Myths of Sifting

Sifting flour has long been a mystery of sorts to me. When I was very young my mother would make cookies, and she would sometimes let me sift the flour, which I always thought was great fun. She had a sifter with a handle trigger you'd pull, which would rotate a single flat rayed disk over a screen on the bottom of the sifter and move the flour within it across the screen so it would drop into the bowl. Assuming you had it over a bowl, which was definitely the best plan.

I always wondered, though, what was the point? She would explain, "It breaks up the lumps." What lumps? I didn't see any lumps.

One day, along came something new. She brought home a bag of flour with "Pre-Sifted" written on it. Never before had that word appeared in our home. Whether it was a new invention in the 60s, I don't know, but I didn't recall ever seeing pre-sifted flour before. I asked her what that meant, and she said it was already sifted, so it didn't have to be sifted again. "Darn," I thought. I would miss the fun.

Since sifting seemed so purposeless anyway, I wasn't convinced that it wasn't important to sift any more. I understood no purpose for it, but supposedly it had some purpose, so I didn't have a basis for being confident that it was no longer needed. So what if it was sifted at the mill? Maybe it needed to be sifted again at home. How could I know? So as I grew older, I would buy pre-sifted flour, but sometimes I would sift it anyway, just for good measure - figuring it was probably a waste of time, but doing it because it was kind of fun anyway. Then the sifter went the way many items do when you move way too many times, and I just skipped the sifting entirely.

As an adult, I realized that "breaking up lumps" essentially meant separating grains of flour from each other. The question was, if there were no lumps, were grains still separated, and did it affect the cookie or whatever else I was making? It was already sifted before it was packaged anyway, and it all sounds pretty esoteric, so I assumed not. I decided to make sure of that, though.

If grains are greatly separated from each other through sifting, that would mean a given quantity of flour would occupy a larger volume. We're all familiar with the idea of leveling off a measuring cup with a straight edge. I tried a simple test - measure out one level cup of presifted flour, sift that cup of flour, and place the results back in the one-cup measure. If it was still a level cup, then the presifting really was sufficient and there should be no point to sifting at home.
I was shocked, I say shocked, at the result! The cup containing flour mounded on top shown in the attached photo is what I wound up with. It started as one level cup, but after sifting it became far more than one level cup! That was a revelation. I was overjoyed! Now I have a good reason to sift flour again. :-) There is no way anyone can say that the mill's presifting produces the same results as sifting at home. And of course, if you are measuring level cups of flour out for your recipe, the quantities of flour you add to the dough will be vastly different depending on whether you sift the flour at home, or not - and the amount of flour you add has a big effect on your cookie results. I went straightaway to buy a new sifter.

I'm fibbing a little. Actually, I bought the new sifter before I did the experiment, or I couldn't have done the test. And in fact, after that little test I bought two of them, shown in the other photo.

The two sifters are of different types. The one on the left contains a rotating wire ring, which you work by turning a crank on the side of the sifter. It has the brand name "Good Cook" on the handle, and it cost me $6.99 at the local supermarket.The one on the right is a fancy version of the handle-trigger sifter my mother had. This version has not one screen on the bottom, but three, and the one trigger works three metal devices over the screens simultaneously. It seems obvious that pushing flour through three screens is like tripling the amount of sifting with one pass through the sifter. I thought it would have to be an improvement over the crank sifter. It is manufactured by Danesco and cost me $9.99 on sale at a local store, Orange Tree Imports. I saw a similar one for the same price at the local Bed, Bath & Beyond. I saw but did not purchase a crank sifter with an extra revolving ring at Williams-Sonoma. With an extra ring, the flour might possibly go through the sifter twice as quickly for each turn of the crank, but I didn't have a reason to believe the results would be any different from the crank sifter I'd bought, so I didn't try it out. Besides, I like sifting, so why make it happen twice as fast? :-) Seriously, sifting usually happens in a few seconds anyway, so I didn't see a good reason to spend extra money to buy one designed to make it happen faster. The sifter at Williams-Sonoma was sturdier than the crank sifter I purchased but also more expensive, and since I'd already bought my crank sifter I let it go. If I was buying another one, though, I might go for the Williams-Sonoma sifter because it is a bit sturdier.

I entitled this post "Myths of Sifting." Actually, I don't know any myths of cooking, cooking not being one of my normal topics of conversation with friends. But there is no shortage of information about this kind of thing on the Internet, and on one Web page with a glossary of terms used to describe flour, I saw "Presifted - sifted at the factory so it does not need to be sifted at home before use." At this point, also recalling what my mother told me, I think I do have to call this a myth. If a recipe was originally constructed using sifted flour, then it is imperative to sift presifted flour at home because the sifting at the factory definitely does not give the same results for measurement purposes. And if the recipe wasn't created using flour sifted in the kitchen, then the measurement is not reproducible (see below).

To emphasize the lack of understanding of sifting, I will mention that the sifter sold by Bed, Bath & Beyond has measurements on the inside. You measure out your flour in the sifter, and then sift the flour. Convenient? Sure! Good measurements? No way! Very bad idea. If you measure before sifting, you're getting an incorrect measurement because the volume will change as it passes through the seive. First sift, then measure. Who designed that product, anyway? Nothing wrong with using it for a sifter, but if you have one like that, forget the measurement markings on the sifter and measure the flour after sifting. And don't pay extra for the volume markings on the side. :-)

Several other questions come up immediately after my experiment.

1. If you sift flour once, does that give you a reproducible way to measure that flour, or does the volume increase further if you sift it again, and again, until it reaches a plateau?
2. Are all flours affected the same way by sifting?
3. Do all flours have the same volume after they are maximally sifted?
4. Do all flours change volume by roughly equal amounts when you sift them?
5. Does the design of the sifter matter to the result?
6. Does the flour actually have to go through a sifter, or can you just stir it up in a bowl?

To answer these questions, I did a grand sifting experiment with the eight flours I purchased. For each flour, I did this:

1. Weigh one level cup of the flour scooped up straight from the bag.
2. Sift the flour with the crank sifter.
3. Weigh one level cup of the flour after sifting once.
4. Sift the flour with the crank sifter a second time.
5. Weigh one level cup of the flour after the second sifting
6. Sift the flour with the crank sifter a third time.
7. Weigh one level cup of the flour after the third sifting
8. Sift a cup of flour scooped straight from the bag through the trigger handle triple-sifter
9. Weigh one level cup of the flour after the trigger handle triple-sifter.

Here are my results with the various flours. The data needs a table, and I'd create one for it if only I could find my book on HTML, which I know I've seen around here lately somewhere. Maybe later, but meanwhile I hope you can process it in this format (numbers are weights of the contents of a level cup in grams):

Shurfine All-Purpose Flour
152 (step 1)
123 121 120 120 (steps 3, 5, 7, and a fourth pass through the crank sifter)
118 (step 9)

Pillsbury Best Unbleached All Purpose Flour
155
118 119 117
117

Pillsbury Best All Purpose Flour
161
117 117 117
115

Gold Medal All-Purpose Flour
157
123 126 125
122

Dakota Maid All-Purpose Flour (milled by North Dakota Mill)
162
132 134 133
129

Robin Hood Premium All-Purpose Flour
147
117 116 115
115

Hodgson Mill Premium Naturally White Flour
150
112 112 111
111

King Arthur 100% Organic All-Purpose Flour
153
122 122 124
123

A word on the accuracy of the measurements, before any further discussion. I noticed in one case that a measurement seemed way too low, and then found there was a fairly large air pocket in the cup, and with that gone a lot more flour went into the level cup so it weighed a lot more. That was with a cup taken straight from the bag. It seemed like the flour filled spaces more freely after the particles were separated by sifting - especially because after sifting I was placing the flour in the cup a spoonful at a time, like you're supposed to - but I can't swear to the accuracy of the measurements from step 1 because of the possibility of air pockets. I could have taken duplicate measurements, but didn't.

That being said, this data answered a number of my questions.

1. After you sift once, the flour grains are as separated as they are going to get, as determined by measuring the resulting volume. Additional sifting produces insignificant changes on the order of less than 2% in either direction.

2. Every brand and type of all-purpose flour from a bag purchased off the store shelf increases greatly in volume by sifting. The volumes of flours in this experiment increased from 22% - 34%. At the higher end, not sifting the flour in a recipe that calls for three cups of flour would be like adding a fourth cup of sifted flour to the recipe, which would make a gigantic difference in the baked cookie. Even the low end is well past the point where a person could tell the difference.

3. After the flours have been sifted, whether once or three times, there is a fair amount of variation in volume. The weights of cups of sifted flour for the eight flours were: 111, 115, 115, 115, 117, 118, 122, 123, 132. It's interesting that all the 5-pound bags are the same size, but the volumes are significantly different after sifting. Just eyeballing the numbers, the weights fall in three groups - 111, 115-118, and 122-132. Both the lightest and the heaviest were unbleached all-purpose flours, so being unbleached doesn't affect sifted volume or weight. All the bags of flour state 3% protein (gluten) per quarter-cup except for the Dakota Maid and King Arthur flours, which state 4% protein. The latter two just happen to be the two that are the heaviest after sifting, and so I tentatively conclude that the weight of sifted flour in one cup increases with increasing gluten content. Since gluten generally makes flour stick together, that sort of makes sense. That implies that after sifting, one might want to use a little less than a cup of flour per stated cup in the recipe if you are using flour with higher gluten content, in order to maintain an equal weight of flour in the recipe. But it is possible that the effect of gluten is broader than this in cookies, and this may not really be the rule to go by. Additionally, who knows the gluten content of the flour originally used to make the recipe? At some point I will test the effect of using equivalent weights of flour versus equivalent measured volumes. I can't guess the reason Hodgson Mill flour is so light, unless it has the least gluten of any of these. It has the same 3% protein measurement as most of the other flours, but since whole numbers are used for the reporting on the bags, there is room for large variation between 2.5% and 3.499999999...%. The two most expensive flours are the King Arthur and the Hodgson Mill flours, and they are at opposite ends of the weight-per-cup spectrum, which is interesting.

4. Different flours do not increase in volume by similar amounts. I observed a range of percent volume increase from 22% to 34%. The amount of increase may depend heavily on how compressed the bags become during shipping, and it makes sense that the amount of increase would be fairly unpredictable - except that mere sifting is probably never going to change the volume enough that unsifted flour enough for forty loaves of bread can be made to feed thousands after sifting. Something more would be needed. :-)

5. The design of the sifter doesn't matter enough to worry about. Whether you use a crank sifter or a hand-trigger triple sieve sifter, you get about the same volume after sifting. You can observe that the triple-sieve sifter did give slightly lower weights after sifting than did sifting with the single-sieve crank sifter three times, indicating increased grain separation. However, the difference was generally less than 2%. Only with the Dakota Maid flour was there more difference. If you think 4% less sifted flour per cup by weight is enough to be concerned about, and you convince yourself that your flour sifts that much better with a triple-sieve sifter, go for it! Otherwise, reproducibility is the major thing, and sifting once with either kind of sifter gives a weight per cup that is a constant for the bag of flour you are using, and hopefully fairly constant between bags over time no matter how much or how little the flour got compressed during shipping to the store. I should consider that there may be more to sifting results than the end volume, and probably I should test the effect of the two sifters by looking at the cookies that result from using them. I guess I will eventually.

That left the last question - do you actually have to use a sifter? I would have said "Of course," but a couple of days earlier I'd seen an Internet site that described the "textbook method" of sifting as "Stir the flour in a bowl, then spoon it into the measuring cup and level." Stir in a bowl? How is that sifting? It sounded pretty inferior to me, but I checked it out.

Using the Robin Hood All-Purpose Flour, I dipped into the bag with the measuring cup and put a heaping cup of flour in a bowl. I rapidly ran a whisk through the flour in the bowl for 30 seconds. Then I spooned the flour into a cup and leveled it, and finally weighed the level cup. A level cup of the flour after sifting once with the crank sifter weighed 117 grams. The level cup of flour I "sifted" by stirring weighed - 117 grams! I'll be darned, just stirring in the flour in a bowl with a whisk was exactly the same as sifting! I tried it again, stirring for only 20 seconds with a whisk. Again 117 grams! I tried one more time, stirring for only 10 seconds. 116 grams!

So in fact, you don't need to buy a sifter to "sift" your sifted flour. All you have to do is stir it in a bowl! For me, another myth busted. Lots of readers will have already known it. What can I say, I'm ignorant - but not about that any more.

I think a whisk is ideal for stirring the flour in a bowl. The multiple wires cut the flour in numerous places, while the space between wires lets the flour through so you don't wind up with a cloud of flour in the air from your stirring. I didn't try it with a spoon, but I know it would be a pain.

But I read something else only today that lets me know that sometimes a sifter really is useful. Somebody was complaining about having bought some dry, hard flour in the store, and they said it had numerous large crumbs that wouldn't break up easily. Stirring in a bowl won't get rid of those hard balls of flour, if there are any. A sifter, because it puts the whole quantity of flour through a sieve, will either break up or filter out all but the tiniest unbreakable balls of flour. I didn't see anything like that in any of the packages of flour I opened, and based on that I would say if I find that in any flour I just bought at the store, as this women had, I got a bad bag of flour, and I wouldn't use it at all. Flour has an expiration date and maybe this woman had bought some flour past the expiration date. But the point is, if you're only stirring the flour in a bowl, you might not even see unbreakable lumps of flour in the bowl, if there are any. Yay, I still have a good excuse for me to continue to sift flour using my sifter! Probably stirring in a bowl is really good enough, though. It's definitely less fuss, unless maybe if you're stirring with a spoon and send a cloud of flour flying into the air or spill over the edge of the bowl.

Besides making the points that sifting flour at home yields a reproducible measurement and that doing so affects the amount of flour in a level cup quite a lot, I want to be sure to mention that if we don't sift flour before using it at home, I'm sure the measurement is not reproducible. It makes sense that if flour is denser in the bag because it gets compressed during shipping or storage, then different conditions of shipping or storage will compress the flour to different degrees. Thus, all the level cups of unsifted flour from a single bag might hold the same amount of flour, but if you buy another bag of the same brand and type a couple of months later, a level cup of the unsifted flour would hold a different amount because something had changed in shipping or storage - perhaps that bag was at the bottom of a stack of bags on a pallet instead of at the top, etc. Once the flour is sifted, that source of variability in ingredient measurement should be gone - no matter how compressed the flour in a bag is or is not, a cup of sifted flour from any two bags by the same manufacturer should weigh the same.

Of course, I'm just saying that. Is that worth testing and proving out? I think so, because it's a significant statement. Most likely two bags I buy at the same time will have been compressed similar amounts, so I'll go buy a bag of flour and set it aside, and buy another one later to check against to see if sifting really increases reproducibility of measurements of flour from bags from the same manufacturer which were compressed to different degrees.

Something else I want to check out is whether flour from different bags from the same manufacturer will always weigh the same after sifting, issues of compression aside. If I bought two bags at the same time I would think they would weigh the same on a per cup basis because they were probably ground at the same time, and all the parameters of grinding would have been the same. What if I buy the two bags a month apart, though? Or six months apart? I'll try to check this out over time with each of these brands of flour, and report what I find at a later date. This is pretty important information. If a cup of sifted flour from a manufacturer has different weights at different times depending on what bag you take it from, you cannot achieve reproducible results through volume measurements - you'd always have to weigh out your flour, and even that might not be good enough because of potential moisture and thus weight loss.

The last thing I want to mention is about my experience with the sifters. I love the idea of the triple-sieve trigger-handle sifter, but I found it was a big hassle to use. It took two or three times as long for the flour to go through. The three screens one under the other trapped flour between them, and I had a devil of a time getting the last of the flour out from between them. The rayed devices rotating over the screens didn't do a good job of scraping flour at the very edges of the screen, so after the flour had been sifted through the screen there was a "ring" of flour around the edge of the inside, and I had to turn the sifter upside down and tap it several times when I was done to try to get all the flour out. And of course both sifters have a thin coating of flour in them after I'm done, and the triple-sieve trigger handle sifter is going to be far more difficult to clean up than the crank sifter. On top of that, the metal guide in the handle at the bottom of the trigger wasn't engineered well enough and so the trigger caught several times and I had to fiddle with it for a second to keep sifting. And the results of sifting are almost identical for both sifters! The last problem might be specific to the particular sifter or that manufacturer's product, but I think the other problems might be inherent to a triple-sieve design. I doubt I'll ever use that sifter again except just for one test of the difference it makes in actual cookies, and barring a surprising result there I sure can't recommend it to anyone. The crank sifter doesn't seem as neat a device, but it works far better. If you know someone you don't like, let me know, maybe I can mail my extra sifter to them.