What Does Magnitude Measure, and What Do You Actually Feel?
An earthquake has one magnitude. The USGS (the United States Geological Survey) puts it simply: magnitude is the size of the earthquake, and the shaking it causes takes many different values from place to place, depending on distance, the type of surface material, and other factors [1].
Japan uses a second number for the shaking itself. Its weather agency reports a shindo, or seismic intensity, for each place where an instrument has measured the ground moving. The Japan Meteorological Agency (JMA) explains the difference with a light bulb. Magnitude is how bright the bulb is. Intensity is how bright the light looks at one spot in the room, and a spot far from the bulb is dimmer [2]. The USGS also keeps size and shaking apart in the same way [1].
A simple way to picture it is a loudspeaker at a concert. The speaker's power is one fixed number, but how loud it sounds depends on where you sit, and on what is between you and the stage.
One more number is easy to misread. Each whole step up in magnitude means about 32 times more energy released, so two steps is about 32 × 32, or roughly 1,000 times [3]. That is a statement about the earthquake. It does not mean the shaking at your location is 32 times stronger.
How Do Distance and Depth Weaken the Shaking?
Shaking generally grows with magnitude and fades with distance from the fault that caused it [4]. So a large earthquake far away can shake a town less than a smaller one right beneath it.
Depth works the same way. The USGS says earthquakes occur anywhere from the surface down to about 800 kilometers (about 500 miles), and that shaking at the surface from an earthquake 500 km deep is considerably less than if the same earthquake had happened 20 km down [5]. The page does not give a number for how much weaker, so we should not invent one.
There is a catch. JMA notes that a very large earthquake produces slow, large movements of the ground, called long-period ground motion [6], and slow movements matter a great deal for tall buildings, as the later sections show. So distance and depth help, but they do not settle the story.
Why Does the Ground Under a Town Change How Hard It Shakes?
Even towns at nearly the same distance can feel different shaking. JMA says shaking is strongly affected by the ground and the shape of the land, so the measured intensity can differ even within one city [2]. Its tables also note that the effect of shaking depends on the nature of the motion (its size, its period, and how long it lasts), the type of construction, and the underground conditions [2].
The USGS adds that, when waves travel up from solid rock, the structure of the layers beneath a city, including softer surface rock and thick sediments, can amplify the motion and has a major influence on damage in urban areas [7]. Strong shaking can also make wet soil behave more like a liquid, a process called liquefaction [7].
The sources do not say "soft ground makes shaking exactly this many times stronger." They say the ground matters a lot, and that is as far as we can honestly go.
Why Do Some Buildings Sway More Than Others in the Same Shaking?
Every building has a natural period, the time it takes to swing back and forth once when pushed. JMA explains that when the period of the ground's shaking matches a building's natural period, the building resonates and sways a lot. Taller buildings generally have longer natural periods, so they respond most to the slow waves. JMA adds that upper floors move more than lower floors, and that tall buildings can sway for a long time [6].
A swing on a playground works the same way. Small pushes at the right rhythm build up a big swing, while pushes at the wrong rhythm do very little. (That swing is only a picture to help you imagine resonance.)
Low buildings are a different case. The USGS says the P and S waves mainly cause high-frequency (fast) vibration, while the surface waves that arrive last mainly cause low-frequency (slow) vibration. So which waves matter most depends on how fast the building itself swings. The USGS also says S waves are the most damaging, because buildings are damaged more easily by side-to-side motion than by up-and-down motion [4].
Putting these sources together, it is reasonable to expect that a tall building can be in trouble from slow waves even when it is far from the earthquake, while a low building is tested more by quick shaking close by. The sources do not test that idea on any one real earthquake, so treat it as a way of thinking, not a verdict.
One more detail explains some news reports. JMA's intensity meters are normally on the ground or the first floor of a low building, so the reported intensity is not what people feel on the top floor of a tall one [2].
So What Decides How Much Damage an Earthquake Causes?
Put the clues together and the answer to the title appears. Damage depends on distance and depth, on the ground, on how a building responds, and on how well it was built. The USGS says there is no single magnitude above which damage occurs. It depends on distance, soil type, and building construction, though damage usually does not begin until the magnitude is somewhere above 4 or 5 [8].
The USGS also says that earthquakes in some parts of the world cause more damage and deaths primarily because of buildings poorly designed and constructed for earthquake regions, and because of population density [9].
This article is about shaking. Tsunamis, which only some earthquakes cause, are a separate process and a separate story.
The hopeful part is that people cannot change an earthquake's size, but engineers and communities can make buildings safer. That is why engineers care so much about this question.
How Can You Look Into This Yourself?
Try two rulers
Do this with an adult nearby. Lay two rulers of different lengths so that half of each sticks out over the edge of a table, and press the other half down firmly with your hand so it cannot slip. Flick the ends gently, keep your face well away from the rulers, and never flick hard. The short one buzzes quickly and the long one wobbles slowly, a small model of how buildings of different heights have different natural periods. Real buildings are far more complicated than rulers, so this only gives a feel for the idea.
Look at your own area
Many local governments and geological surveys publish hazard maps showing how the ground and buildings in an area may respond to earthquakes. Search for your city or region and see whether the ground in your neighborhood is rock or loose sediment.
Open the sources
The USGS pages below are written for the general public and are a good place to start. A question that remains open in our reading: why does damage sometimes become extreme in one small area while nearby areas are spared? The sources here point to the ground and the buildings, but none explains each case in detail.