A pulsar wind is a cold relativistic electron-positron outflow originating from a pulsar magnetosphere. Because pulsar winds are radiatively inefficient, it is difficult to obtain their properties, such as the magnetization, the bulk Lorentz factor and the pair multiplicity which corresponds to the number of the particles. In my talk, I introduce the pair multiplicity problem which is independent from the well-known problem of the magnetization called sigma-problem. We study induced Compton scattering off radio pulses by the pulsar wind to constrain pulsar wind properties by imposing that radio pulses are not scattered by the wind. We find that the pair multiplicity of the Crab pulsar can be more than two orders of magnitude larger than the previously inferred value only when the magnetization of the Crab pulsar wind is close to unity at the light cylinder.