5 ms·

I believe the data should add the comparison using purchase power parity. The price in USD being higher in the US than in Brazil doesn’t mean that the internet

by soneca 6y ago

I believe the data should add the comparison using purchase power parity. The price in USD being higher in the US than in Brazil doesn’t mean that the internet is more expensive in the US

goshx 6y ago
Thank you. It is silly to convert everything to USD, unless the list is built to know how much who earns their salary in USD would pay in those countries.
- smabie 6y ago
  PPP converts everything into USD, just using a different method. PPP has nothing to do with a person's salary. If they used PPP, every GB would be the same price for every single country. That's where the "parity" part comes in.
smabie 6y ago
That doesn't make any sense, PPP is used to come up with "real" exchange rates based on price differentials. If you used PPP as the exchange, all countries would have the exact same price per GB. For example, if a GB is $1 in America and 2 euros in Germany, then we would use a 1:2 ratio for converting the price. So Germany's PPP price would now be $1 in dollars as well. As you can see, this isn't very helpful. PPP's fundamental assumption is the law of one price, so it's no surprise that all the prices would be the same using it.
- soneca 6y ago
  It would compare the price relative to other goods in the same country. It makes sense to me, what am I missing?
  - smabie 6y ago
    PPP's fundamental assumption is that a product in one place and the same product in another place have the same price. It's used to determine a relative exchange rate, by setting the price of the goods to be equal. Using PPP methodology for GBs of data, we would first assume that all GBs in every country are the same price. We would then use that fact to come up with a conversion rate between currencies. This isn't very helpful for price comparison, since it's very literally assuming that the prices are equal. Working back from the exchange rate makes much more sense, because we are interesting in the cost difference between a GB here and one there. We can't compare the cost difference if we assume a priori that they must equal each other.
    - soneca 6y ago
      But if you calculate PPP for a basket of goods instead of exclusively for internet GB, it shows the comparison that I would like to see, doesn’t it? I had the impression that using PPP for a basket of goods is how it is usually used.
      - smabie 6y ago
        PPP is used to talk about money. For example, I could compare salaries in different countries to each other and using PPP might give me a better representation than the exchange. If we try and use it to compare goods not in a basket, it's problematic. Everything in the basket is the same price, but stuff out of the basket is now a different price. It doesn't make logical sense with assumptions PPP has. If you used PPP in that way, you're basically saying that the law of one price holds for only the things in the basket, and not for stuff out of the basket. This fundamental distinction between the things in the basket and out of the basket is artificial and doesn't make sense. After all, we could then drastically change the price of something by just adding or removing it from the basket. What you're after is relative cost of a good, normalized by the wealth of an average buyer. This is a fair point, but PPP doesn't do that for your. PPP gives you relative value of money, using the market price and the law of one price.
        nl 6y ago
- goshx 6y ago
  Something is cheap or expensive in each country based on what % of the salary of a person living in those countries a good or service costs. Converting to USD and assuming that it looks cheap to an american is silly.
  - smabie 6y ago
    Sure, that's a fair point, but that has nothing to do with PPP. PPP is about exchange rates, not about goods.
- nl 6y ago
  No, there is a PPP index, eg: https://data.oecd.org/conversion/purchasing-power-parities-ppp.htm https://data.oecd.org/conversion/purchasing-power-parities-p... You would be correct if you created a new PPP index based on GB of data (which is an interesting idea actually), but the proposal is to use an existing one and compare using that.

by soneca 6y ago

I believe the data should add the comparison using purchase power parity. The price in USD being higher in the US than in Brazil doesn’t mean that the internet is more expensive in the US

goshx 6y ago

Thank you. It is silly to convert everything to USD, unless the list is built to know how much who earns their salary in USD would pay in those countries.

smabie 6y ago
PPP converts everything into USD, just using a different method. PPP has nothing to do with a person's salary. If they used PPP, every GB would be the same price for every single country. That's where the "parity" part comes in.

smabie 6y ago

That doesn't make any sense, PPP is used to come up with "real" exchange rates based on price differentials. If you used PPP as the exchange, all countries would have the exact same price per GB. For example, if a GB is $1 in America and 2 euros in Germany, then we would use a 1:2 ratio for converting the price. So Germany's PPP price would now be $1 in dollars as well. As you can see, this isn't very helpful. PPP's fundamental assumption is the law of one price, so it's no surprise that all the prices would be the same using it.

soneca 6y ago
It would compare the price relative to other goods in the same country. It makes sense to me, what am I missing?
- smabie 6y ago
  PPP's fundamental assumption is that a product in one place and the same product in another place have the same price. It's used to determine a relative exchange rate, by setting the price of the goods to be equal. Using PPP methodology for GBs of data, we would first assume that all GBs in every country are the same price. We would then use that fact to come up with a conversion rate between currencies. This isn't very helpful for price comparison, since it's very literally assuming that the prices are equal. Working back from the exchange rate makes much more sense, because we are interesting in the cost difference between a GB here and one there. We can't compare the cost difference if we assume a priori that they must equal each other.
  - soneca 6y ago
    But if you calculate PPP for a basket of goods instead of exclusively for internet GB, it shows the comparison that I would like to see, doesn’t it? I had the impression that using PPP for a basket of goods is how it is usually used.
    - smabie 6y ago
      PPP is used to talk about money. For example, I could compare salaries in different countries to each other and using PPP might give me a better representation than the exchange. If we try and use it to compare goods not in a basket, it's problematic. Everything in the basket is the same price, but stuff out of the basket is now a different price. It doesn't make logical sense with assumptions PPP has. If you used PPP in that way, you're basically saying that the law of one price holds for only the things in the basket, and not for stuff out of the basket. This fundamental distinction between the things in the basket and out of the basket is artificial and doesn't make sense. After all, we could then drastically change the price of something by just adding or removing it from the basket. What you're after is relative cost of a good, normalized by the wealth of an average buyer. This is a fair point, but PPP doesn't do that for your. PPP gives you relative value of money, using the market price and the law of one price.
      - nl 6y ago
goshx 6y ago
Something is cheap or expensive in each country based on what % of the salary of a person living in those countries a good or service costs. Converting to USD and assuming that it looks cheap to an american is silly.
- smabie 6y ago
  Sure, that's a fair point, but that has nothing to do with PPP. PPP is about exchange rates, not about goods.
nl 6y ago
No, there is a PPP index, eg: https://data.oecd.org/conversion/purchasing-power-parities-ppp.htm https://data.oecd.org/conversion/purchasing-power-parities-p... You would be correct if you created a new PPP index based on GB of data (which is an interesting idea actually), but the proposal is to use an existing one and compare using that.