• Firm bias against inexperienced workers leads to inefficient unemployment and persistent productivity losses.
• Targeted hiring subsidies that elevate entrants in firm rankings yield the largest welfare gains among all interventions.
• When sufficiently strong, these subsidies bring the economy close to the constrained efficient outcome.

• Coordination friction: Workers and firms underinvest in entry and vacancies without each other.
• Dual policy dominates: Hits the target with 39 bps lower unemployment, 5 bps of GDP lower fiscal cost, and 23 bps higher welfare than one-sided subsidies.
• Dynamic optimal design: Support is hump-shaped; mix rotates worker → firm → worker.
• Welfare threshold: Net gains when environmental benefits exceed 0.87% (consumption-equivalent).

Manufacturing in many economies faces a chicken-and-egg problem: firms create few manufacturing jobs when they expect too small a pool of workers with suitable skills, while workers are reluctant to acquire those skills when manufacturing job opportunities are limited. We analyze optimal design of industrial policy in a two-sector search model with job-specific capital and workers choosing educational tracks. The decentralized equilibrium features three inefficiencies: capital underinvestment from holdup, matching externality, and inefficient sectoral allocation of labor. Under the standard Hosios condition, only the underinvestment distortion remains, implying that policy should focus on investment incentives, and job creation and skill acquisition will endogenously catch up. When Hosios is violated with workers' bargaining powers being too small, an additional wedge distorting workers' sectoral choice emerges and targeted training subsidies become welfare-improving. In quantitative exercise, we show that greater services-trade openness raises manufacturing employment share.

A large empirical literature has linked weather shocks to GDP to estimate future losses from climate change and has obtained a wide range of results. We develop a framework that nests past approaches, distinguishing between modelling assumptions about how changes in climate variables translate into GDP impacts. We show how four prominent model choices in the literature are related to different assumptions about the persistence of impacts and adaptation. We follow recent developments and apply our framework to extreme weather shocks, including extreme heat and harsh droughts, using high-resolution climate projections to obtain long-term GDP impacts. We compare results under the four different modelling assumptions and show that they play an outsized role in determining the size of impacts. Finally, we quantify the role of different sources of uncertainty. The goal is to provide a structured method that can be used for policy-relevant macro-fiscal projections and frameworks.

• Design: Subsidies are most effective when targeted at tasks where green inputs closely match traditional productivity.
• Trade-offs: Welfare gains arise only when the environmental externality exceeds 1.2, with a 4.3% consumption-equivalent threshold.
• Financing: Lump-sum taxes impose the lowest welfare cost, followed by capital-income taxes; labor taxes are most distortionary.