Search results

Online ride-hailing platforms match drivers with passengers by receiving ride requests from passengers and forwarding them to the nearest driver. In this context, the low acceptance rate of offers by drivers leads to friction in the process of driver and passenger matching. What policies by the platform may increase the acceptance rate and by how much? What factors influence drivers' decisions to accept or reject offers and how much? Are drivers more likely to turn down a ride offer because they know that by rejecting it, they can quickly receive another offer, or do they reject offers due to the availability of outside options? This paper aims to answer such questions using a novel dataset from Tapsi, a ride-hailing platform located in Iran.

The authors specify a structural discrete dynamic programming model to evaluate how drivers decide whether to accept or reject a ride offer. Using this model, the authors quantitatively measure the effect of different policies that increase the acceptance rate. In this model, drivers compare the value of each ride offer with the value of outside options and the value of waiting for better offers before making a decision. The authors use the simulated method of moments (SMM) method to match the dynamic model with the data from Tapsi and estimate the model's parameters.

The authors find that the low driver acceptance rate is mainly due to the availability of a variety of outside options. Therefore, even hiding information from or imposing fines on drivers who reject ride offers cannot motivate drivers to accept more offers and does not affect drivers' welfare by a large amount. The results show that by hiding the information, the average acceptance rate increases by about 1.81 percentage point; while, it is 4.5 percentage points if there were no outside options. Moreover, results show that the imposition of a 10-min delay penalty increases acceptance rate by only 0.07 percentage points.

To answer the questions of the paper, the authors use a novel and new dataset from a ride-hailing company, Tapsi, located in a Middle East country, Iran and specify a structural discrete dynamic programming model to evaluate how drivers decide whether to accept or reject a ride offer. Using this model, the authors quantitatively measure the effect of different policies that could potentially increase the acceptance rate.

Does the long-term growth rate of a firm increase by exporting? If yes, how large is that increase in a developing economy? The paper aims to discuss this issue.

The authors incorporate data from the manufacturing plants in Iran as a developing economy for 2003–2011 to address this question. Using fixed effect panel and propensity score matching method, the authors examine whether exportation can affect a firm’s growth rate to test for the learning to grow hypothesis.

The findings document that: not only the exporters are larger and more productive than non-exporters, but they also grow faster in size and productivity measures as well. Additionally, the authors find that the rise in the growth rate is a short-term phenomenon and it disappears in the second year; meaning that exportation does not have a permanent growth effect. The findings are consistent with a spot effect of learning, compared to a permanent growth engine. Results are robust to different analysis tests.

The authors investigate the learning effect of exporting within recently released firm-level data of a developing country.