as decisive.

The brokers had been in practice for a median of 9 years (range 4–17). Nine of the 11 graded the expected response of the origin firm on a shared three-tier scale, from administrative protest through covert recovery attempts to lethal retaliation, and all nine priced their fees by that grade; the remaining two set flat fees. Eight required medical screening of the employee before transfer, chiefly to detect toxin sacs or other coercive implants bonded into the body by the origin firm; screening took place in Chiba black clinics, and the clinic registry records explantation of such devices. Seven brokers routed at least some transfers through Freeside, whose banking and transit status made it a neutral waypoint.

4. Discussion

Among employees who received a credible offer, the judgement that the origin firm had broken its governing bargain was associated with acceptance, and the size of the offer was not detectably so. Pay mattered to some individuals, as the eight extractions that brokers attributed to pay or position show, but across the corpus it did not separate those who accepted from those who stayed. That is the pattern expected of subjects assessing a government's legitimacy, and it is hard to reconcile with a purely contractual account of zaibatsu employment.

The items that weighed most heavily are telling. Withdrawal of a lifetime guarantee and reprisal against kin are breaches of protection, the function most closely associated with sovereign authority. An employee who has lost confidence in the firm's protection, or fears the firm itself, has reason to seek another sovereign.

In this respect the zaibatsu govern in a way that differs from the Tessier-Ashpool model. As the succession literature on that family shows, its authority was bounded by its habitat and passed through its bloodline, and its subjects were residents and staff of Freeside. Zaibatsu authority travels with the employee and changes hands only through extraction. Loss of legitimacy in the family model threatened succession within one line; in the zaibatsu model it produces exit to a rival.

Extraction practice itself fits this reading. A shared risk grade, fees tied to it and routine screening for coercive implants indicate a trade organised around the expectation that origin firms will enforce their claims on personnel much as a state pursues a defector. Whether these conventions have been stable over a longer period cannot be established from a single round of interviews.

5. Limitations

All cases come from the files of practices willing to deposit them, and those practices may differ from the wider trade. Approaches that never reached intake, and extractions arranged without a broker, are absent. The illegitimacy score rests on the employee's account as recorded by a broker with a commercial interest in the outcome, although the notes predate the decision.

Survivorship remains a concern. Employees recovered or killed by the origin firm during an attempt are represented by the six aborted cases only when the broker's file survived; attempts that ended with the loss of the broker as well leave no record. With 45 events the model estimates are imprecise, and the null result for pay excludes a large effect without excluding a modest one. The brokers' attributions were made in 2046–2047, up to a decade after some of the transfers, by people who knew how each case had ended and had handled only 34 of the 39 completed extractions. They are a weak independent check on the intake-based score and should be read as corroboration.

The second author is employed by Hosaka, which was the origin firm in 22 of the 103 approaches and the origin or destination in 22 of the 39 completed extractions. Hosaka provided access to its security files for verification on condition that the employees concerned remain unidentified, had no role in coding or analysis, and did not review the manuscript. Readers should nonetheless weigh the possibility that access shaped which Hosaka cases could be verified.