Cost Subtlety

[elansegarra]

As it stands, the firm only incurs one cost each period and the current price adjustment heuristic only has them change price if their periodic costs are not fully covered (on average). One unfortunate result is that sellers continue to drop their prices (by undercutting neighbors) overtime. Seems like we need to add some heuristic to allow them to raise their price so prices are more dynamic (perhaps when their sales are "high"?). Furthermore there is currently nothing stopping them from lowering their prices "too low". We probably need to introduce variable and fixed costs so that there is some lower price boundary (like variable cost) which they are much more hesitant to cross.

[ysaikai]

Sounds like it is time to go back to our textbooks and learn just a little bit more sophisticated price competition than the Bertrand model.

[ysaikai]

Here is an idea base on my own experience. Vendors desperately wait for someone (in their neighbor) to raise the price (i.e. expectation for implicit collusion). A trigger for initiative is a signal of "doing well", e.g. 10% profit in t consecutive periods. Once she autonomously raises it, she 'reset' and restart counting "good days" (if any), and all the neighbors react. The process is very dynamic. One implementation may be having "sanguine periods", in which they do not reverse their decisions (either increasing or decreasing price) no matter what the neighbors are doing. That is temporarily ignore collusive opportunities. I guess we need adjust the current undercutting rule accordingly.

[ysaikai]

Another key concept in LFS is authenticity. Simply put it, the larger, the less authentic as local farmers. Some consumers negatively see the adoption of the conventional scaling-up business strategy (becoming a Wal-Mart). So, how about writing the trust dynamics as a negative function of sales? Combined with a bit more sophisticated costs mechanism, we may be able to generate a more interesting and realistic picture.

[elansegarra]

In theory I agree with the idea that scaling up could lead to perceptions of unauthenticity (or "unlocalness"). However, do we think this would manifest in our model? In other words, when one of our local producers has a huge number of sales, are we envisioning that they are becoming corporate or that they just have a long line at their farmer's market stall?

[ysaikai]

There is no magic to reduce costs but technological & managerial innovations and scale economies. And the latter is likely the option that the majority of local farmers think of for serving a long line at their farmers' market stall (esp. a single producer in our model represents some aggregated one). So, I wonder, if no qualitative jump from local to corporate, we may write embeddedness, which is more appropriate now than trust, as a negative function (say, quadratic or sigmoid functions) of sales. This is another effort to make us comfortable with and endogenize embeddedness.

Wittman, H., Beckie, M., & Hergesheimer, C. (2012). Linking Local Food Systems and the Social Economy? Future Roles for Farmers’ Markets in Alberta and British Columbia. Rural Sociology, 77(1), 36–61.