Self-Employment as a Signal: Career Concerns with Hidden Firm Performance Journal: Working Paper
Georgy Lukyanov, Konstantin Popov, Shubh Lashkery · 2025
Read Self-Employment as a Signal: Career Concerns with Hidden Firm Performance Journal: Working Paper by Georgy Lukyanov, Konstantin Popov, Shubh Lashkery (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-01265.txt.
Introduction
Choosing self-employment can say something about a worker's talent, because the results become public, while firm jobs hide individual performance. Georgy Lukyanov, Konstantin Popov, and Shubh Lashkery study that signaling role in a steady labor market. Risk-averse workers know their own lasting talent and choose each period which path to take.
Because workers decide whether to add another public result or apply to a firm, both the record holders and the applicant pool are shaped by those choices. The authors build market beliefs in two stages, from the flow of types through each history, then from the current choice. They show an equilibrium exists under a stated condition, with talent cutoffs guiding choice at each record. That equilibrium need not be unique, and firm work persists under stated conditions.
Catalog record
| CREATOR | Georgy Lukyanov, Konstantin Popov, Shubh Lashkery |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-01265
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Self-Employment as a Signal: Career Concerns with Hidden Firm Performance Journal: Working Paper Georgy Lukyanov, Konstantin Popov, Shubh Lashkery Abstract We analyze a dynamic labor market in which a worker with career concerns chooses each period between (i) self-employment that makes output publicly observable and (ii) employment at a firm that pays a flat wage but keeps individual performance hidden from outside observers. Output is binary and the worker is risk averse. The worker values future opportunities through a reputation for talent; firms may be benchmark (myopic) (ignoring the informational content of an application) or equilibrium (updating beliefs from the very act of applying). Three forces shape equilibrium: an insurance–information trade-off, selection by reputation, and inference from application decisions. We show that (i) an absorbing employment region exists in which low-reputation workers strictly prefer the firm’s insurance and optimally cease producing public information; (ii) sufficiently strong reputation triggers self-employment in order to generate public signals and preserve future outside options; and (iii) with equilibrium firms, application choices act as signals that shift hiring thresholds and wages even when in-firm performance remains opaque. Comparative statics deliver sharp, testable predictions for the prevalence of self-employment, the cyclicality of switching, and wage dynamics across markets with different degrees of performance transparency. The framework links classic career-concerns models to contemporary environments in which some tasks generate portable, public histories while firm tasks remain unobserved by the outside market (e.g., open-source contributions, freelancing platforms, or sales roles with standardized public metrics). Our results rationalize recent empirical findings on the value of public performance records and illuminate when opacity inside firms dampens or amplifies reputational incentives. JEL D82 , D83 , J24 , J31 , L26 , M51 1 Introduction Markets often see some but not all of what workers do. 1 1 1 Think open-source commits and package releases for developers, contest entries or preprints for researchers, portfolio pieces for designers, client reviews for freelancers, and seller ratings on platforms. A developer’s open-source commits, a freelancer’s platform ratings, or a salesperson’s client-visible metrics are public; most day-to-day contributions inside firms are not. We study a dynamic environment in which a risk-averse worker cares about both current consumption and a reputation for talent that shapes future wages. Each period the worker chooses between self-employment (publicly observed binary output) and hidden firm employment (a flat wage, with individual performance not disclosed to the outside market). 2 2 2 “Hidden” means the market does not see an informative individual-level signal. Inside the firm, managers may observe rich performance signals; we discuss partial public observability in Section 7 . The market maintains a public belief about the worker’s talent summarized by a Beta posterior; self-employment updates this public state via observed successes or failures, while hidden employment leaves it unchanged. 3 3 3 Conjugacy keeps the public state finite-dimensional. None of the comparative statics requires Beta priors; any regular one-dimensional prior with monotone likelihood ratio yields the same single-crossing logic. Firms post flat wages. In a benchmark (myopic) case, wages are set equal to the public posterior mean, ignoring the information in application decisions. In the equilibrium case, firms infer from who is expected to apply and post the mean talent of the implied applicant pool. This setting captures a common choice in modern labor markets: whether to invest in a public track record that travels across jobs, or to accept an opaque wage inside a firm. The trade-off is first-order. Self-employment exposes the worker to risk but creates portable information that improves future outside options; hidden employment insures current consumption but slows market learning. The paper’s contribution is to characterize the dynamic interaction of this insurance–information trade-off with selection into public production and with firms’ inference from applications, delivering sharp comparative statics and testable implications for occupations with portable public histories. The worker’s problem has a simple structure. Because current public production raises expected utility more for higher-ability workers and because the posted wage is type-independent in a period, the difference between choosing public production and hidden employment is strictly increasing in talent and strictly decreasing in the posted wage. We formalize this single-crossing property in Lemma 1 . 4 4 4 The result survives multi-period wages that are independent of contemporaneous individual output and extends to strictly concave u beyond our normalizations. The only knife-edge failure is when current wages do condition on individual output publicly, which restores public learning inside the firm. It implies cutoff policies in the public state and underpins all subsequent results. Our first main result is an absorbing hidden-employment region . In the benchmark (myopic) case, the posted wage equals the public posterior mean. If a worker below the cutoff takes a hidden job, the public state does not change next period, so the same comparison repeats—hidden employment is absorbing (Proposition 2 ). The same logic extends to equilibrium posting, where wages account for application-based inference (Proposition 1 ). Intuition: for sufficiently low public beliefs, the insurance value of a flat wage dominates the informational value of producing another public signal, and because hidden employment generates no public updates, the choice becomes self-confirming. Our second result establishes existence and uniqueness of the equilibrium cutoff when firms infer from applications. Equilibrium posting ties the wage to acceptance: the posted wage equals the truncated-posterior mean of the types who apply, and the cutoff solves the fixed point. Existence and uniqueness follow from the monotonicity in Lemma 1 : the applicant-pool mean is strictly increasing in the cutoff, while indifference is strictly decreasing in the wage (Proposition 4 ). Intuition: adding marginally stronger applicants raises the equilibrium wage, which makes employment relatively more attractive for slightly weaker types, pinning down a unique fixed point. Our third result compares equilibrium to the benchmark. Because equilibrium wages price the selected applicant pool, they are (weakly) below benchmark wages at the same public state. Consequently, the equilibrium acceptance cutoff is (weakly) lower (Proposition 5 ). Intuition: inference from applications corrects the myopic benchmark for adverse selection, compressing wages and tightening self-selection. The framework yields transparent empirical implications. Early in careers—when a single public outcome moves beliefs more—stronger workers should be more willing to generate public signals; later, absorbing hidden employment becomes more prevalent. Transitions into hidden employment should be more common after runs of low public realizations. When employers price the applicant pool, wage dispersion after observed successes and failures should be compressed relative to a myopic benchmark. Roadmap. Section 2 reviews related work. Section 3 presents the model. Section 4 characterizes equilibrium under benchmark (myopic) and equilibrium firms and establishes the absorbing-employment result. Section 6 studies comparative statics and testable implications. Section 7 discusses extensions and empirical connections. Section 8 concludes. 2 Related Literature Our analysis speaks to facts from modern labor markets. Online platforms and open-source ecosystems provide portable, public records of performance; causal evidence shows that public signals substantially improve workers’ subsequent outcomes ( Pallais 2014 ; Stanton and Thomas 2016 ; Horton 2017 ) . At the same time, much productive activity is not externally visible, and employers gradually learn about workers through noisy, partly private signals ( Altonji and Pierret 2001 ) . By allowing workers to choose between public and hidden production